Page 1

ITU A|Z • Vol 12 No 3 • November 2015

Contents Yurdanur Dülgeroğlu Yüksel Editorial


Dossier: Space syntax and architectural design Alper Ünlü Dossier Editorial On space syntax and architectural design: A story of the suitcase Daniel Koch Architecture as Material Discourse: On the spatial formulation of knowledge and ideals in four library extensions



Kerstin Sailer The dynamics and diversity of space use in the British Library


Erincik Edgü Success in Basic Design Studios: Can seat selection be an advantage?


Mehmet Emin Şalgamcıoğlu, Fitnat Cimşit Koş, Ervin Garip Tracing a biennial layout: Experiencing an exhibition layout through the syntactic analysis of Antrepo No. 3 at the 2013 Istanbul Biennial


Nilüfer Kozikoğlu, Pelin Dursun Çebi Thinking and designing with the idea of network in architecture


Esra Özsüt Akan, Alper Ünlü Behavioral responses of the elderly regarding spatial configuration: An elderly care institution case study


Ervin Garip, Mehmet Emin Şalgamcıoğlu, Fitnat Cimşit Koş The influence of architectural configuration on the pedestrian network in Büyük Beşiktaş market


Fitnat Cimşit Koş, Mehmet Emin Şalgamcıoğlu, Ervin Garip A syntactic analysis of social interfaces in Istanbul Biennial patterns in case of biennial buildings in 2013


Nevşet Gül Çanakçıoğlu Can cognitive maps of children be analysed by space syntax?


Suat Apak Disintegration of urban housing areas: Districts and new gated housing settlements


İlgi Toprak, Alper Ünlü A diachronic approach on heterochronic urban space

159- 173

Erincik Edgü, Meray Taluğ, Nezire Özgece Divided shopping: A syntactic approach to consumer behaviour


Ayşe Özbil, Demet Yeşiltepe, Görsev Argın Modeling walkability: The effects of street design, street-network configuration and land-use on pedestrian movement


ITU A|Z • Vol 12 No 3 • November 2015

Theory Harun Ekinoğlu, Gülname Turan Does favorite design lead to good design?: Taxi design competitions in Istanbul and New York City


Alev Erarslan Continuity of architectural traditions in the megaroid buildings of rural Anatolia: The case of Highlands of Phrygia


Özlem Atak, Gülen Çağdaş The reflection of religious diversity and socio-cultural meaning on the spatial configuration of Traditional Kayseri Houses


Ervin Sezgin, Gülden Erkut Cross border cooperation in Edirne-Kırklareli border region: New institutionalist perspectives


Mine Aşcıgil Dincer, Sevtap Yılmaz Modelling road traffic noise annoyance by listening tests



Editorial Yurdanur Dülgeroğlu Yüksel The new issue of A|Z has a special dossier theme on Space Syntax and Architectural Design. It contains a valuable collection of articles, each of which is contributing to the theme as well as to the designers who are interested to use space syntax models as a tool in their architectural design. There seems to be no need to say more to the dossier editor’s full explanation of the individual articles in the collection and history of development of the theme. In this issue, also there are a number of theory articles on varying subjects, mostly involving field study or case study. In the article titled as “Does favorite design lead to good design?: Taxi design competitions in Istanbul and New York City”, Ekinoğlu and Turan discuss the changes in both urban and national bureaucracy in the last decades. With the rejection of the rationalist design methods in the Seventies, due to their setting the goals in a top-down approach in defining the problem; the participatory design approach were found by the scholars to fit into the diverse needs, values, and interests and priorities of the people. Such a participatory decision-making was concretized in two different design competition conducted for taxi design in Istanbul and in NY City during 2011 was set separately. The common aim was to increase the quality of taxi service as a mobile space. However, The Concept of Participation has the controversy when majority’s choice dominates minority’s expectations. Therefore, the article explains this controversy in taxi design competition cases: that is, a good design as defined and agreed by the experts does not necessarily be the most popular design for people, and the most popular design is not necessarily the best-quality design. The article shows us that, this paradox does not underestimate the value of participatory process, as it reveals the tastes and values of the people, no matter how different they may be from those of the experts.

Erarslan’s article, titled as “Continuity of architectural traditions in the Megaroid buildings of rural Anatolia: The Case of Highlands of Phrygia” is an attempt to prove that architectural continuity, despite changing times and societies. The case selected for the study is Anatolia. Phrygian highlands. In the underlining study, it has been found out that the megaroid structures are similar to the historical megara houses,with respect to their plan types, spatial organization of the house, functional layout, construction materials, and systems. Thus, there seems to be a regional memory of rural architecture. This is evidenced by the cultural adaptation of the newcomers and nomads by the former settlers’ cultures. Such an acculturation and cultural adaptation process explains architectural continuity throughout the centuries, despite the transition from nomadic to settled life style. Traditional house patterns and plans, the megaroid buildings in the region, are found to demonstrate not only the regional taste but also the suitability and functionality of these house by their courtyards and sofas. The article sets for an interesting picture and shows how the Anatolian Turkish house has evolved through overlapping cultures and centuries to become a unique architecture of the region. This article, despite the contextual differences, confirm the continuity of certain aspects of culture, to be reflected in the architecture of the houses in the article making Kayseri houses as case. Atak and Çağdaş, in their article “The reflection of religious diversity and socio-cultural meaning on the spatial configuration of Traditional Kayseri Houses” examine two different cultures’ courtyard houses by using a field study. The methodology is space syntax and visibility graph analyses to inquire into inwardness-outwardness, the determination of spatial privacy, control, social hierarchy within the household, and the degree of relations between the household and visitors. These methods are constructed through the relations of permeability (based on movement) and visibility (based on the perception of a moving observer). In addition, while a very


significant proportion of the houses exhibit outward-looking structures, in terms of accessibility relations. Only a small portion exhibit inward-looking structures. The dominancy of extraversion is expressed to be correlated with the extroversion trend in the social structure in the last two centuries. The authors conclude from this study that the courtyard plays a major role in connecting all units of the house to the outer space; and by structuring these spaces of the house. Permeability and visibility relations reveal the spatial mechanisms of houses and how the dwellers and visitors experience these mechanisms. ln the fieldwork, the spatial organization of traditional Kayseri houses are explained by their social and cultural factors. It is found that the permeability and visibility structures of the inner spaces of houses function together. Courtyards, gardens and central halls are predominantly more visually integrated spaces. This study by having systematically examined the accessibility and visibility structures of the traditional Kayseri houses, can guide other studies in future to be conducted a new,on similar spatial organizations. “Cross border cooperation in Edirne-Kırklareli border region: New institutionalist perspectives” by Sezgin and Erkut elaborate on the border issues, especially after the Cold War. They are justifying their study by the expanding scholarly interest on border regions; and explaining it by the the changing status of the borders from being barriers into becoming bridges between two countries. However, tensions occurring by such change are understood to be caused by the global forces imposed by supranatural conditions and EU for the purpose of devel-

oping flexible socio-economical and political links between the bordering communities; and by the constraints of the existing national institutions. The local/regional dynamics are inevitably affected by this tension between the two different scales of pressure.. While the article gives an in-depth theoretical background on the theme, it also provides a case study from Bulgarian-Turkish Border Communities. It concludes that such change from competition to cooperation status is regularized by the national institutions, but put into practice at the local level by the creative inventions of communities to bypass the restrictions of these institutions. Only then change could be achieved. Aşcıgil Dincer and Yılmaz in their article “Modelling road traffic noise annoyance by listening tests”, aim to measure the response of the dwellers and their level of annoyance when confronting the increasing traffic noise. Their methodology involves simulation technique: they gave sound clips to their subjects by telling them to assume that they are resting at home and hear the noise put into sound clips. Some of their salient findings include the following: settlement types and geometries can cause critical changes in the annoyance level of the subjects; dwellers with bedrooms across the street are more disturbed by the traffic noise than others; gender difference is experienced and women seem to have more sleep disturbance than men when subjected to traffic noise; and horns and motorcycles are found the be the sources for the most annoyance while resting. The findings carry clues for designers in taking into consideration the traffic annoyance for better comfort of the dwellers.


Dossier Editorial: On space syntax and architectural design A story of the suitcase Alper Ünlü A person carries some suitcases in the short life. I may define the suitcase here as the person’s interest, involvements, bringings or accumulations. I suppose the theory of Space Syntax has been one of my suitcases that I have been carrying more than seventeen years. It is interesting that the subject of Space Syntax has always attracted me as a magic suitcase. I noticed this subject, it was many years ago, it was an article, probably the first one of Bill Hillier’s, titled as same as the theory, “Space Syntax”. In fact, I was really in trouble about understanding this article at that time, however it was telling me about another but unknown face of the architecture. It was a kind of unusual face of my profession. I noticed that Hillier and Hanson were mentioning about something as a latent discourse like an iceberg under the sea. I thought, maybe it was reflecting something as a magic side of our professional duties. The subject that I confronted was reflecting another face, which I did not know before. Architecture in general, design as specific was my daily routine, either in teaching or in my professional projects. What was Space Syntax and what was it telling us? What was it explaining as a structure? How could it be linked to our life and spaces? I started working on Space Syntax in 1998, and I seriously began reading about it. In that time, some researchers were doing their researches by UCL based softwares, the others like me, “poors” were making calculations by hand. The publication of the bible book titled as “Social Logic of Space” was the only one comprising many discussions; it was proposing the linkage between societal and syntactic values from macro, meso and micro scales. I participated to Space Syntax Sym-

posium 2, Brasilia in 1999 with some syntactic formulas modified from this book. During the conference, participants were taking the depth values as inclining numericals, but I was doing reverse. I noticed my values were reverse, and they were opposite. However, my reverse model also was clearly working, like being depth values as reverse of the integration values as a constant acceptance. So, the paper that I presented in Brasila was about searching for the existence of the central hall concept of Turkish houses between 16th and 19th centuries. My accidentally emerging reverse model was working out and it was explaining the truth between these centuries. The true story was not descriptive, but it was using mathematical aggregates, more cells, depth levels, formulas and numericals. By time, the journey for Space Syntax motivated many students around me, and there was an increasing motivation among the students. The architects liked to work with space syntax softwares even for testing the design schemes and proposals. I noticed their motivation in these investigations. They liked to insert design layouts in their personal laboratories and they were checking, if the schemes were working or not due to integration and depth levels as reverse. They were searching for the result that the design is working out or not. Space Syntax rather than as methodology or tool, it was helping the researcher or the architect. The classical question was very simple and it was concentrating on the qualities of the design, was it working out or not? Was it exposing a low performance and where? Architects were testing the design schemes with softwares, and they liked to involve in figures. They were reinventing the figures but this time through the architectural field, and they were travelling in the latent and magic discourse. The architects were inventing the latent discourse based on mathematics that they lost this straightforward strategy in their past, maybe in the former school years. By the help of new feeling, students are settled for a new confirmation system


for their design strategies not only based on descriptive models, but also based on ratios. On the other hand, the series of symposiums as once in two years have been continuing, after Brasilia, London, Delft, Istanbul and Stockholm. My former PhD students Erincik, Mehmet and me were talking about something was missing in conferences. The first thing was apparently the difference between as high attendance for urban planning issues and small demand for architectural design issues. The ratio of architectural design issues considered in continuing series of conferences in general was less than 30 percent of whole accepted papers. This emerging case has reached to the upper limits in Santiago and Seoul. The ongoing symposiums were densely concentrating on urban issues and the architectural subjects seemed to be neglected and participants were tended to be declining. By the time, the segregation or division during the ongoing symposiums, the reasons behind the subject have become more clear for me. The urban dynamics seemed to be as concrete dynamics, whereas the building dynamics are tended to be more subjective, more human behaviour based. Of course, we may not separate the environment as place vs. space, whereas there are many similarities in both sides. We may also identify both of them as more formal, but if we may consider that the nature and the essence of the dynamics in the positivist world, the content for both sides seemed to be more reversible. That’s why the contribution from the architectural side to the space syntax was so limited. The planning issues and urban design issues were more appropriate as part of the immanent essence of research dynamics, on the other hand the architectural design issues were more based on emotional issues, in other words, they were more as transcendental interpretations which it might be more powerful than positivist explanations. At the basic scale, the immanent structure was determining the research strategies of urban and architectural spaces, but the lack of transcendental situation in critiques was the import-

ant issue in the architectural design field. The issues held in space syntax symposiums were mainly criticized based on these augmented problems. The lack of transcendental strategy is not solely a critical discussion for architectural spaces, but this is also a debating issue for space syntax symposiums like 2001 Atlanta. The famous philosopher and urban geographer, David Harvey addressed this issue in SS3, where he pointed out the lack of this issue in his critical conference. He annotated the space syntax as turning to be a vain subject in urban geography. Many thinkers commented that the space syntax theory is a part of immanent thinking and it is isolated from phenomenological interpretations. Latterly, David Seamon in SS6, Istanbul, criticized the lack of phenomenological approach in space syntax. His paper “A Lived Hermetic of People and Place: Phenomenology and Space Syntax”. Seamon’s paper, rather than being as segregating, it was promising for constructing an integrative model between differing opinions. By the effect of these debating considerations, space syntax theory has been continuing to influence us. By time, we are driving into another debate that it might be resulted more effective research and strategies at the future. In architecture, specifically at the design field, this is my guess that all discussions might be derived from the essence of the professionalism and the ongoing debates eventually may lead to the design strategies and methodologies as feedback. All these debates underpin the new methodologies of the architectural design field. This special dossier has been prepared under these considerations. Our main aim is to extent the debating issues and to convey them to the architectural design field and the profession itself. So, the title of this dossier is selected as “Space Syntax and Architectural Design” that is prepared to shed a light for the occurrence of human activities specifically in the buildings or in the vicinity of the buildings, how they are acted, by whom and where. The selected subjects due to buildings in this dossier are many. In the dossier, there are recently completed researches


on libraries, exhibition halls, shopping centers, elderly institutions, hospitals, gated communities, educational areas and houses. The first article by Daniel Koch is an invited one and the title of his article is “Architectural Works as Material Discourse: On the spatial formulation of knowledge and ideals in four public library extensions”. Koch evaluates four buildings in his article and exemplifies his ideas through Malmö City Library, Växjö City Library, Stockholm City Library and Gothenburg City Library. He argues the idea of library in the contemporary life and continues his discussion with the essence of knowledge and literature emerging through architectural programme. He extends his focus on collections, activity and library visitors. Koch mutually tries to find out answers linked to the architectural form and aims to discuss through these examples. He tries to find out new architectural principles and ideals of “new library design”. The results of his scrutinized works on these buildings bring out that appropriation and inhabitance are empirically related to the spatial organization in modern and digital life. The term of appropriation in here, rather than being an activity or Barker’s famous “milieu” term, it is much more spatial and cognitive issue and a more behavioral issue. Koch points out the complicated sides of architectural programming and behaviour based protocols in the design process. The space what we concentrated is how beyond of its pragmatic nature and how it is more media transacted meanings and values. Koch argues his considerations around this theme. The second invited article belongs to Kerstin Sailer. Her article is titled as “The Dynamics and Diversity of Space Use in the British Library”. This article also concentrates on another case study of the library. Sailer elaborates the dynamics of library in her article, and concentrates on diverse user behaviors and temporal order in the British Library. Her aim, through the cases, is conveying deep analytical explorations about spaces and their usage, where she presents a detailed picture of usage in the building. Sailer elaborates usage patterns, points out varia-

tions in behaviors, adds up temporal order in her approach, and tests the affordance of the configuration. Sailer presents elaborated spatial potentials of the building. She shows how physical environment is coexisted and presents behavioral modes and tendencies. By this way, she criticizes the architectural programming and beyond this, syntactical assumptions and how they are modified by new behavioral modes and patterns. If we start from blind reviewed articles, continuing from micro to meso and eventually macro scale, the third article belongs to Erincik Edgü and her article is “Success in basic design studios: Can seat selection be an advantage?” This research interestingly brings out the series of observation and syntactic data of an educational space, specifically the Basic Design Studio of the architectural school, CIU (Cyprus International University). Edgü explores students’ preferences of seating and she criticizes success of the student, in terms of social interaction and movement pattern based on the arrangement of rows and columns that is far from being the ideal scheme. Edgü takes a case of educational scheme in the architectural school and she argues physical layout, syntactic visual field and potentials. By the effect of these arguments, she extends her discussion around the essence of social communication, and mechanisms that are emerging social pedal or social fugal identity in the educational space. The fourth article is about the main exhibition venue of 2013 Istanbul Biennial, which is a former warehouse just as the building next to it, Istanbul Modern, where both buildings are altered as exhibition buildings in the last decade. The authors are Mehmet Emin Şalgamcıoğlu, Fitnat Cimşit Koş, and Ervin Garip. Their article is titled as “Tracing a Biennial Layout: Experiencing an Exhibition Layout through the Syntactic Analysis of Antrepo No. 3 at the 2013 Istanbul Biennial”. The authors explore syntactically integrated or segregated locations of exhibition spaces, and how spatial layouts influence visitors’ explorations in gallery spaces. It means, which spaces are more or less visited. Eventually, authors scrutinize


the predominate path that is depending on the number of visitors during a specific period of time. The authors try to find out the importance of coexistence cognizant and syntactically integrative spaces in accordance with the flexibility of design spirit of biennials. By this aim, authors discuss the impact of the morphology of space on exhibition visitors. The fifth article brings out and discusses the double aspect of the architectural space. The authors are Nilüfer Kozikoğlu and Pelin Dursun Çebi. Their article is titled as “Thinking and Designing with the Idea of Network in Architecture”. This article focuses on the idea of networks in architectural design and discusses the use of “graph theory” based on tools in the design process. Authors take the architectural layout cognitively as a mapping scheme that is accorded with spatial elements and they emphasize that these relations are highly cognitive and brain based. The mapping has dynamic essence and it is contrary to preliminary assumptions in the design process, which is non-hierarchical. Kozikoğlu and Dursun takes two examples, one from an academic setting-a mapping from the university, the other is elicited from practice-a hospital building. The first one describes a workshop on systems thinking that is demonstrated with a game and the second one is an iterative hospital campus design scheme. The aim of authors is to draw attention of the readers to the importance of cognitive mapping and cognized nodes; and how these dynamics are segregated from the immanent graphs. As continuum of the series of specific buildings, the sixth article’s subject elaborates an elderly care institution. The authors are Esra Özsüt Akan and Alper Ünlü. Their article is titled as “Behavioral Responses of the Elderly Regarding Spatial Configuration: An Elderly Care Institution Case Study”. This article elaborates a case study, an elderly institution in Istanbul, and authors define a scrutinized observation for behavioral occasions in the elderly institution. The authors draw the attention of the users to how designated spaces are used by the elderly, and the how segregated syntactic spaces are

transformed as sociopedal or sociofugal spaces. The seventh article belongs to Ervin Garip, Mehmet Emin Şalgamcıoğlu and Fitnat Cimşit Koş. The article is titled as “The Influence of Architectural Configuration on the Pedestrian Network in Büyük Beşiktaş Market”. The authors in this article take a specific building as the center of pedestrian movements, and exemplify their considerations through the “Büyük Beşiktaş Market” in Istanbul. The building is a distributor and connection of the pedestrian movements. The article tests an urban building that is continuously treaded by the pedestrian almost in every second. The authors question how syntactic pattern plays a role in this specific planning. They also test outcomes of their search as availability and usage of courtyards and open spaces regarding as possible design criteria in the creation of specific and private areas in public buildings. The eighth article belongs to ” Fitnat Cimşit Koş, Ervin Garip and Mehmet Emin Şalgamcıoğlu as their second outlook to the exhibition spaces with a title of “A Syntactic Analysis of Social Interfaces in Istanbul Biennial Patterns in Case of Biennial Buildings In 2013”. This article is different than the first one and it comprises activities and behavioral patterns in the urban scale. The authors question both the biennial spaces and buildings, and how they are preferred or not in the urban scale based on space syntax parameters. Their evaluation covers the history of biennials and the authors search for the importance of performances and they evaluate how they are interacted with syntactic values and visitors’ activities. The ninth article belongs to Nevşet Gül Çanakçıoğlu. Çanakçıoğlu summarizes her main theme as the question of the title “Can Cognitive Maps of Children be analysed by Space Syntax?” Çanakçıoğlu implements cognitive mapping in two different housing areas and she gets the responses from children aged 11 living in two different socioeconomic status levels. The analysis orients to be focused on cognitive mapping and syntactic graph outcomes. She establishes an interesting methodology between mapping


outcomes and interpretation of them in the syntax language. She finds out important clues about gender and social status level of children and the cognition of their home environments that is the primary subject of the article. The tenth article comes from Suat Apak, based on the relationship between gated communities, space syntax and emergence of graffiti in urban areas. The title of this interesting research is “Disintegration of Urban Housing Areas: Districts and New Gated Housing Settlements”. Apak takes gated community examples from different districts of Istanbul and observes these areas based on physical appearances, human movements and existence of graffiti. The author argues the existence of architectural appearance, existence of graffiti and syntactic outcomes; how they are interacted and what brings out to the urban life, and also how they are differed and why they are differed. Results show that the interaction within the existing texture gets weaker at substantial levels around the walls of the island, controversially the existence of graffiti increases, and sociologically existing users living around the gated communities turn to be as the author said “societies around / bottom the walls”. The eleventh article is derived from Michel Foucault’s “heterotopia” and “heterochronic” terms. The authors are İlgi Toprak and Alper Ünlü, and their article is titled as “A Diachronic Approach on Heterochronic Urban Space”. Toprak and Ünlü act from Foucault’s term and investigate the urban and social morphology of a coastal town, Kuzguncuk of Istanbul. The scrutinized work aims to take out geographical layers, to conduct a diachronic approach, and to investigate the cases and situations derived from the Heidegger’s “dasein” term. Toprak and Ünlü try to combine urban geography and space syntax terms in their paper. They evaluate the reflections of socio-cultural background of the historical neighborhood and they concentrate on the morphological and the semantic changes of its heterochronic elements throughout the history. The following twelfth article belongs to authors Erincik Edgü, Meray Ta-

luğ and Nezire Özgece. Their article is named “Divided Shopping: A Syntactic Approach to Consumer Behaviour”. The article focuses on the comparison of consumer shopping behaviour in a historical city centre, Walled City of Nicosia, the capital of Cyprus Republic and TRNC. Despite the political isolation and segregation, Nicosia has gone through different physical and social development patterns in terms of land uses and functional changes as a divided city. The authors try to examine syntactic hints underlying the physical development concerning in three different periods of the city. Moreover, they search for he preferences of the pedestrians emphasizing functional and spatial patterns. The highlighted issue in their search is that the divided city has consumer behaviors and patterns as an outcome of the physical environment that it might be interpreted as part of the syntactical definitions. The last thirteenth article has come from the authors Ayşe Özbil, Demet Yeşiltepe and Görsev Argın titled as “Modeling Walkability: the effects of street design, street-network configuration and land-use on pedestrian movement”. This study involves comparative roles of urban design qualities of the street environment and street network layout on pedestrian movement. In their comparative study, preliminary findings imply that not withstanding the significance of certain aspects of the street environment, but they relate to local urban design qualities. They also defend the syntactic basis of this comparison, where the overall spatial configuration of street network may be a concrete role and may prove to be a significant variable for the description and modulation of pedestrian movement. Consequently, the presented articles reinforce that we may be still circling around the problem or hopefully we are close to the target/ core. This special dossier may be a small step, but it is an important one for space syntax family. I consider this step as a more specific and a more architectural one that mainly sheds more light to our problems. The aim of this dossier is to help new generations and to motivate them


by giving a reason for inventing new methodologies in the future. The updated considerations in architectural design as part of the dynamic relations of the profession have been going on as slightly aparted from the functionalist-immanent foundations. I may apparently see this problem. So, the architectural design field should gain new methodologies concerning more involvement on transcendental interpretations, and more creative solutions in the design methodologies. As I indicated in the first page, this dossier will be unique and precious in my suitcase. My suitcase will be slightly filled with this, but there is still a space for being full. Sometimes, I call my space syntax suitcase and I label it as “hope chest”. If I translate it directly into Turkish, it is called as “çeyiz sandığı” that means “bride’s chest”. In

other words, all suitcases in our lives partly seem to be as the “chest for hope”. I think this dossier will be the “suitcase for hope”, especially for new generations. When I come to the end of very after very long introduction, presumably as a very long preface, I honestly admit that I should thank to my colleague, first Dr. Mehmet Emin Şalgamcıoğlu who is always showing high motivation and sincerity for the preparation of this publication, and solving our problems in hundred ways and always standing side by side in my life. My other gratitude is to Feride Şener Yılmaz who is always courteous and patient in the preparation of dossier for publication even for the typesetting and for being as a cordially representative of the editorial board.

ITU A|Z • Vol 12 No 3 • November 2015 • 7-22

Architecture as Material Discourse: On the spatial formulation of knowledge and ideals in four library extensions Daniel KOCH daniel.koch@arch.kth.se • School of Architecture, KTH Royal Institute of Technology, Stockholm, Sweden

Received: September 2015

Final Acceptance: October 2015

Abstract In recent decades libraries have been challenged in many ways, perhaps most pointedly by the digital revolution. This is, however, not the first time – a series of booms in library architecture emerging rather when knowledge ideals are challenged than established allows us to discuss library architecture more clearly as investigations into what knowledge, learning and literature could be rather than as expressions of what knowledge, learning and literature is. These questions are complex and multifaceted and require both careful examination of architectural proposals and works and a step back to analyse the propositions they make through their formulations into architectural form. Utilizing four public library extensions in Sweden, of which three have been built and one has been rebooted, and competition and parallel commission proposals for their making, this article discusses how ideas of libraries, knowledge, and literature emerge through the mediation of programme, collections, activity, and visitors in interaction, related to other aspects of architectural form. Building on a series of empirical findings of correspondences between use patterns of libraries and spatial configuration, the article takes this discussion further into what this means for a discussion of architectural principles, ideals, and propositions.

Keywords Architecture, Architectural competitions, Library architecture, Space Syntax, Spatial configuration.


1. Introduction In an interview in the Architectural Magazine RUM in 2011, Annette Gigon discusses the practice of building as a way to understand the world (Singstedt, 2011). Set in relation to a series of booms in library architecture, when they have appeared and the discussions around their creation, this statement seems to make an important point even if it does not specifically relate to it. It allows us to discuss library architecture more clearly as investigations into what knowledge, learning and literature could be rather than as expressions of what knowledge, learning and literature is. This also makes it easier to understand, one might argue, how come libraries have often been built not in situations of stability in the views of knowledge (or the degree of literacy; c.f. Markus, 1993; Bennet 1995), but in situations where such is challenged or under radical transformation. The recent boom in the so-called ‘western world’, extending back to the 1990s and tapering off somewhat after 2010 with some notable exceptions, has taken place in a situation where not only knowledge is under transformation, but libraries and books as such have been under attack through the advent or in the wake of the ‘rise of the network society’ (Castells, 1996) and the growing influence of digital media (Gillespie, Boczkowski, & Foot, 2014; Niegaard, 2011; c.f. Bruijnzeels, 2008; van der Velden, 2010). This has in many parts of the world taken the form of a wide range of investments made to build large, central public libraries (c.f. Roth, 2011). While some of these projects have been about radical transformation of the very idea of ‘libraries’, most have largely operated within a paradigm of large institutional buildings and within the frameworks of alterations of a typology. Internationally we have examples such as the Bibliotheca Alexandrina (Snøhetta, 2002), Sendai Mediatheque (Ito, 2001), Amsterdam Public Library (Jo Coenen & Co, 2004), Seattle Public Library (OMA, 2004), Chilean National Library (A & F Architects, 2009), and the Rolex Learning Center (Saana, 2009), amongst many others. In Scandinavia it includes the Culture House and Li-

brary in Copenhagen (COBE, 2010), Urban Mediaspace in Århus (schmidt hammer lassen, 2014), Halmstad Library (schmidt hammer lassen, 2006), and the City Library in Turku (JKMM, 2007), to name a few. Partially, this is a response to a general perception of an attack on and commercialisation and privatisation of public space (Zukin, 1995; Kärrholm, 2014; van der Werf, 2010) where libraries have been seen as one of few remaining bastions of the unquestionably public. This makes the question of what these libraries do even the more interesting. The interest here is a set of principal questions studied through a series of specific projects of alteration. This essay will therefore build on the material of four library extensions, three of which have been realized and one that has been put aside for now. In three of these cases thorough empirical analysis of inhabitance patterns has been made, which has partially been presented earlier (Koch, 2004). This material will be used to discuss the relations between architecture and library use as a foundation for a following discussion on architectural principles. It will therefore be as thoroughly presented as reasonable within the bounds of the essay, before the focus turns to two libraries and two proposed extensions of each. The discussion is deeply informed by the work of Thomas A. Markus (1993) and Sophia Psarra (2009) and their analysis of public and cultural buildings over time as well as Julienne Hanson’s (1998) extensive work on analysis of buildings. It is also heavily indebted to John Peponis’ (2005) notion of proposals and propositions in architecture, and his and others’ work on the formulation of architectural meaning (Peponis, Conroy Dalton, Wineman, & Dalton, 2003) and subsequent work on configurational meaning (Peponis, Bafna, Dahabreh, & Dogan, 2015). In addition, considerable work on libraries through history by e.g. Battles (2003), Lerner (2009) and Dahlkild (2011) have been pivotal to allow a broadened perspective on what a library is, has been, and could be. 1.1. A (very brief) historical point To understand the situation library ITU A|Z • Vol 12 No 3 • November 2015 • D. Koch


architecture is wrestling with there are key historical aspects to take into account that do not always come to surface in the debate, partially belonging to a typological process that embeds (perceived) history and future into the understanding of any type at any point in time (Steadman, 2014; Koch, 2014; c.f. Rossi, 1982), and partially in a ‘silent-but-present’ history of libraries informing expectations, values, and choices in their making. The type discussed in this essay – the public library in concurrent, European-American society – is by and large a modern product (Markus, 1993; Dahlkild, 2011). However, this type explicitly as well as implicitly carries a lot of its precursors in its underlying conceptual definitions conditioning how it is or can be treated. The type as it is perceived today also casts a shadow back in history affecting our interpretation of the role of libraries historically as well as extends itself into the future. Amongst the historical roots worth reminding of, one seems to be found in the etymological roots of the term. Both the English library and the Swedish bibliotek here comes from the root ‘book’ (Latin ‘liber’ and Greek ‘biblion’, both translating to ‘book’). While etymology should be handled carefully in relation to the development of building types (c.f. Forty, 2000), it is worth to consider how the concept is specifically tied to books and the handling or collections of them rather than buildings or practices of reading or learning. This suggests that the challenges to books raised on occasion goes right to the very core, origins and purpose of the type as such. In practice, the origins seem to be intertwined with various forms of archives. There is an important origin in the storage of texts – be it legal documents as in Egypt, written versions of oral traditions as in Ancient Greece, or tenants of philosophy as in China (Lerner, 2009). As many of these documents existed in only one copy, the earliest libraries often served a double role as archive and library: preserving texts in archives was a central part of the emergence of libraries. Preservation brought with it the formation of a place where one could find informa-

tion, which led many of the earliest libraries to take on a mediating function as well as becoming sites of material production where texts were copied for further propagation. While the degree of publicness have varied, the artefact holding a text, its materiality, storage, arrangement and subsequent use thus is arguably integrated in the very foundation of the library as a type and concept. One can of course argue, as is done in some contemporary discussions about libraries, that the importance of the book is a practical result of the types of media available within which to store and reproduce the content (c.f. van der Velden, 2010; Bruijnzeels, 2008), but this rather reflects a contemporary and not entirely unchallenged view of what a book ‘is’ that does not easily translate back to how it has been considered throughout history. It is in this situation important to not project concurrent view of technologies back onto earlier periods and cultures and their treatment of material (c.f. Lievrouw, 2014). The contemporary public library, as emerging largely in the 19th century, has several roots, including traditions such as monastic libraries and personal collections in Europe as well as the Islamic dar-al’ilm (Lerner, 2009, p. 5566). Already from the beginning they held an educational and enlightening purpose, perhaps most clearly similar to the dar-al’ilm, where learned librarians were to mediate knowledge and literature to a wider populace. This educational purpose also formulates one of many clear links to the university libraries, where the university library of Göttingen is often referred to as a key behind their transformation from storages to hearts of knowledge, largely attributed to the main librarian Christian Gottlob Heyne in the 18th century (Lerner, 2009, p. 112-119). However, the contemporary public library as a type should also be set in relation to a growing production of printed material, an increasing level of literacy, and an increasing amount of time available for especially the bourgeoisie but also the general urban population. This led to concerns for providing quality literature to the masses, but also for control of what was read,

Architecture as Material Discourse: On the spatial formulation of knowledge and ideals in four library extensions


and that the read material was appropriate for the various audiences who were reading it (Battles, 2003; Markus, 1993; Bennet, 1995). The public library, thereby, also incorporated a quality control aspect into itself, and formed a part in a societal power struggle over taste, knowledge, literature and ‘spare time’. In this development they are, therefore, also distinctly responses to a changing urban environment as well as a changing idea of society and the function-types required to uphold it. In such a historical view, a transition from more concretely collections of books to centres of learning and knowledge for and of the public therefore can be said to have begun long before any digital era challenging the role of the printed material; the very notion of this being a linear progress can be questioned. This must reasonably be considered as part of the problematic that is being revisited repeatedly, once again recently. There is, simply, more in the function-type ‘library’ that is being challenged than the form of the media at any turn, and perhaps especially with the digital revolution. 2. The empirical material: Two plus two times two The empirical material behind the discussion of this essay is, as noted, four public libraries and their planned or built extensions. These are the Malmö City Library (John Smedberg 1946, Henning Larsen 1997; Figure 1), the Växjö City Library (Erik Uluots 1965, schmidt hammer lassen 2003; Figure 2), the Stockholm City Library (Gunnar Asplund 1928; Figure 3), and Gothenburg City Library (Rune Lund 1967, Erséus Architects 2011; Figure 4). All are studied through interviews, on-site visits, and studies of plans and literature, and the first three extensively through on-site observations. In the latter two cases, alternative proposals for extensions have been studied. For Malmö and Växjö the observation data concerns the library with extension, and for Stockholm the library without extension. The reason to focus on extensions is twofold: on the one hand, extensions form a common practice in the latest boom, and on the other hand, extensions, specifically, raise interest-

Figure 1. Malmö City Library with extension, view from the nearby park towards the main entrance. Original to the left in the image. Photograph: Press Image/ Malmö City Library/Niclas Blomgren.

Figure 2. Växjö City Library. Photograph of exterior with the original in front. Press Image/ Växjö City Library.

ing architectural questions in having to negotiate between earlier and concurrent ideas. This forces choices and interpretations of what the ‘previous’ was, what respect for and preservation of it is, and how far and in what ways architectural transformations can take place. Analysis of competition proposals have been used in similar ways before (e.g. Schmeideknecht, 2012; Rustad, 2010), and there is an extensive discourse showing how proposals and judgements are often highly informative of architectural views through quality judgements and motivations of both architects and jury (c.f. Rönn, Kazemian, & Andersson, 2010). Naturally, competition briefs and conditions heavily influence the proposals,

Figure 3. Stockholm City Library, photograph from the park next to it. Photograph: Pressbild/ Stockholms Stadsbibliotek/Olle Nordgren. ITU A|Z • Vol 12 No 3 • November 2015 • D. Koch


Figure 4. Gothenburg City Library, exterior from Götaplatsen. Photograph: Håkan Grissler.

but this will here be given less attention in favour of focusing on the results as material discourse. As the cases of this essay are Swedish, it is of interest to note that public libraries in Sweden are bound by a specific law, ‘bibliotekslagen’, according to which the main purpose of public libraries is to – in line with the development of a democratic society – contribute to the distribution of knowledge and the free formation of opinions (Swedish Government, 2013). In addition to this aim, the law also states that libraries are to promote the position of literature and interest in education (‘bildning’), enlightenment, research, and cultural activities in general. Earlier iterations of the law more clearly pointed to mediation of state and municipal information, which is more or less completely integrated in the identity of the public libraries today. During the ‘million programme’ in the 20th century especially, but also before and after, public libraries were considered an integral part of any development. This has left a clear presence and expectation on libraries embedded in Swedish society. 2.1. Patterns of appropriation and spatial configuration Of particular interest for this discussion is the dual of symbolic values and public appropriation of the libraries in question, always ever in connection to one another. What I here mean by ‘appropriation’ is how the libraries come

to be used and inhabited by visitors but also how the arrangement of literature and reading places as well as other functions are negotiated, which interacts with volumetric, organisational, and aesthetic qualities of the library in the forming of an identity as a meaningful whole (Markus, 1993) – or a ‘public identity’ if one chooses to paraphrase Zukin (1995). It is therefore of note here that first, there is a high and significant correspondence between several patterns of inhabitance and the spatial configuration of the libraries analysed through means as developed from the principles set out by Hillier and Hanson (1984). Central to this is the understanding of spatial configuration defined as relations taking into account other relations (Hillier, 1996), which has been developed to be analysed through models that allow to mathematise these aspects of material arrangements of boundaries and spaces as graphs. The libraries have been analysed in a range of different ways and with a range of different resolutions; more specifically both as convex and axial systems (Hillier & Hanson, 1984) and as isovist fields (VGA) (Turner & Penn, 1999; Turner, 2001). The results have been carefully compared to various observed patterns of user behaviour (Figure 5), both through visual comparison and interpretation and statistical correlation studies. Most distinctly clear for the patterns of collective rates of movement, this is further traceable in other activities such as the differentiated practices of reading (c.f. Verschaffel, 2010) (Table 1). Of importance for the coming discussion on how principles and propositions are formulated in architectural

Figure 5. Snapshot observations, City Library of Malmö. Examples of observations. To the left a balcony in the centre of floor two, to the right the top floor of the original. Both interaction between people and percentage of seats filled is noticeably higher in the left. Each observed person forms a small circle grey, each larger grey circle symbolises interaction.

Architecture as Material Discourse: On the spatial formulation of knowledge and ideals in four library extensions


Table 1. Relations between spatial configuration and occupancy patterns. Only occupancy where relations have been found, have been included in the table. It should be noted that due to limits of programmatic distribution and furnishing, some of the found correspondences are additionally conditioned by e.g. availability of (type of) seats. Since furthermore distinctions between e.g. types of reading are highly interpretative, these have not been mathematically correlated as this risk giving a false impression of precision. Type of inhabitance Movement Flows

Type of observation

Analytic method


Counted as passing by pre-defined observation gates

Statistical Correlation (gate counts vs. integration)

Strong correlation to global integration with exterior system included


Occupancy of seats

Average correlation to global integration


Reading with notebooks and/or pen and paper

Statistical correlation, occupancy per seat (snapshot data vs. integration) Quantitative count, number of sitting per seat; qualitative interpretation of type of reading; individual trails

Focused studies

Reading focused for longer time period, often with headphones or similar Observed relaxed reading

Quantitative count, number of sitting per seat; qualitative interpretation of type of reading; individual trails

Tends to correspond to low integration, and limited views or low integration and few seats (one to two per table)

Qualitative estimation; qualitative interpretation of type of reading;

Tends to correspond to closeness to social interaction or spatial segregation

Group work

Observed interaction around books or other media

Quantitative count, share of available tables


Observed waiting until other person arrives

Qualitative estimation of share of waiters; qualitative judgement of behaviour

Corresponds to close-to integration with limited number of seats; somewhat related to literature Corresponds to a combination of integration and control

Social Interaction

Observed relaxed interaction

Qualitative estimation

Corresponds to distance from studying, relates to high integration

Arrangement of literature

Observed location of books and other media

Analysis of spatial distributions of literature compared to bookcases, intervisibility and spatial articulations

Follows primarily clearly articulated spaces or volumes; with less spatial articulation, bookcases gain increased importance for category differentiation

Reading novels

design work is that increased detail in the spatial analysis does not by default increase degree of correlation; rather, correlations tend to increase as details are reduced. Depending on analytic model the correlations do not peak at the same level of detail in all libraries, though for an axial analysis the pattern is consistent (Table 2). For an isovist field analysis, the visibility-blocking objects tend to have a higher impact especially for open hall sections of the libraries (Koch, 2005). While further research is needed to confirm its generality, a reasonable interpretation of the result seems to be that the correlation peaks at a certain level of complexity of the system, which is not too detailed

Corresponds to being close-to integration and multiple seats; largely unrelated to literature

and complicated, yet not too simple so as to not support orientation. Another reasonable interpretation, arguably intertwined with the earlier, is the degree to which the architecture, considered as the building in its own right, articulates spaces on a specific and detailed enough level or not. That is, in systems like Malmö’s and Växjö’s large halls, additional material structuring through the higher bookcases have a noticeable effect on global movement patterns, whereas in a distinct and clear articulation of a system as in the Stockholm case, global movement seems primarily related to the architectural definition of walls and spaces. It is worth to note here, that moveITU A|Z • Vol 12 No 3 • November 2015 • D. Koch


Table 2. Correlations (r-square values) between global axial line integration and movement flow rates for the libraries. For the axial line correlations specifically, the building articulation provides highest correlation in all cases, whereas VGA analysis gives the highest correlation at an intermediate resolution including vision-blocking objects for Växjö and Malmö (albeit a lower correlation than the axial analysis). Correlations are provided with the entry gate removed from the statistics due to how it alone raises correlations dramatically. Library Malmö Stockholm Växjö

Building Articulation 65,81% 82,37% 55,76%

Visibility limitations 60,76% 68,22% 52,37%

ment specifically correlates with the spatial configuration of the building structure in all cases, also when it peaks for more detailed analysis. The reason to point this out is to stress the importance of the architectural formulation and articulation of spaces and configurations as having significant impact on inhabitance regardless of subsequent furnishing and programming, recursively even more so through how it conditions the same. In a design process, of course, it does not work in such a single-directional manner and the process is always ever intertwined between program, building layout, interior furnishing, and a range of other concerns and questions (e.g. Koch & Miranda Carranza, 2014; Peponis, et al., 2015; Anderson, 1984) – but at some point the constructed building gain more inertia than other parts. It also points to the importance of overall architectural configuration for subsequent use, especially when we consider questions of programmatic distribution, identity and meaningful appropriation. Rather than more detailed levels of analysis, it is often more explanatory to abstract the analysis further to gain further insight into the patterns of appropriation, such as the volumetric distribution, clustering of spaces to subsystems and the overall configurational logic in relation to the specific configuration as analysed via isovist fields, convex spaces and axial systems. As an example, the grouping of literature into the three volumes of Malmö is a more meaningful way of understanding the distribution of the programme and collections than configurational or metric distances on a furnishing level of detail. This is not to disregard detailed analysis as it can clearly relate to other forms of use, such as the repeatedly found pattern of

Permeability limitations 49,7% 63,95% 48,27%

‘waiting’ in locations with a combination of high axial or isovist integration and high isovist control value. Second, it is important to point out that in the empirical studies of the libraries, the found correspondences between appropriation patterns and configurational properties require the analysis to take the exterior into consideration for most studied forms of use. Somewhat counter-intuitively, however, this is not as simple as that as public domains and spaces they form extensions of exterior public space and therefore the spatial system used for the analysis must include the public exterior. Rather, the correlations increase in all studied cases if the specific exterior is disregarded and the configurational formulation of this relation of the library building ‘itself ’ is analysed, although the effects on buildings with singular entrances like the analysed libraries are minimal. Technically, this is done through a process of ‘mirroring’ and thereby analysing the configuration as related to itself internally and through the entrances (c.f. Koch, 2013). This suggests that a large portion of the mediation of the library content to the public is defined by the internal configuration rather than the external, which increases the importance of understanding how the libraries ‘themselves’ make this description. While these correlations and observations are not the central discussion of this essay, I believe it is what enables the coming discussion as it means that changing the internal configuration of the libraries will re-describe the relation between program and public no matter what effort is put externally or in the specific new or existing entrance spaces to handle it. The internal configuration is simply a powerful means

Architecture as Material Discourse: On the spatial formulation of knowledge and ideals in four library extensions


through which this is communicated, and it is through the global internal configuration of the public portions that it is at its most powerful. The point here is to raise the discussion from pragmatics of ‘use-values’, pros and cons of different proposals, to a discussion of the proposition made in each specific proposal (Peponis, 2005) of what a library and what knowledge is, operating on an overall global-configurational level through its subsequent mediation of content and programme to visitors, and its relation to people both individual, as relational collectives, and as a general public, through how this comes together as possible meaningful wholes (Markus, 1993). How public is a public library? For whom? What kind of activity is given priority and under what conditions? What is arranged for to be common and shared, and what is arranged for to be a private, or at least personal, concern? And not the least, what is the general ideas of hierarchies and distributedness, and what degrees of differences are introduced or maintained? 2.2. Global configurations and structures of knowledge Architectural configuration participates in such propositions in more ways, however. For instance, the differentiation between ‘kinds’ of literature is within the bounds of possible configurations emphasised in the current City Library of Stockholm, making the visitor chose branch early on and making it comparatively more difficult to change branch the further into one one has venued. Comparatively, Malmö City Library offers a more flexible situation that does not to the same extent emphasise this differentiation and has a more networked character, but still makes clear differences between what is readily accessible and what is distant in the library in general. Växjö here, partially by means of placing all literature comparatively deep into the configuration, makes differences smaller. However, in spite of its circular form, it places an emphasis on two key transitions in the central atrium that operate very similar to the tree structure of Stockholm. In fact, in general Växjö operates in a very tree-like manner

simply due to how the solution emphasises the central walk around the atrium and de-emphasises the peripheral walk along its boundary – which is also observed to affect the movement pattern of visitors by making more common the inner route and less common the outer one even when the second would be the functionally more efficient. This tree-like structure is significantly strengthened as compared to the original library which was organised to perform much more like a grid or a field (c.f. Allen, 1997). The distributions of space hence clearly signals ideas of the structure and hierarchies of knowledge in general (e. g. as series, trees, or networks as discussed by Foucault, 1997) as well as the relations between depth and width thereof, in relation to which the distribution in space of functions, literature, and activities communicates both internal relations between these, and their relation to the public as well as what constitutes branches, entities, or categories therein. In practice the libraries spatially demonstrate a less clear and rational organisation than is perhaps expected, and one that is logically incoherent as it comes to their classifying operations – which appears to rather provide clarity than confusion for most of the visitors. Spatially measured as being co-located in the same spaces of intervisibility on the scale of architectural articulation of spaces, for instance, Malmö organises its literature into a clear separation between fact and fiction into two buildings, ‘youths’ as a separate group of in-between, arts and music as being as related to fiction as to humanities and social sciences, and a range of other categories. On a finer level, natural sciences and technology are separated, but co-located as compared to humanities and social sciences. Et cetera. (Figure 6) While one should not take these distributions too literally as representations of the ideal organisation of knowledge or literature for either architects or librarians, it can clearly be interpreted from the point of view of an emergent structure thereof based on a negotiation between an overall ideal and specific local choices which responds to values and ideals on an often non-discursive level (c.f. FouITU A|Z • Vol 12 No 3 • November 2015 • D. Koch


topias (Foucault, 1997) and therefore be left outside of consideration.

Figure 6. Malmö City Library, literature structure. Some simplifications have been made for readability. (a) Popular, (b) Newspapers and Magazines, (c) Children, (d) Humanities & History, (e) Social sciences, (f) Jurisprudence, (g) Technology, (h) Sustainability, (i) Natural Sciences, (j) Arts and music, (k) Reference books, (l) Fiction, (m) Thrillers, Detective Novels, etc, (n) Novels in foreign languages, (o) Old novels, (p) Literature by people from Malmö or about the region, (q) Genealogy and microfiche, (r) DVD and CD records, (s) Listening Books and easy reading, (t) Youth books, fact, (u) Youth books, fiction. As can be seen, especially fiction becomes a blended category. Another possible interpretation would have been that the southern length of rooms from the atrium (M, P, Q, S) constitutes ‘anomalies’ on each floor and that fiction rather consists of L, N and O.

cault, 2003; Markus & Cameron, 2002). In such a process the materiality of architecture and collections forces such choices to be made and to take shape within the spatial organisation through which it operates. It can therefore also be seen as to ‘unearth’ otherwise non-discursive portions thereof, and subsets of values and priorities that in ideal images can be relegated to hetero-

2.3. The extensions When discussing the extensions, it is important to consider that the libraries in question are historically and culturally significant; Stockholm has international recognition, Växjö was Sweden’s first open-hall library, and Gothenburg forms an important piece of Götaplatsen, a square surrounded by several important cultural and social buildings in Gothenburg. The outlier here is Malmö, where the old library was originally not built as a library and arguably had issues in how the library had been integrated into the existing old building before the extension was to be made. Furthermore, two of the built extensions are highly regarded: Malmö was awarded the Kasper Salin prize – the most prestigious architectural reward in Sweden (Hultin, 2001) – and Växjö was a nominee for the same. If we begin with the extensions less central to the essay – the ones of Malmö and Växjö City Library – it can be noted how they in both cases come in the form of clear geometric formal play (Figure 7). In Växjö an original square volume is paraphrased by a cylindrical volume, where the vertical arrangement of solid mass and transparency is inverted. In Malmö the square, nearly cubical, form of the original building is repeated in the extension with a similar size but including a small offset of a cube in a cube, connected to the old via a much smaller cylindrical form. There is, however, a noticeable difference in how the solution in Malmö places the new entry in the in-between cylinder, engaging both larger volumes equally and equidistantly from the new entry with several connections in-between and how the Växjö solution places the added cylinder behind the original square, leading to a differentiated relation to the public. In Malmö, furthermore, the old library still holds a large portion of the literature, whereas Växjö – arguably in a common form of solution – places all literature in the new volume and reworks the original book hall into casual reading places, information and utility desks, magazine and

Architecture as Material Discourse: On the spatial formulation of knowledge and ideals in four library extensions


newspaper sections, a café, and a few other service functions. On one hand, this preserves the main hall as present in any library visit to Växjö, while the old building can remain unvisited in Malmö. On the other hand, this makes a radical transformation of what a visit to the ‘old library’ contains, what practices it includes, and the aesthetics and functionality of the spaces in Växjö whereas the part of being a ‘library’ that concerns the housing and mediation of access to books is preserved in Malmö. At the same time, in order for the old library in Malmö to become an integrated part of the new whole in use and not only in volumetric composition, it has had to be dramatically altered; an earlier courtyard has been turned into an atrium in which balconies and stairs form the main communicatory space of the volume (Figure 8), and one of the most active parts of the library has become its connection to the other volumes, i.e. where there used to be bricks and mortar. Neither as a configuration nor as a generic material-spatial arrangement guiding and allowing inhabitance, then, the old library operated remotely close to how it does currently. Arguably, this comes as a result of an altered idea of what a library is combined with a wish to include the existing library into such an idea, where the consequences of such an altered view have been followed through architecturally. This solution in Malmö furthermore allows literature to remain closer to the public, whereas they are – in Hanson’s (1998) terms – insulated from the public in Växjö both distinctly spatially (number of spaces one needs to pass, especially in relation to total depth) and spatio-functionally (number of functions one needs to pass). It thus shows differences in relation to both architectural preservation and ideas of a library, but also in what careful consideration of the existing as argued for in both cases leads to as architectural proposals. From this point of view I will continue through the material of two times two proposals: two proposals regarding the Gothenburg City Library, and two proposals for the Stockholm City Library, in both cases comparing the winning proposal (which are radi-

Figure 7. The volumetric composition of Malmö (left) and Växjö (right).

Figure 8. Malmö City Library. Interior of the atrium, previously a courtyard, original building. Photograph: Malmö City Library/ Niclas Blomgren.

cally different between the two to start) to an alternate proposal that shows clearly other configurational strategies (c.f. Peponis, et al., 2015). 2.4. Stockholm and Gothenburg The Stockholm competition was an open international competition, receiving a vast number of contributions. Of these, I will mainly present two – Heike Hanada’s winning proposal ‘Delphinium’ (Figure 9) and JaJa Architects’ (Jakob Steen Christensen and Jan Yoshiyuki Tanaki) awarded proposal The Book Hill (shared 4th place). Hana-

Figure 9. Delphinium; extension proposal for Stockholm City Library by Heike Hanada. Press image from Sveriges Arkitekter for the competition. ITU A|Z • Vol 12 No 3 • November 2015 • D. Koch


da’s proposal clearly separates new and old in two volumes, respectfully keeping the Asplund building intact in its volumes and as a solitary, whereas Christensen and Tanaki’s makes a radical cut into the building connecting the new addition right into the rotunda of Asplund’s work (Figure 10). This makes them two of the most radically different outcomes of the competition. The Gothenburg extension was instead handled by a parallel commission given to five architectural offices. Also in this case I have chosen to make use of the winning proposal by Erséus Architects in comparison to one of the other proposals by Malmström & Edström. Here, the winning proposal (Figure 11) is claimed to work within the logic of the original work by “letting the library spaces grow forth and develop from within” (personal communication, Peter Erséus, May 2008, trans. by the author), whereas the other proposal maintains the exterior integrity of the building by adding a second volume to the side of the original seen from the square against which it is located (Figure 12). This means that the two are different both from one another and from the Stockholm proposals as in Malmström & Edström’s work, the new volume holds auxiliary functions whereas in Hanada’s work the new volume holds the main part of the book collection. Hanada’s work also aims to transform the library into more of a ‘public living room’ whereas Malmström & Edström’s aims to preserve the identity and character of the originally by adding space for the additional functions and activities expected of a library today. The relation

Figure 10. Book Hill; extension proposal for Stockholm City Library by JaJa Architects. Press image from Sveriges Arkitekter for the competition.

Figure 11. Extension proposal for Gothenburg City Library by Erséus Architects and WSP. Gothenburg/Press Image/Erséus Architects.

Figure 12. Extension proposal for Gothenburg City Library by Malmström & Edström. Gothenburg/Press Image/ Malmström & Edström.

to Götaplatsen here adds an additional dimension as the original library forms a coherent image with the surrounding buildings in material and stylistic expression, which is challenged by the extension by Erséus both in the volumetric composition and in aesthetics and materiality. It is thus possible to create a kind of matrix of solutions amongst proposals, all argued for as carefully preserving the core of the architecture of the existing libraries. On the one hand proposals creating a new building in order to preserve volumetric composition and exterior aesthetics of the existing, and proposals focusing on the ‘internal spatial logic’ at the expense of considerations for preservation of exterior expression. On the other hand, it is possible to set up the extent to which the traditional ‘library’ is significantly altered or moved. Here, Malmström & Edström’s proposal with a new volume is uncommon amongst extensions in that it specifically strives to maintain the existing as far as possible and let the ‘new’ functions and activities be housed in the addition. In light of the above discussion, however, it seems clear this matrix is insufficient as more

Architecture as Material Discourse: On the spatial formulation of knowledge and ideals in four library extensions


than a starting point. Considering the question of spatial organisation, taking into consideration the distinct relation of Asplund’s building to on the one hand an international library culture and on the other a prevalent idea of knowledge at the time of its making, it is arguably difficult to speak of preservation of the library without including how it represents and organises knowledge as a tree structure with distinct branches ordered around a central space of fiction. This is not a pragmatic solution to a programme, but a distinct formulation of an idea on several levels including spatial configuration as expression and representation as well as through how it organises narratives of library visits (c.f. Psarra, 2009) and distributes flows and other activities in and through the building. This organisational idea comes clear through how the tree structure has been maintained even when it might for practical reasons been better not to, and in how the deviations from the structure appear to be more or less hidden and subsequently less used (Koch, 2004). Arguably, the specific order of the collections – which has been altered over time – is of less importance than the systemic tree-representation and its relation to the central space of fiction. This representation also requires, more or less, all of the collection to follow this logic or it looses its power. The organisation furthermore tends to lead to a quieter space in the branches, as thoroughfare is more or less eliminated, supporting focused, undisturbed reading as a central activity while potentially inhibiting browsing. This is not to say the volumetric composition can be ignored; Asplund’s work as a whole clearly includes the volumetric arrangement and how these volumes meet the surrounding streets – carefully making the library offset and monumental from all directions – the streets, the park next to it, and the annexes (Figure 13). Keeping in mind the historical context of Asplund’s work, the library as a solitary composition is an important part of its urban identity; an identity arguably also dependant on housing at least most of the library within its boundaries. Since his work remains within an

Figure 13. Asplund’s Plan for the Stockholm City Library from 1928, showing the library without its fourth branch in the back, and with four annexes planned – as well as drafts of the park and hill. It is clear how the composition was a city planning question to a large extent. The square building at the top of the hill is a preliminary draft for a building for what is now Stockholm University. Public Domain by Age.

understanding of ‘library’ as primarily being concerned with the housing, organisation of, and access to a collection of books, de-coupling the building and the books under a new paradigm of library typologies becomes problematic from a point of view of preservation. At some breaking point, preserving the volume would still not preserve its library identity. This clearly has raised a kind of conundrum where the different proposals have opted to give priority to one aspect at the expense of other. This contradiction is noted already in the competition programme, stating that “[t]he assignment includes proposing a link that ensures the best possible spatial and functional connection between the buildings, all while being sufficiently respectful to the Asplund building” (Stockholm Stad, 2006, p. 32). However, when it comes to the judgement of the competition, it comes clear how the respect for Asplund’s volumetric composition and exterior overrules the programme’s stress on the connection (see Stockholm Stad, 2007). The jury also points to a not awarded but mentioned proposal by Wingård’s that arguably draws on the historical evolution of the Asplund building by rather than running a chirurgical corridor to the rotunda, replaces Asplund’s own addition of a western foyer with the beginning of the new extension, which, ITU A|Z • Vol 12 No 3 • November 2015 • D. Koch


Figure 14. Brass; extension proposal for Stockholm City Library by Wingård’s. Press image from Sveriges Arkitekter for the competition.

according to the jury, better connects to the rotunda and forms a more complete whole than the other proposals making the connection (Figure 14). From this point of view, if we insist on the library as a representation of an idea through how it organises and distributes its material collection of books, it is then also not enough to – as in some of the proposals – ensure that the rotunda is a central part of most visits to the library. The narrative logic of arrangement would have to be maintained as well. That is, the tree-structure branching out from the rotunda. In the material of the competition, it is not fully clear to what extent the proposals breaking their way through the wall into the rotunda do this, though in as far as it can be told it appears the focus has been on reaching the rotunda specifically to incorporate it in the new structure. As noted earlier, the winning proposal in Gothenburg, in comparison, explicitly attempts to work with the ‘formal logic’ of the existing over and above its volumes or its exterior expression; what is to be preserved is the original’s ‘inner logic of growth’ from the inside out. Under such logic it is reasonable to preserve the overall arrangement of functions and collections and let them grow outwards. While this logic only holds so far – the extension is after all rather modest in meters from the facade and there is little difference to this growth related to any need – it formulates a radically different stance. In the situation, the rhetoric around the proposal suggesting that the old library will still be visible through the glass parts of the new

facade must be considered more of a post-rationalisation (c.f. McMorrough, 2008); at the very least, it can be noted that as the work has been realized, the old library is hardly visible as reflections and viewing angles hide it behind the new glass facades. It is, however, part of the motivation of the jury. All the same, the approach have allowed Erséus to create a new whole based on an inner logic with consequences for other architectural questions such as the exterior expression and the contextual relation across Götaplatsen to the surrounding buildings. Malmström and Edtröm’s proposal here reminds more of the solution in Malmö, although even more consideration may be allowed for the existing situation. An added volume housing ‘other’ or ‘new’ activities attached to the old with some connections in between. Naturally this at least skews the configuration and centrality somewhat and rearranges certain flows, but arguably to a comparatively small extent. It does, however, clearly signal the split between the traditional and the new and questions can be raised how this affects the whole. In comparison to the Malmö case, one can raise the question whether there is a contradiction between global integration of the whole and preservation that can only be resolved by deliberate choice of priority, but this may seem premature. What is clear, however, is how different priorities taken to certain extents radically conflict with one another, possible to read as differences on the level of architectural principles and ideologies. What the configurative analysis allows is a deeper and firmer understanding of the structural representation and mediation of ideas of knowledge, its orders and arrangements, and the way it is to be accessed, sequenced, insulated, separated, and connected. It furthermore strengthens this understanding by showing how, specifically, the spatial configuration is powerful in communicating this by how it structures flows of movements and other activities, as expressed through correlations between configurational measures and various activities on the one hand, and additionally through how measures relate to common individual behav-

Architecture as Material Discourse: On the spatial formulation of knowledge and ideals in four library extensions


iours of visitors on the other. That this statistical correlation increases when the volumetric composition and more detailed configuration of delimitations and allowances of visibility coincide in their basic network logic is of interest but not the main point. 3. Transformations and Additions between Geometry, Aesthetics and Configuration What I have tried to do here, is not to suggest one strategy as superior over the other. I believe it is fair to say that I have my preferences, but that more than that I am deeply fascinated by the process and the proposals in how what could be considered as small shifts in priority can lead to such radically different results – and how this leads to all the proposals being argued for as being respectful to the existing building. To an extent, the proposals conduct a discussion where each of them challenge the very notion of preservation in all of the other proposals, suggesting they have misunderstood just what the foundational principles of the libraries are, which leads further to a discussion about architecture and architectural principles that reach deep into the heart of the concept. The argumentation presented for competition or parallel commission proposals must here, as McMorrough (2008) argues, be treated partially as ruminations on established case. While it must also be considered that there are many reasons that the proposals have reached their final shape, studying the proposals themselves as statements in such a debate is what allows for this discussion to happen. I do not here propose that it is as simple as differences between valuing aesthetics or use; this is clearly not the case and I would challenge such a division to start with. However, relations between geometry, configuration, aesthetics, and many other factors come into play, as well as how they relate to programmed content and to consecutive appropriation and inhabitance, as aspects of integrated architectural questions. These are statements made regardless of whether they are intended or not, and priorities made in the decision of a winner regardless of ex-

pressed intent or explicit motivation. In light of this discussion, one could argue that Asplund’s solution is resistant to extension by the dual relation of a tree-structure with a central rotunda inscribed in a distinct solidity of geometrical definition. This is only partially true, however, as the original library lacked one of the branches leaving one side open. On the other hand, the extension possibility was effectively cut off with Asplund’s own addition of the final volume in the ‘back’ that geometrically seals it off. The issue can of course be further complicated by raising questions of what the original architectural intentions were, both explicitly and implicitly, which additionally is challenged by the interplay of the discursive versus the non-discursive in architectural design suggesting that we cannot understand the original intentions solely by looking at what was formulated in speech or writing, which, in addition to not communicating non-discursive ideas, is also balanced by how the architects have had to argue for their proposals as well as how they have sought to present the work as conforming or challenging concurrent traditions – and if so, which. What we can say, however, is that the appropriation and inhabitance of the buildings, clearly shown empirically to relate to the spatial organisation in several distinct ways, reasonably must be one piece of the discussion that cannot reasonably be disregarded even if in the end it can be given higher or lower priority than other factors depending on aims and values. However, the more the inhabitance or the way a library mediates an idea of knowledge through how it structures narratives of uses and visits is valued, the more significant it becomes to consider the spatial-configurational effects of the extensions. In some cases, this may provide keys that allow clearer choice to be made on priorities and proposals, whereas in other cases it may help clarify if within the set ambitions a proposal taking care of all expectations is plausible. In order for a configurational analysis to contribute beyond mere pragmatics, however, a more complex understanding of mediation of ideals, values and meanings ITU A|Z • Vol 12 No 3 • November 2015 • D. Koch


not as separate from but intertwined with practical considerations and user concerns must be brought into play and analyzed from the point of view of architectural propositions. Acknowledgements The writing of this paper had not been possible without the discussions with John Peponis, Sophia Psarra and Ermal Shpuza back in Stockholm in the spring of 2013 and following. It also relies heavily on the discussions within the research group Spatial Analysis and Design at the KTH School of Architecture and, for this paper specifically, especially with Ann Legeby. Part of the writing work of this paper is done through the FORMAS-funded research environment Architecture in the Making. Finally I would like to acknowledge the support, openness and willingness to contribute in various ways shown by the libraries in question. References Allen, S. (1997). From object to field. Architecture After Geometry, Architectural Design, Profile 127, 67 (5/6), 2431. Anderson, S. (1984). Architectural design as a system of research programs. Design Studies, 5 (3), 146-150. Battles, M. (2003). Library: An unquiet history. London: W. W. Norton & Company. Bennett, T. (1995). The Birth of the Museum London. London: Routledge. Bruijnzeels, R. (2008). Design Criteria for the Library of Tomorrow. Volume, 15, 10-12. Castells, M. (1996). The Rise of the Network Society. Oxford: Blackwell Publishing. Dahlkild, N. (2011). The Emergence and Challenge of the Modern Library Building: Ideal types, model libraries, and guidelines, from the enlightenment to the experience economy. Library Trends, 60(1), 11-42. Forty, A. (2000), Words and Buildings: A vocabulary of modern architecture. New York: Thames & Hudson. Foucault, M. (1997). Of Other Spaces: Utopias and Heterotopias. In N. Leach (Ed.), Rethinking Architecture: a reader in cultural theory (pp. 350-356).

London: Routledge. Foucault, M. (2003). The Order of Things: an archaeology of the human sciences. London: Routledge. Gillespie, T., Boczkowski, P. J., Foot, K. A. (Eds.). (2014). Media Technologies: Essays on Communication, Materiality, and Society. Cambridge, MA: MIT Press. Hanson, J. (1998). Decoding Homes and Houses. Cambridge, UK: Cambridge University Press. Hillier, B., Hanson, J. (1984). The Social Logic of Space. Cambridge, UK: Cambridge University Press. Hultin, O. (2001). The Kasper Salin Prize 1998-2001. Stockholm: Arkitektur FÜrlag. Lerner, F. (2009). The Story of Libraries: From the invention of writing to the computer age. London: Continuum books. Lievrouw, L. A. (2014). Materiality and Media in Communication and Technology Studies: An unfinished Project. In T. Gillespie, P. J. Boczkowski, K. A. Foot (Eds.), Media Technologies: Essays on Communication, Materiality, and Society (21-51). Cambridge, MA: MIT Press. Kärrholm, M. (2012). Retailising space : architecture, retail and the territorialisation of public space. Burlington: Ashgate. Koch, D. (2004). Spatial Systems as Producers of Meaning: The idea of knowledge in three public libraries. Stockholm: KTH. Koch, D. (2005). Parallel Spatial Scales: discerning cognitive levels of space. In A. van Ness, (Ed.), Space Syntax 5th International Symposium: Proceedings Volume II (pp. 373-386). Delft: Techne press. Koch, D. (2013). The Architectural Interface Inside-Out: Interior-exterior relations, spatial models, and configurational mirroring. In Y. O. Kim, H. T. Park, K. W. Seo (Eds.), Proceedings of Ninth International Space Syntax Symposium (pp. 67:1-67:16). Seoul: Sejong University Press. Koch, D. (2014). Changing building typologies: The typological question and the formal basis of architecture. Journal of Space Syntax, 5(2), 168-189. Koch, D., Miranda Carranza, P. (2014). JOSS Spring/Summer 2014:

Architecture as Material Discourse: On the spatial formulation of knowledge and ideals in four library extensions


Models and diagrams in architectural design. Journal of Space Syntax, 5(1), i-xix. Markus, T. A. (1993). Buildings & Power: Freedom & Control in the Origin of Modern Building Types. London: Routledge. Markus, T. A., Cameron, D. (2002). The Words Between the Spaces: Buildings and language. London: Routledge. McMorrough, J. (2008) Ru(m)inations: The Haunts of Contemporary Architecture. Perspecta 40: Monsters, 164-169. Niegaard, H. (2011). Library Space and Digital Challenges. Library Trends, 60(1), 174-189. Psarra, S. (2009). Architecture and Narrative: The formation of space and cultural meaning. London: Routledge. Peponis, J. (2005). Formulation. The Journal of Architecture, 10(2), 119-133. Peponis, J., Conroy Dalton, R., Wineman, J., & Dalton, N. S. (2003). On the formulation of Spatial Meaning in Architectural Design. Peponis, J., Bafna, S., Dahabreh, S. M. (2015). Configurational Meaning and Conceptual Shifts in Design. The Journal of Architecture, 20(2), 215-243. Rossi, A. (1982). The Architecture of the City. Cambridge, MA: MIT Press. Roth, M. (2011). Library architecture + design. Salenstein: Braun. Rönn, M., Kazemian, R., Andersson, J. E. (Eds.). (2010). The Architectural Competition: Research inquiries and experiences. Stockholm: Axl Books. Rustad, R. (2010). What is Contemporary Architecture? Changes in architectural competitions and architectural discourse. In M. Rönn, R. Kazemian, & J. E. Andersson (Eds.), The Architectural Competition: Research inquiries and experiences (pp. 549-560). Stockholm: Axl Books. Schmeideknecht, T. (2012). Curating the Mainstream: the case of the german competition journal Wettbewerbe Aktuell. Nordic Journal of Architectural Research, 13-30. Singstedt, N. (2011). Att bygga är att

förstå världen. RUM, 8 2011, 71-74. Spieker, S. (2008). The Big Archive: art from bureaucracy. London: MIT Press. Steadman, P. (2014). Building Types and Built Forms. Leicestershire: Matador. Stockholm Stad (2006). /Asplund: Architectural Competition: The Stockholm City Library – Competition Brief. Stockholm: Stockholm Stad. Stockholm Stad (2007). /Asplund: Arkitekttävling om Stockholms Stadsbibliotek: Juryns rapport: slutligt resultat av allmän arkitekttävling i två steg. Stockholm: Stockholm Stad. Swedish Government. (2013). SFS 2013:801: Bibliotekslag. Stockholm: Kulturdepartementet. Turner, A. (2001). Depthmap: A program to perform visibility graph analysis. J. Peponis, J. Wineman, & S. Bafna (Eds.), Space Syntax 3rd International Symposium: Proceedings, (pp. 31.1-31.9). Michigan: A. Alfred Taubman College of Architecture and Urban Planning. Turner, A., Penn, A. (1999). Making isovists syntactic: isovist integration analysis. Paper presented at the 2nd International Symposium on Space Syntax, Universidad de Brasilia, Brazil, April 1999. van der Velden, D. (2010). The Netbook and its Library. In H. H. van der Werf (Ed.), The Architecture of Knowledge: The library of the future (pp. 2437). Rotterdam: NAi Publishers. van der Werf, H. H. (2010). The Architecture of Knowledge – Introduction. In H. H. van der Werf (Ed.), The Architecture of Knowledge: The library of the future (pp. 10-19). Rotterdam: NAi Publishers. Verschaffel, B. (2010). Guessing the Future of the Library. In H. H. van der Werf (Ed.), The Architecture of Knowledge: The library of the future (pp. 8495). Rotterdam: NAi Publishers. Zukin, S. (1995). The Cultures of Cities. Oxford: Blackwell.

ITU A|Z • Vol 12 No 3 • November 2015 • D. Koch

ITU A|Z • Vol 12 No 3 • November 2015 • 23-39

The dynamics and diversity of space use in the British Library

Kerstin SAILER k.sailer@ucl.ac.uk • Space Syntax Laboratory, The Bartlett School of Architecture, University College London, 140 Hampstead Road, London NW1 2BX, United Kingdom

Received: September 2015

Final Acceptance: October 2015

Abstract The Space Syntax study of buildings typically distinguishes between weak and strong programming, where social behaviours either follow or defy the spatial logic of a building. This is often based on analysing collective and aggregate patterns of behaviour. This paper builds on recent work redefining our understanding of weak and strong programming, yet aims to analyse usage patterns and spatial affordances in a much more fine-grained way by taking diversity of user groups as well as the temporal unfolding of behaviours into account. The British Library acts as a case study and is investigated based on a rich empirical dataset of observed user behaviours. Results suggest that the British Library shows both strong and weak programming: movement flows only partially followed spatial configuration, and the interface the building constructed kept people apart rather than bringing them together. In addition, large variations in user activities existed in some parts of the Library, all of which points towards strong programming. At the same time however, certain activities showed clear spatial preferences and significant differences in local and global visibility patterns, which illustrates weak programming. It was also shown how dynamic and diverse user behaviours emerged in the British Library, highlighting the need to draw a nuanced picture of usage. The contribution of the paper thus lies in a detailed and deep analysis of usage patterns, unpacking variations in behaviours between different users at different times and linking this both to the affordances of configuration as well as programmatic influences.

Keywords Public libraries, Space syntax, Space usage, Temporal dynamics, Strong and weak programming.


1. Introduction: More than just books… Buildings are dynamic settings that accommodate a range of different uses. A hospital for instance is not just a place for curing the ill, but also a workplace for nurses, doctors, cleaners and porters (see for instance Heo, Choudhary, Bafna, Hendrich, & Chow, 2009 on nurses assignments and behavioural patterns). A school not only enables teaching and learning, but is also an important place for socialisation, making friends and hanging out (Minuchin & Shapiro, 1983; Sailer, 2015). Shops have clear social functions beyond their retail purpose (Koch, 2014). Museums do not only exhibit pieces of art, but also educate, entertain and sell merchandise (Kwon & Sailer, 2015). Likewise, libraries have always been meeting places for communities in addition to storing books and organising access to information (Capille & Psarra, 2015). All of the above descriptions centre on what people do in buildings. Indeed, most buildings are defined by functions or so called ‘use types’ (Forty, 2000) or ‘activity types’ (Steadman, 2014): a hospital is a hospital because of what happens there, and likewise a school is a school again because of what goes on inside it (Hillier, Hanson, & Peponis, 1984). Yet, it seems that what people do in buildings becomes even more important, as a new focus on the human side of architecture as well as on people’s experiences, behaviours and usage patterns can be observed in recent discourses. The question of usage and daily life has already been popular in the 1970’s with architects like Herman Hertzberger defining architecture as concerned with ‘daily life lived by all people’ (Hertzberger, 1991), however only recently, scholars have argued that the social agenda of architecture has too long been a blind spot that needs re-addressing (Cupers, 2013). Other recent publications on usage and the social role of architecture and design (Awan, Schneider, & Till, 2011; Bergdoll, 2010; Maudlin & Vellinga, 2014; Till, 2009) underlined the important reading of buildings as ‘lived in’ (Brand, 1994; Hollis, 2009).

If we consider buildings based on usage, change becomes essential. Public libraries, like most other building types have seen a dramatic change in how they are used, perceived and experienced. The increasing digitisation of content means that new ways of accessing collections emerge, thus shifting the necessity of a physical site away from providing access and towards other uses. Drawing on a study of 24 recently built monumental public library buildings, Shoham and Yablonka (2008) came to the conclusion that the new-built libraries had increased user numbers, were full of life and served wider purposes as symbols of culture, as tourist attractions, but also as pleasant meeting places in a quiet cultured environment. The British Library forms a particularly interesting case in this context. The architect of the British Library, Sir Colin St John Wilson (1998) described the multitude of functions to be accommodated as: a day-to-day workplace, an institution that embodies and celebrates national memory, a storage of collections, places of study, exhibitions of its treasures, an eventspace hosting lectures and seminars, and back-of-house functions such as conservation laboratories and administration. This already points to a real diversity of space usage patterns. How the publicly accessible areas of the Library are indeed used in their everyday functioning will be explored in this paper, drawing on a rich data set of empirical and both quantitative and qualitative participant observations, collected in 2009 and 2010 by MSc students at the Bartlett, UCL. It will be asked how people move around in the building, to which degree the spatial layout (analysed with Space Syntax) informs usage patterns and how usage varies between different user groups, but also over time. Its main aim is to provide a sketch of the multi-functionality of the building and describe user groups and usage patterns in as much detail and variation as possible. This is an important task, if we want to reflect on how to design ‘social’ buildings in the future, where usage and people’s activities, preferences and experiences are actively anticipated, embedded, ITU A|Z • Vol 12 No 3 • November 2015 • K. Sailer


and allowed to grow and change. This paper is structured as follows: Chapter 2 will provide theoretical foundations for the empirical explorations of usage patterns in the British Library by sketching research on library buildings and human behaviours, but also by elaborating on the Space Syntax theory of ‘strong and weak programming’ in buildings. Chapter 3 will introduce the British Library as a case study, followed by a detailed description of methodology in Chapter 4. In four consecutive steps, Chapter 5 will discuss the diversity and dynamics of usage patterns in the British Library and a final Chapter 6 will draw conclusions, discuss limitations of the study and provide an outline of future research in the field. 2. Usage patterns and building types: On strong and weak programming In their paper ‘Visible Colleges’ Hillier and Penn (1991) conceptualised buildings as either strongly or weakly programmed depending on the degree to which the activity patterns inside the buildings followed strict rules, procedures and models. This theory is crucial in understanding the relationship between spatial layout and usage patterns inside different building types. A programme was defined as “not the organisation it houses (…) [but] the spatial dimension of an organisation, and the key element in any programme is the interface, or interfaces, that the building exists to construct (…) [i.e.] the spatial relation between or among two broad categories of persons (…) that every building defines: inhabitants, or those whose social identity as individuals is embedded in the spatial layout and who therefore have some degree of control of space; and visitors, who lack control, whose identities in the building are collective, usually temporary and subordinated to those of the inhabitants”. (Hillier & Penn, 1991, p. 33) Therefore, buildings were considered strongly programmed if the interface between user groups was highly controlled and the patterns of encounter followed so called ‘long models’ with a high degree of prescription and determinism (Hillier & Hanson, 1984). A court was the classic example

of a strong programme building, since different user groups with varying degrees of inhabitant or visitor status such as judges, barristers, witnesses, defendants and public were channelled through the building along separated paths so that their movement was highly controlled and encounters were actively hindered until all users met in the highly orchestrated and ritualised court room proceedings (Hanson, 1996). In contrast, buildings were seen as weakly programmed if the interface between user groups was not controlled and everyone could encounter everyone else freely, following ‘short models’ with a high degree of randomisation and morphogenesis (Hillier & Hanson, 1984). The most used example for traditionally weakly programmed buildings was the editorial floor of a newspaper, which flourished through generative and unstructured encounters among different users. The implications of this theory for the understanding of buildings and usage patterns lie mainly in the question how closely movement flows and resulting patterns of encounter correspond to spatial configuration. Traditional Space Syntax theory would suggest that movement flows are highest in areas of high spatial integration – so called ‘natural movement’ (Hillier & Iida, 2005; Hillier, Penn, Hanson, Grajewski, & Xu, 1993), however, adding strong and weak programming, we would only expect this relationship to hold in the case of weak programming, where randomisation is at play and the layout can act morphogenetically. In contrast, it could be argued that movement flows follow programme in strongly programmed buildings. Over recent years, the theory of strong and weak programming was taken up by different researchers and articulated further, for instance Koch and Steen (2012) proposed a new criterion for strong programming, thus adding more nuance and variation to the original concept. Likewise, Capille and Psarra (2013) suggested that the unequal distribution of activities across different spaces and functional areas of a building meant strong programming, whereas an equal distribution highlighted weak programming.

The dynamics and diversity of space use in the British Library


Studying two public libraries in London, it was concluded that one library was weakening the influence of programme on activities, while the other one was strengthening it. Additionally, it was shown that elements of strong programming could appear in traditionally considered weakly programmed building types such as workplaces and offices, for instance in the form of attractors that may deflect movement flows away from spatially integrated areas (Sailer, 2007, 2010). Similarly, buildings considered strongly programmed such as hospitals could show aspects of strong and weak programming, even within a single case depending on which criterion was applied (Sailer et al., 2013). What can be learnt from these studies, is the insight that space usage activities are dynamically enfolding systems, embedded in spatial situations and practices, driven by organisationally defined roles and programmes, but also distributed in space by configuration. Building types (such as hospitals, libraries or offices) cannot be associated with one type of programming per se; neither does a particular building necessarily follow clear categorisations as strong or weak programme. The theory of strong and weak programming of buildings can help scrutinise phenomena, yet detailed analysis is needed before a judgement on the degrees and levels of programming in its interplay with spatial configuration can be made. For the study of libraries, which in the traditional dichotomised description of either ‘strong’ or ‘weak’ would have been seen as a typical example of a weak programme (Zook & Bafna, 2012), this means scrutinising the space for aspects of strong programming (rules, procedures, attractors, strong distribution of activities by function, movement flowing against configurational logic) in addition to understanding weak programming and spatial practices. Particular studies on libraries within the framework of Space Syntax seem worth mentioning, too. With the aim to discover how spatial systems produce meaning, Koch (2004) studied three public libraries in Sweden and concluded that three dif-

ferent forms of knowledge representation (tree-like, network-like, as a control system) were found. Overall, it was concluded that contemporary libraries could be seen as systems that increasingly aim to integrate people and promote social encounter rather than keep people apart by providing silence, solitude and concentration. To that end, it was shown that the activity of reading occurred mainly next to heavily used corridors and areas of movement flows, thus giving rise to social encounters. This very phenomenon of reading in close proximity to highly integrated areas was found in a study of nine academic libraries in Portugal as well, however, here it was reported as a noise problem inhibiting concentration (Both, Heitor, & Medeiros, 2013). Another recent study on two academic libraries in London (Zong, 2015) focused on the diversity of activities as a result of new pedagogic ideas and digital access. Analysing both spatial configuration and furniture arrangements as affordances for usage, it was proposed that a diversity of spatial characteristics allowed for a diversity of usage patterns to unfold. The theme of libraries changing to accommodate different functions was also the subject of a syntactic study of 18 public libraries in France, where it was investigated how traditional libraries with closed collections and a central catalogue changed into so called ‘mediatheques’ providing access to a diversity of media sources and information types, which meant a spatial change towards open bookshelves and reading spaces (Lim & Kim, 2009). Results suggested that newer building types which followed the mediatheque model had lower overall values of visibility on average, but also a wider and more diverse range of configurational options. Visibility relations also feature in the paper by Zook and Bafna (2012), which highlights how everyday activities (borrowing a book, attending a scheduled meeting, meeting a friend in the reading rooms) in the Seattle Public Library – a building with a highly unusual spatial composition – still follow genotypical patterns of visual access, where paths lead through expected levels of openness and encloITU A|Z • Vol 12 No 3 • November 2015 • K. Sailer


sure, thus publicness and privacy, underlining the view of the library as a known institution. This so called ‘social staging’ is contrasted by a view of the ‘phenomenal staging’ – the subjective, individual experience of a user as they make their way through the building, which is characterised by unexpected vistas, hidden outlooks and surprising elements. Libraries were also used as settings to understand issues of wayfinding and signage (Carlson, Hölscher, Shipley, & Conroy Dalton, 2010; Li & Klippel, 2010, 2012) and to test new methods for user feedback and Post-Occupancy Evaluations (Dalton, Kuliga, & Hölscher, 2013). In summary, previous research has shown how libraries have become set-

tings staging a multitude of different usage patterns beyond the traditional access to collections and the accommodation of the process of reading. Instead libraries were shown to be social spaces, experienced differently by people and supported by specific configurational properties of the library buildings. How these phenomena resonate in the case of the British Library will be explored in the following chapters. 3. Case study: The British Library This paper draws on rich observations of space usage in one particularly interesting building: the British Library. As National Library of the United Kingdom, its aim is to store every book published in the UK and

Figure 1. Annotated floor plan of the British Library. The dynamics and diversity of space use in the British Library


make its collections freely accessible to the public. With a building size of more than 122,000 sqm it stores around 170 million items (among them almost 14 million books)1 and employs around 1200 staff in its main St Pancras building. As an institution, the British Library was founded in 1972 by an Act of Parliament, but first continued having its collection and reading rooms housed in the British Museum, until it moved to its own site in 1997, when the first reading room opened. The British Library’s main building in the centre of London near the railway station St Pancras was fully opened to the public in 1998 after a more than thirty-five year-long conception, design and construction process. The British architect Colin St John Wilson created a building ‘based on purpose’ and human scale in an approach that he called ‘the other tradition’ of Modernism (Wilson, 2007). For Wilson, architecture was grounded in use, creating an ordered framework for activities to happen; this was also called an ‘architecture of experience’ (Stonehouse, 2004). The building was described as ‘inviting’ and ‘democratic’ by critics: “A library might be expected to be conceived as a monument. (…) Yet this building [the British Library] has found a kind of democratic equivalent. (…) This kind of monumentality is not imposed upon us; it is assigned by us. So the building is symbolic, but this symbolism is not assertive and it is not about great occasions or collective events. The building seeks relationships with the individuals who use it and visit it, through a sense of invitation first evident in the forecourt and entrance. You are invited to be a participant, not merely a spectator.” (MacCormac, 2004, pp. xii-xiii)

The way in which the building negotiates between individual usage and institutional representation was described as ‘intimate monumentality’: “The building is book-like, revealing its inner world only when entered, an individual, intimate act. The hard, rather sober exterior allows the interior to be revealed and discovered on entering and using the library (…) – all part of an intimate experience of monumentality.” (Stonehouse, 2004, p. 69)

It was also praised to provide an “inherent versatility of form” (Stonehouse,

2004, p. 79), able to adapt to future usage and organisational needs, for instance the new ways in which information will increasingly be stored and accessed digitally. 4. Methodology This paper combines the syntactic study of the British Library, based on axial accessibility maps (drawn on knee level) and Visibility Graph Analysis (VGA, constructed on eye level) (Turner, Doxa, O’Sullivan, & Penn, 2001) with detailed and structured observations of space usage patterns. The following three standard Space Syntax observation techniques (Al-Sayed, Turner, Hillier, Iida, & Penn, 2015; Grajewski, 1992) were used: gatecounts, traces by following people and snapshots. For the gate-counts, movement flows across a total of 127 imaginary gates on all six public floors of the British Library were counted for five minutes each in the morning, midday and afternoon on three days (including Saturday) in 2009 and on two days (including Sunday) in 2010. Gender and readership status2 was recorded. Data was aggregated across all observations and collective hourly flow at each gate was calculated. Movement was also captured through traces, where observers picked up building users at entrances or other movement distributors on each floor plate (lift, staircases) and discretely followed them for 10 minutes (in 2009), 5 minutes (in 2010) or until they had reached a destination (for instance a desk in a Reading Room or a seat in the café), or in fact left the floor plate or building. The route they took was traced on a floor plan and digitised in GIS. A total of 679 building users were shadowed during Library opening hours, and additional demographic and user specific information (gender, estimated age range, formal or informal attire, Reader or Non-Reader) was noted. Snapshots recorded the exact location and type of activity of building users at a precise moment in time. All six publicly accessible floors of the building were observed repeatedly throughout the course of the day. Most areas ITU A|Z • Vol 12 No 3 • November 2015 • K. Sailer


Some of the books of the British Library are stored in its branch in Boston Spa in Yorkshire, from where they can be ordered to the Reading Rooms in St Pancras within 48 hours; basic statistics are from Wilson 1998 and updated statistics on the building are taken from: https:// en.wikipedia. org/wiki/British_ Library

2 Building users were distinguished by visual cues: Readers could be recognised by their Reader’s passes, sometimes worn around people’s necks, or more obviously, by carrying their belongings in plastic carrier bags, which were the only bags allowed inside the reading rooms. This means that everyone categorised as a Reader in the observations definitely had a Reader’s pass, but the observations could be biased by not recognising all Readers as such (for instance those leaving their possessions in lockers or at a desk in the Reading Rooms and going for a coffee).


were observed eight times in total with more intensively used areas captured up to 17 times to get a higher data resolution. Primary and secondary activities were recorded: primary activities included sitting, standing and walking as individual behaviours and interactions as group behaviours, while secondary activities distinguished a total of 22 different behaviours (such as searching, shopping, working on a laptop, reading, looking at an exhibition) and combinations of behaviours (such as laptop and reading, looking at an exhibition and talking, etc.). A total of 7993 people were observed. 5. The British Library: Diversity and dynamics of usage patterns Patterns of usage in the British Library will be analysed in this section, discussing the distribution of activities in space, the diversity of behaviours, rhythms and temporal patterns as well as emerging communities and their specific needs and activities. 5.1. Movement flows As a first step in the analysis, it is of interest to understand the overall distribution of people in the building and investigate to which degree movement flows are driven by configuration (indicating a weakly programmed building) or in contrast by programme and function (indicating a strongly programmed building). Four spatial variables of the axial map were analysed regarding their relationship to the overall flow of people in the British Library, as well as to the flow of Readers and Non-Readers (using gate-count data): Connectivity, Integration Radius 3 (Local), Integration Radius N (Global) and Choice. No sin-

gle relevant correlation was found, although Choice and Global Integration yielded highly significant / significant results for Non-Readers (p<0.0068** and p<0.0142* respectively) yet with very low R2 values of 0.03 each. Clearly, spatial configuration cannot explain the overall distribution of moving people very well in this case. Interestingly better correlations appear for a floor by floor analysis, particularly for the upper floors and for the non-reader demographic, as shown in Table 1. It seems that building users do follow the configurational logic of space to some degree when they are moving through the building, yet this is only the case for Non-Readers and only for the 1st, 2nd and 3rd floors, where correlation coefficients higher than 0.20 were achieved. Even then this is not a very strong relationship. Various factors interfere with the configurational logic. It could be argued that Readers know the building well and come for a particular purpose, i.e. to use the collections and take a seat in one of the Reading Rooms. Therefore, their movements are much more programmed and as such do not follow configuration. Secondly, the ground floor as well as the floors above (mezzanine) and below (lower ground) provide many different facilities and places of interest, most of them specifically targeted at Non-Readers, such as the shop, exhibition spaces, the cafĂŠ and canteen, the information desk, the cloak room, etc. This means attractors (Sailer, 2007) divert the flow of movement of Non-Readers and may counteract configuration as a way to distribute people. The role of the entrance should not be underestimated either. Every single building user pass-

Table 1. Coefficient R2 for correlation of movement flows (total, Reader, Non-Reader) with global Integration [INT] and Choice [CHOI]. Floor Whole Building Lower Ground Ground Mezzanine 1st Floor 2nd Floor 3rd Floor













































* Values marked in bold with * were significant at the 0.05 level and ** at the 0.01 level. Negative correlations are shown in green, low R2 (<0.2) are shown in grey and above that in black.

The dynamics and diversity of space use in the British Library



Figure 2 a-d. Movement traces of 427 Readers (red) and 149 Non-Readers (black) on the ground floor, mezzanine, first and second floor.

es through the entrance, however it is not necessarily also the most integrated place in the building. Buildings with many floors often show the integration core placed around the geometric centre of the building as a whole, which is also is the case for the British Library, where the most integrated areas are found on the first floor. Again, this distorts the correlation between movement flows and configuration. The relationship between different user groups such as Readers and Non-Readers can also be analysed as a matter of the interface constructed by the building to bring people together or keep them apart. First of all movement traces of the two different user groups can be compared visually. It can be seen in figure 2a-d that movement flows between Readers and Non-Readers overlap in certain parts of the building such

as the ground floor, mezzanine and first floor, yet, there are many spaces with a distinctive dominance of either Readers (upper floors, circulation, staircases, Reading Rooms) or Non-Readers (exhibitions, café, canteen). It can also be seen from the traces that Readers (shown in red in Figure 2a-d) move in a rather targeted fashion with straight routes, while Non-Readers (shown in black) tend to wander more aimlessly along curvy paths. The degree of co-presence between Readers and Non-Readers can also be investigated statistically by correlating total numbers for each group across the different locations in the building. With gate-count data a correlation of R2=0.28, p<0.0001 is obtained, showing that Readers and Non-Readers distributed rather differently across the building: areas with high counts ITU A|Z • Vol 12 No 3 • November 2015 • K. Sailer


of Readers showed rather low counts of Non-Readers and vice versa. This means the building creates a controlled interface between the different user groups and keeps them apart rather than bringing them systematically together. It seems the Library offered attractive spaces to each group separately; they co-existed rather than cohered and came together. In summary, the analysis of movement flows has highlighted that the British Library is a predominantly strongly programmed building: overall flows do not follow configurational logic consistently and different user groups with distinct usage patterns (Readers vs Non-Readers) were separated to a high degree. A further analysis of strong and weak programming in buildings as a function of the diversity and distribution of activities will follow in the next section. 5.2. Diversity and distribution of activities To investigate diversity and distribution of activities, a two-step approach was followed: firstly it was analysed whether primary and secondary activities differed according to their spatial properties of connectivity and integration (i.e. visual Mean Depth), retrieved from the VGA. This will highlight whether certain activities show preferences for areas with high or low direct visibility (connectivity) and for areas with strategically short or long visual paths (mean depth). Secondly, the distribution of activities will be brought together with the functional allocations of spaces to analyse whether specifically allocated areas attract usage differently from the overall building averages. Regarding the spatial logic of primary activities, i.e. sitting, standing and walking as individual behaviours

and interactions as group behaviours, highly significant differences of connectivity and mean depth can be found between these activities in a statistical Analysis of Variance (ANOVA). Results of the ANOVA tests are shown in Table 2; Figures 3a-b show the VGA of the building for connectivity and mean depth. Differences in connectivity are more pronounced, leading to a higher coefficient of R2=0.07 (p<0.0001), while mean depth only shows an R2=0.01 (p<0.0001), yet both are highly significant effects also due to the large sample size. The fact that connectivity seems to relate more strongly to user behaviour is in line with findings reported in previous research (Haq, 2003). In detail, interactions have the lowest average connectivity (284 VGA pixels, which equates to an area of 284 sqm, since the VGA grid was set to 1x1m); interactions are also relatively high in mean depth compared to other activities (5.83 on average), which means people interacted in rather segregated and smaller areas. Since the average values of connectivity and mean depth for the whole building are 585 and 5.95 respectively, all observed activities were more integrated than the building average (<5.95). Standing and walking took place in medium sized areas (354 and 418 sqm on average), but walking clearly happened in the most integrated places (lowest average mean depth of 5.54), while standing occurred in the most segregated ones of the places observed (MD=5.89). Sitting enjoyed the largest view-sheds with an average of 606 sqm, which is larger than the building average; this is clearly due to sitting being most prominent in the reading rooms, which are also relatively large in size. Secondary activities also showed significant differences between the direct and strategic visibility of various

Table 2. Number and statistics [mean, standard error] of spatial properties connectivity [CONN] and mean depth [MD] of observed primary activities from ANOVA tests. Activity Interaction Sitting Standing Walking

Count 591 6137 904 697

Mean [CONN] 284.836 606.156 354.165 418.305

Std Err [CONN] 16.535 5.131 13.370 15.226

Mean [MD] 5.83134 5.75574 5.88449 5.53567

Std Err [MD] 0.03167 0.00983 0.02560 0.02916

* Connectivity values larger than the building average and mean depth values lower than the building average are highlighted in red.

The dynamics and diversity of space use in the British Library


Figure 3 a-b. Visual Graph Analysis of the British Library: Connectivity (a) and Mean Depth (b).

behaviours; the ANOVA results for connectivity were highly significant (p<0.0001) with an R2=0.27, while R2=0.10 was obtained for mean depth (p<0.0001). Details are presented in Table 3. Interestingly for instance, laptop users preferred smaller and more integrated areas, especially if they were also talking (CONN=265, MD=5.47), whereas laptop users reading in parallel sat in larger and more segregated areas (CONN=929, MD=6.03). People looking around tended to be in very integrated areas (MD=4.37); similarly those occupants using their phones were also found in integrated spaces (MD=4.70). This means people seek out specific types of spaces in order to go about specific activities. Spatial configuration therefore played a role in distributing activities in space according to its degree of local and global integration and segregation. Some activities however, for instance shopping or engaging with

an exhibition could be argued to be driven by the functional programme rather than the preference for a particular spatial character of integration or segregation, hence the distribution of activities by function will be investigated next. In order to do so, the 22 different observed activities3 were clubbed together into ten broader core activities (as listed in figure 4a-b), for instance all activities involving talking (Eating Drinking Talking, Exhibition Talking, Laptop Talking, Reading Talking) were grouped together into ‘Talking’ rather than distinguishing by additional activities. This procedure also ensured that results were comparable to those reported by Capille and Psarra (2013), who distinguished nine different activities in their study of public libraries. The method presented by Capille and Psarra to quantify the degree of programming in a library by calculating the distribution of activities for the ITU A|Z • Vol 12 No 3 • November 2015 • K. Sailer


Only the data collected in 2009 was taken into account, since in 2010 sitting in some areas (specifically in the Reading Rooms) was not broken down further (e.g. reading, laptop usage, etc.) and this would distort the following analysis.


Table 3. Number and statistics [mean, standard error] of spatial properties connectivity [CONN] and mean depth [MD] of observed secondary activities from ANOVA tests. Activity Bag Cloak Locker Eating Drinking Eating Drinking Laptop Eating Drinking Reading Eating Drinking Talking Exhibition Exhibition Talking Laptop Laptop Reading Laptop Talking Lift Looking Order Phone Reading Reading Talking Searching Shopping Sitting Standing Talking Walking

Count 22 133 7 12 115 297 33 819 311 34 4 19 74 47 420 70 8 66 3534 208 880 625

Mean [CONN] 213.261 211.232 159.619 203.076 218.767 341.650 402.942 399.693 929.028 265.319 333.188 323.947 153.375 314.777 514.097 155.665 195.458 180.619 752.236 430.779 275.280 425.535

Std Err [CONN] Mean [MD] Std Err [MD] 77.42 6.70176 0.15607 31.49 6.03696 0.06348 137.26 6.51515 0.27669 104.83 5.54179 0.21132 33.86 5.88826 0.06826 21.07 6.05512 0.04248 63.22 6.17684 0.12743 12.69 5.40453 0.02558 20.59 6.03018 0.04151 62.28 5.47077 0.12554 181.57 5.93379 0.36602 83.31 4.37154 0.16794 42.21 6.14963 0.08510 52.97 4.69967 0.10678 17.72 5.75063 0.03572 43.40 6.37389 0.08750 128.39 6.21016 0.25882 44.70 7.03216 0.09011 6.11 5.76712 0.01231 25.18 5.75250 0.05076 12.24 5.74283 0.02468 14.53 5.57156 0.02928

* Connectivity values larger than the building average and mean depth values lower than the building average are highlighted in red.

building as a whole and comparing it to the distribution in different functional areas was applied here. In addition to investigating functional areas (Figure 4a), the variation across the different floors was also scrutinised (Figure 4b). Looking at the building as a whole, it can be seen that using a laptop is the most predominant activity in the Brit-

ish Library, amounting to 40% of all observed activities, followed by talking (20%), reading (15%) and sitting (8%). While the mix of activities in the corridors and cafĂŠ strongly resembled those found in the entire building (see figure 4a), some more variation was found in the foyer (where 19% of people engaged with exhibition material

Figure 4 a-b. Distribution of ten core activities (based on 2009 data) across different functional areas [a] and floors [b] of the building in comparison to the whole building average. The lower ground and 2nd floor were not observed in 2009. Percentage values are rounded. The dynamics and diversity of space use in the British Library


rather than just 3%) and the reading rooms (where laptop use rose to 63%); the highest level of variation however was found on the staircases and the exhibition spaces themselves. The figures obtained for the average variation in those two areas are in line with the ones reported by Capille and Psarra for a strongly programmed building. This means that some parts of the British Library enact a strong programme, distributing and shaping behaviours and the mix of activities in addition to the effect of spatial configuration, as argued in the previous section. Analysed floor by floor, the mix of activities on the mezzanine level, as well as the 1st and 3rd floors is comparable to the overall building distribution; only the ground floor showed significant variation, which is due to the high number of specialised functions on the ground floor, for instance exhibition spaces. The role of the ground floor in helping people to orient themselves also becomes obvious in the disproportionately high percentage of people using their phones (14% rather than 2%) and more than twice the percentage of people talking (43% rather than 20%). To summarise, the British Library is a building combining both elements of weak programming (since activities showed statistically significant preferences for integrated or segregated spaces) and strong programming (since activities were distributed unevenly across the functional areas of the building). 5.3. Rhythms and temporal patterns of usage In addition to the analysis of the overall diversity of activities unfolding in the British Library, changes in usage patterns over time were also investigated. Regarding the distribution of movement, it can be seen that busy areas (with high flow intensity) during the week are not necessarily those also populated to a higher degree on weekends. A correlation of gate counts based on traces for the week versus the weekend reveals an R2=0.32 (p<0.0001), which means that 32% of flow intensity on the weekend can be predicted by flow intensity during the week, but the ma-

jority of flow intensity across the different spaces differed between weekday and weekend. Generally speaking, the upper floors showed higher usage intensity during the week, whereas lower floors showed more intensive usage over the weekend. Another interesting difference between weekday and weekend usage patterns can be revealed by re-doing the analysis of variation of connectivity values of each activity (as done in section 5.2 and shown in Table 3 above), but now executed separately for weekday versus weekend. The same was repeated for mean depth. The ANOVA is highly significant for all datasets (p<0.0001); correlation coefficients for connectivity were higher for weekdays (R2=0.33) than weekends (R2=0.20), whereas it is the other way around for mean depth with lower coefficients for weekdays (R2=0.14) than weekends (R2=0.34). It should be noted that coefficients generally rise by splitting the data by day of the week, which shows that different patterns were evolving on weekdays versus weekends. The differences between average connectivity and mean depth values between weekday and weekend are plotted in Figure 5a-b. The most pronounced differences can be observed for activities involving the use of laptops and talking, but also for reading, walking and sitting. People working on their laptops preferred smaller (CONN=234) and more integrated spaces (MD=5.04) on the weekends as opposed to weekdays (CONN=451, MD=5.51); this is even more pronounced for those using their laptop alongside reading (CONN=236 vs CONN=944 and MD=5.31 vs MD=6.03). Reading itself as well as sitting showed a similar preference for smaller and more integrated spaces on the weekends than on weekdays, despite the fact that weekend observations were done on Saturdays, which meant the Reading Rooms (as rather large and segregated spaces) were open. Talking occurred in similarly sized areas on weekdays and weekends, however, on weekends talking happened in much more integrated spaces (MD=5.27 rather than 5.96). Possibly it could be the case that users sought ITU A|Z â&#x20AC;˘ Vol 12 No 3 â&#x20AC;˘ November 2015 â&#x20AC;˘ K. Sailer


Figure 5 a-b. Average connectivity [a] and average mean depth [b] of activities on weekdays versus weekends (based on 2009 data only, since weekend observations were not done in 2010).

more socialising opportunities on weekends and thus preferred to place themselves in more buzzy (hence integrated) areas. Looking at the distribution of ac-

Figure 6. Variation of ten core activities plus interactions between people (based on 2009 data) over the course of the day. The time marked in the diagram highlights the starting time of the observation period.

tivities across functional areas again, not much variation appeared between weekday and weekend, however interesting differences can be detected over the course of the day (as illustrated in Figure 6). Concentrated work such as reading or working on a laptop peaks in the afternoon (from 3-4pm); eating peaked at lunchtime and in the early afternoon as expected (between 1-3pm); social activities such as talking and interactions peaked mid-morning (11-12pm), at lunchtime (1-2pm) and in the late afternoon (5-6pm); and the engagement with exhibits showed a high in early morning (10-11am) after lunch (2-3pm) and in the afternoon (4-5pm). In summary, this analysis highlights how temporal patterns of usage evolve, creating a rhythm of activities and experiences over the course of a day with shifting preferences and locations of activities between weekday and weekend. Qualitative accounts of peopleâ&#x20AC;&#x2122;s engagement and behaviours in the

The dynamics and diversity of space use in the British Library


British Library (Thomas, 2013) seem to underline this interpretation of user experience of rhythms and temporal patterns, where people chose to change activities as the day went on. 5.4. Emerging communities and usage patterns Last but not least, specific usage patterns of emerging communities and particular groups of people can be investigated. From qualitative observations we know that a group of people using the British Library for working purposes queue outside the building every morning to take up seats in front of the Kings Library on the first floor, which offered good seating, nice individual lighting and power plugs in addition to the free wifi available in the whole building. Those spaces shown in Figure 7 offer both opportunities for socialising (as they are in a highly frequented route) as well as solitude (by the nature of the furniture) and were the most popular seats in the Library, essentially being occupied first thing in the morning and throughout the whole day. As an emerging community of so called ‘nomadic workers’, people have come to know each other and watched out for other people’s belongings. Other specific communities of people with particular space usage patterns were entrepreneurs using the IP and Business Centre of the British Library. All areas connected with the IP and Business Centre showed disproportionate numbers of males (four times as many males as females as opposed to a ratio of 1:1.15 for the building as a whole), but also higher numbers of users in the age range 40-60 (1.7 older people per younger people in contrast to a 1:1.03 ratio for the entire building). Other areas with an uneven distribution of users by additional demographic information include a female dominance in the Social Sciences Reading Room (3.4 females per male) and the shop (1.6 females per male) and a higher presence of older people (40-60 years of age) in the Philatelic Exhibition (3.6 older people per young person), whereas twice as many 20-40 year olds as compared to 40-60 year olds were found around the areas of the Folio Society Gallery.

Figure 7. Community of nomadic workers in front of the Kings Library. Photograph by Kerstin Sailer.

This account of emerging communities and differences in the distribution of people with certain user demographics highlights how the building affords behaviours by particular groups of people in distinct ways. 6. Conclusions: On diversity, dynamics and built form This paper presented evidence from observations of space usage patterns in the British Library in conjunction with an analysis of the spatial configuration of the building and its affordances for user behaviours. It was shown how movement flows in the British Library mostly defied configurational logic. In addition to a rather controlled interface between different categories of people such as Readers and Non-Readers, this drew a picture of a strongly programmed building. However, the analysis of the distribution of activities across space highlighted that activities and behaviours of people followed configurationally defined preferences and thus showed weak programming. Functional areas in contrast, in particular exhibition spaces and to a smaller degree the foyer and Reading Rooms maintained elements of strong programming, since the mix of activities there differed significantly from the overall building average, pinpointing the many ways in which the functional allocation and affordances of different spaces drove usage behaviours. The analysis of temporal dynamics, emerging communities and different user experiences over the course of the day ITU A|Z • Vol 12 No 3 • November 2015 • K. Sailer


and between weekday and weekend highlighted how the building constantly evolved, shifted and changed, depending on perspective. The main contribution of this paper is therefore the conceptualisation of a building as a layered, dynamic and changing experience rather than as a definite entity impacting collective user behaviour in one particular way. It also shifts the attention of Space Syntax analysis away from top-level collective user behaviours to more nuanced and detailed understandings of the diversity and dynamics of the relation between configuration and space usage. Due to the nature of the used data set, this paper has clear limitations; issues include inconsistent data (for instance different observation standards in 2009 and 2010), missing data (for instance not all areas were covered equally well), and possibly limited quality of the data due to issues with interobserver reliability, specifically given that the data was collected by Master’s students in their first weeks of their degree. Wherever possible those limitations were taken into account for the different types of analysis. To conclude, this paper has investigated the diversity of different space usages of a building over time in relation to its spatial configuration. It has explored both temporal dynamics as well as usage diversity to incorporate a more differentiated perspective on who uses a building when for what purpose, or in short the ‘multiplicities of occupation’ (Groák, 1992). Space Syntax can offer a fruitful framework for this exploration beyond mere aggregate and collective social patterns. Future research could focus on the nuances of temporal and user-specific dynamics more systematically to address what Brand (1994) called a ‘shocking lack of data’ on building usage. He highlighted the need for studies of all kinds of buildings in use and what changes from hour to hour, day to day, week to week, month to month and over the years. 20 years later this is still an open research question, which this paper hopes to contribute towards. Through its architecture and ‘versatility of form’, the British Library has clearly managed to be a space that

‘builds relationships with individuals’ as evident in the diverse and dynamic usage patterns showcased in this paper. Acknowledgements I would like to express my sincere gratitude to the staff at the British Library, in particular their Welcome Team and Mark Walton, who have made it possible to investigate the building and its usage. I would also like to thank two generations of students of the MSc ‘Advanced Architectural Studies’ (now called: ‘Spatial Design: Architecture and Cities’), who have collected the data used in this paper during long days in the field as part of their coursework; these are in alphabetical order: Eleni Alexiou, Min Hi Chun, Monica Datta, Jean-Francois Goyette, Birce Eren Karafazli, Eun Hye Kim, Tim Mason, Fiona McDonald, Gillian McNally, Nikolina Nikolova, Stella Parpa, Amanda Pluviano, Rosamund Pomeroy, Zhen (Alex) Qian, Carolina Rodriguez, Aabid Raheem, Khondoker Mobinur Rahman, Fernanda Lima Sakr, Ria George, Frederik Weissenborn and Jingcao Zhang (MSc AAS cohort of 2009-2010); and Rahwa Ayob, John Bingham Hall, Michelle Chan, Vasiliki Gogou, Rosie Haslem, Zahra Khaniki, Efstathia Kostopoulou, Annita Miltiadis, Krisangella Camacho, Rosica Pachilova, Fei Que, Zheng Xie and Kelin Yue (MSc AAS cohort of 2010-2011). Last but not least, I would like to thank Dr Jaehong Lee for his help with the spatial analysis as well as with cleaning and processing observation data. References Al-Sayed, K., Turner, A., Hillier, B., Iida, S., & Penn, A. (2015). Space Syntax Methodology. A teaching textbook for the MSc Spatial Design: Architecture and Cities Retrieved from http:// discovery.ucl.ac.uk/1415080/ Awan, N., Schneider, T., & Till, J. (2011). Spatial Agency. Other ways of doing architecture. Abingdon / New York: Routledge. Bergdoll, B. (2010). Introduction. In A. Lepik (Ed.), Small Scale, Big Change: New Architectures of Social Engagement (pp. 7-11). New York: Museum of Modern Art.

The dynamics and diversity of space use in the British Library


Both, K., Heitor, T., & Medeiros, V. (2013). Spaces for Knowledge: Strategies in academic library planning and design. Paper presented at the 9th International Space Syntax Symposium, Seoul. Brand, S. (1994). How buildings learn. What happens after they’re built. New York / London: Penguin Books. Capille, C., & Psarra, S. (2013). Space and planned informality: Strong and weak programme categorisation in public learning environments. Paper presented at the 9th International Space Syntax Symposium, Seoul. Capille, C., & Psarra, S. (2015). Disciplined informality: Assembling unprogrammed spatial practices in three public libraries in Medellin. Paper presented at the 10th International Space Syntax Symposium, London. Carlson, L. A., Hölscher, C., Shipley, T. F., & Conroy Dalton, R. (2010). Getting Lost in Buildings. Current Directions in Psychological Science, 19(5), 284-289. doi: 10.1177/0963721410383243 Forty, A. (2000). Words and Buildings. A Vocabulary of Modern Architecture. London: Thames and Hudson. Grajewski, T. (1992). Space Syntax Observation Manual (2001 unpublished revised edition: L. Vaughan) (2001 unpublished revised edition: L. Vaughan ed., pp. 18). London: UCL Bartlett and Space Syntax Ltd. Groák, S. (1992). The idea of building. Thought and action in the design and production of buildings. London: E & FN Spon. Hanson, J. (1996). The architecture of justice: iconography and space configuration in the English law court building. Architectural Research Quarterly, 1(4), 50-59. Haq, S. (2003). Investigating the syntax line: configurational properties and cognitive correlates. Environment and Planning B: Planning and Design, 30(6), 841-863. Heo, Y., Choudhary, R., Bafna, S., Hendrich, A., & Chow, M. P. (2009). A modeling approch for estimating the impact of spatial configuration on nurses’ movement. Paper presented at the 7th International Space Syntax Symposium, Stockholm, Sweden. Hertzberger, H. (1991). Lessons for Students in Architecture. Rotterdam: Uitgeverij 010 Publishers.

Hillier, B., & Hanson, J. (1984). The social logic of space. Cambridge: Cambridge University Press. Hillier, B., Hanson, J., & Peponis, J. (1984). What do we mean by building function? In J. A. Powell, I. Cooper & S. Lera (Eds.), Designing for building utilisation (pp. 61-72). London: Spon Ltd. Hillier, B., & Iida, S. (2005). Network effects and psychological effects: a theory of urban movement. Paper presented at the Proceedings of the 5th International Space Syntax Symposium, TU Delft. Hillier, B., & Penn, A. (1991). Visible Colleges: Structure and Randomness in the Place of Discovery. Science in Context, 4(1), 23-49. Hillier, B., Penn, A., Hanson, J., Grajewski, T., & Xu, J. (1993). Natural movement: or, configuration and attraction in urban pedestrian movement. Environment and Planning B: Planning and Design, 20, 29-66. Hollis, E. (2009). The Secret Lives of Buildings. From the Parthenon to the Vegas Strip in thirteen stories. London: Portobello Books. Koch, D. (2004). Spatial Systems as Producers of Meaning - the idea of knowledge in three public libraries. Licentiate Thesis, KTH Stockholm, Stockholm. Koch, D. (2014). Changing building typologies: The typological question and the formal basis of architecture. The Journal of Space Syntax, 5(2), 168189. Koch, D., & Steen, J. (2012). Analysis of strongly programmed workplace environments. Architectural configuration and time-space properties of hospital work. Paper presented at the 8th International Space Syntax Symposium, Santiago de Chile. Kwon, S. J., & Sailer, K. (2015). Seeing and being seen inside a museum and a department store. A comparison study in visibility and co-presence patterns. Paper presented at the 10th International Space Syntax Symposium, London. Li, R., & Klippel, A. (2010). Using space syntax to understand knowledge acquisition and wayfinding in indoor environments. Paper presented at the 9th IEEE International Conference on Cognitive Informatics (ICCI), Beijing. Li, R., & Klippel, A. (2012). Wayfinding in Libraries: Can Problems ITU A|Z • Vol 12 No 3 • November 2015 • K. Sailer


Be Predicted? Journal of Map & Geography Libraries, 8(1), 21-38. doi: 10.1080/15420353.2011.622456 Lim, H., & Kim, S. (2009). Changes in Spatial Organization in French Public Libraries. Journal of Asian Architecture and Building Engineering, 8(2), 323-330. doi: 10.3130/jaabe.8.323 MacCormac, R. (2004). Foreword. In R. Stonehouse & G. Stromberg (Eds.), The Architecture of the British Library at St Pancras (pp. xii-xiv). London: Spon Press. Maudlin, D., & Vellinga, M. (Eds.). (2014). Consuming Architecture. On the occupation, appropriation and interpretation of buildings. Abingdon / New York: Routledge. Minuchin, P., & Shapiro, E. K. (1983). The school as a context for social development. In E. M. Hetherington & P. H. Mussen (Eds.), Handbook of child psychology: Socialization, personality and social development (Vol. 4, pp. 197274). New York: Wiley. Sailer, K. (2007). Movement in workplace environments - configurational or programmed? Paper presented at the 6th International Space Syntax Symposium, Istanbul. Sailer, K. (2010). The space-organisation relationship. On the shape of the relationship between spatial configuration and collective organisational behaviours. Doctoral Dissertation PhD, Technical University of Dresden, Dresden. Retrieved from http://nbn-resolving.de/urn:nbn:de:bsz:14-qucosa-38427 Sailer, K. (2015). The Spatial and Social Organisation of Teaching and Learning – The Case of Hogwarts School of Witchcraft and Wizardry. Paper presented at the 10th International Space Syntax Symposium, London. Sailer, K., Pachilova, R., Kostopoulou, E., Pradinuk, R., MacKinnon, D., & Hoofwijk, T. (2013). How Strongly

Programmed is a Strong Programme Building? A Comparative Analysis of Outpatient Clinics in Two Hospitals. Paper presented at the 9th International Space Syntax Symposium, Seoul. Shoham, S., & Yablonka, I. (2008). Monumental Library Buildings in the Internet Era: the future of public libraries. IFLA Journal, 34(3), 266-279. doi: 10.1177/0340035208097227 Steadman, P. J. (2014). Building Types and Built Forms. Kibworth Beauchamp: Matador. Stonehouse, R. (2004). Composition and Context. In R. Stonehouse & G. Stromberg (Eds.), The Architecture of the British Library at St Pancras (pp. 43-79). London: Spon Press. Thomas, I. (2013). In the Library. London Review of Books, 35(8). Retrieved from http://www.lrb.co.uk/ v35/n08/inigo-thomas/in-the-library Till, J. (2009). Architecture Depends. Cambridge/MA: MIT Press. Turner, A., Doxa, M., O’Sullivan, D., & Penn, A. (2001). From isovists to visibility graphs: a methodology for the analysis of architectural space. Environment and Planning B: Planning and Design, 28(1), 103-121. Wilson, C. S. J. (1998). The Design and Construction of the British Library. London: The British Library. Wilson, C. S. J. (2007). The other tradition of modern architecture: the uncompleted project. London: Black Dog Publishing. Zong, W. (2015). Higher Education Libraries: Exploring the Library Users’ Activities Facilitated with the Change of Learning Conceptions. MSc Dissertation, UCL, London. Zook, J. B., & Bafna, S. (2012). Imaginative Content and Building Form in the Seattle Central Public Library. Paper presented at the 8th International Space Syntax Symposium, Santiago de Chile.

The dynamics and diversity of space use in the British Library

ITU A|Z • Vol 12 No 3 • November 2015 • 41-53

Success in Basic Design Studios: Can seat selection be an advantage?

Erincik EDGÜ erincik@gmail.com, erincikedgu@duzce.edu.tr • Department of Architecture, Faculty of Arts, Design and Architecture, Düzce University, Düzce, Turkey

Received: May 2015

Final Acceptance: October 2015

Abstract Socio-petal spaces have proven to be crucial for students’ social life especially in outdoor spaces and common gathering areas; however, actual design studio seating deserves to be examined as well. In various studies, it is revealed that there may be a correlation between seat location, seat selection and student performance. As social interaction is among the essential qualities of design education where training is based on table critiques and face to face discussions, studios ideally should provide the desired interaction. This research explores the students’ preference of seating assuming that it affects the consequent success of the student, in terms of social interaction and movement pattern, conducted in classically arranged rows and columns based studio layout, far from being ideal, where the movement pattern among the tables and the visual field become the most important modes of communication between students and instructors. The syntactic values of tables located adjacent to windows or aisles, middle rows, or back seats, front lines or wall corners help to determine the reason behind preference and selection of these seats. Integration values along with mean depth data are used to explore the socially active and passive sections of the studio layout, while isovists are examined to analyse the visual scope of each assigned seat. The results indicate that when the medium is crowded the position of the tables located alongside of circulation path gains importance. When the medium is less crowded, students prefer to prioritize their visual scope rather than physical accessibility.

Keywords Social interaction, Space syntax, Spatial preference, Studio layout, Visual field.


1. Social interaction in studios As is the case in all design programs, studio courses constitute the essence of architecture program. The data, inputs, outputs and problems of the spatial design issues are alike regardless of the scale differences. It is important for students to see at the beginning of their education that different variations may occur in designs and different results can be achieved. Since, it is essential that each student makes original design, drawing and presentation, in a design course multiple instructors may be present in order for each instructor to deal with the student individually and supervise the project development process, whereas the students can have the opportunity to acquire different design views. The design students are distinctive with their designed products, the equipment they use, working hours, patterns of behaviour and their perceived image. Thus, design students usually form Gemeinschaft society thinking, working, consuming and living together, as suggested by Dobriner (1969). This method of education necessitates a well balanced communication between the instructor and the student as well as a sociopetal form of behaviour where face to face seating arrangements may be used for both parties. However, in most cases the advantages of this unique method of teaching takes time for a first year student to notice and discover. Previous researches conducted by Ünlü et al., (2001 and 2009) indicate that sociopetal spaces have proven to be crucial for students’ social life especially in outdoor spaces and common gathering areas; however, actual design studio seating deserves to be examined as well. There are researches examining the students’ seating preferences in relation to territorial behaviour in various classroom layouts (Guyot et.al, 1980; Pedersen 1994; Kaya and Burgess, 2007; Costa, 2012). In these studies, territoriality is regarded as a behaviour mechanism occurred in public territory (Altman & Chemers, 1980) which is in fact related with self protection or defence (Sommer, 1969) rather than visual control. Miura and Sugihara (2011) emphasize in their research that large-sized classrooms

may decrease the learning effect on the basis that as the distance between the teacher and the student increases, it would be difficult for the student to pay attention to the teacher. Another study within the context of economics courses conducted by Benedict and Hoag (2004) showed that in large lecture rooms, students who prefer to sit towards the front of the room, have higher probability of receiving good grades compared to the ones sitting at the back. Perkins et al. (2005) conducted a seating research in the context of a physics classroom, where they have found that the initial seat location significantly affected student attendance, performance and attitudes. Through these studies, it is seen that there may be a correlation between seat location, seat selection and student performance. However, these studies are usually executed within conventional lecture halls where students are assigned either tablet arm chairs or desks where they need to express their individuality by controlling their environment. In case of design courses however, the relationship modes of the students with the instructors and with their peers change extensively. Miura and Sugihara (2011) define studio as a place where students constantly interact within a group, with their peers and mentors. As Webster (2008) and Dutton (1991) emphasize, architectural education orients students into some aesthetic and ethical values along with specific manners and language, in which peer motivation gains more importance compared to conventional lecture based methods. Social interaction is among the essential qualities of design education where training is based on table critiques and face to face discussions, therefore studios should provide the formation of desired interaction. Ideally architectural school layouts are supposed to provide the optimum settings as an exemplar for the design students. Especially design studio layouts equipped with movable drafting tables, computer stations, modelling spaces and reference shelves are considered to be a necessity for widening the scope of design intellect. However, especially in newly established institutions, limitations of ITU A|Z • Vol 12 No 3 • November 2015 • E. Edgü


the classroom facilities and teaching resources necessitate cases where the pragmatic solutions are deemed to be crucial. In these cases the courses vary according to the weekly schedule within the limits of the same studio space, where classically arranged rows and columns based seating pattern is observed. This type of layout indicates a focus on the instructor similar to a theoretical lecture; on the other hand, studentsâ&#x20AC;&#x2122; seating also gains importance on the basis of providing concentration on the individual work or keeping uninterrupted eye contact with the instructor. Also, in classically arranged rows and columns based studio layout, the instructorsâ&#x20AC;&#x2122; movement pattern between the tables and the visual field of students become the most important modes of communication between the students and instructors. The layout of the physical setting and the seating arrangements, are interrelated with the user behaviour patterns such as participation, social interaction and consequent success. Tables located adjacent to windows or aisles, middle rows, or back seats, front lines or wall corners have all various syntactic values in terms of integration. For example, in a study investigating the relation between privacy preference and the location of selected seats in a classroom Pedersen (1994) indicate that, students who chose to sit in the back of the classroom desire to be out of the visual field and wanted less

Figure 1. Student participation in a classic row and column layout (Adapted from Sommer, 1969).

involvement with others. On the other hand, seating pattern studies searching the best layout for prevention of cheating by Pomales-Garcia et al. (2009) have concluded that concentric rectangles and look away arrangements are better alternatives to traditional classroom seating. Prevention of cheating necessitates non-contact between the students; so this situation is just the opposite of what is expected and desired in a design studio. Sommer (1969) found out that in row-and-column arrangements student participation in the front row and in the middle of each row is the highest as it is indicated in Figure 1; while for example, in the U-shaped arrangement the class participation was the highest among students sitting directly across from the instructor. Kaya and Burgess (2007) on the other hand compare traditional setting and U-shaped arrangement in the context of social interaction. They emphasize that U-shaped configuration in classroom layouts generates an increased sense of community, eases discussion and promotes social interaction while, the traditional rows and columns layout helps the concentration especially on teacher centred lecture based courses. In their research, they have also concluded that in rows and columns layout, seats that are located on the sides are territorially claimed compared to middle seats. This finding may be similar to the situation of this research where the assumption includes that the tables located alongside the movement axis are considered as syntactically integrated and therefore are likely to be preferred by the students. Wang, et al. (2010) emphasize that the needed knowledge in architectural design studio, is dynamic and complicated, in a way that an individual studentâ&#x20AC;&#x2122;s knowledge is no longer sufficient to complete a good design project. McCormick, (2004) mentions the importance of knowledge sharing and resource exchange in dealing with complex design projects, whereas, Chiu and Shih (2005) emphasize the notion of peer to peer learning, indicating the importance of cooperation in a design studio as a learning alliance. These aspects are crucial to differentiate the seat

Success in Basic Design Studios: Can seat selection be an advantage?


selection of a design student who has to keep communication and social interaction both with the instructors and the peers, in order to come up with a good design. However, during the first few weeks of the freshman year, this social interaction is usually not settled yet. Ünlü et al. (2001) remark that social intelligibility of a space is not fully linked to social interaction level among users, but it is correlated to visual capacity of the environment. Thus, the research hypothesis assumes that regarding inexistent habitual attachment to specific seats, the low levels of acquaintances and yet lacking friendship bonds, the students are free to choose the seats they will occupy. Therefore, this research explores the students’ preference of seating on the basis that it affects the consequent success of the student, in terms of social interaction and movement pattern of instructors. The mentioned social interaction both with peers and instructors and the movement pattern of instructors are tested in an actually unfit medium for design studio with fixed physical layout of rows and columns. 2. Case study area and limitations Physical characteristics of studios in terms of shape or size, drafting table layouts, position and width of the circulation axes are among the important aspects of social interaction between the students and thus, seat selection. Referring to Georgiadou’s (2003), research done in the context of child care centres, in settings where internal configuration produces easily supervised

areas, there seems to be less rigorous control needed and so autonomy for children can be offered. This situation is similar in a design studio context; indeed it is observed that in studios with smaller dimensions and smaller cohort sizes, it is easier to maintain social interaction through discussions. However this is unfortunately not the case for this research. In this research, an actually unfit medium for design studio with fixed physical layout of rows and columns is tested on the basis of students’ social interaction and the movement pattern of instructors. The case study is conducted with the freshman year basic design studio students of Architecture and Interior Architecture departments of Cyprus International University; a privately owned university with a student population less than 10.000 located on a single campus. As mentioned before, lack of physical resources necessitates the studio to be kept in traditional row-column layout to enable theoretical courses to be conducted in the same location as well. As a combined hall of two smaller units, B221 (Figure 2) is the largest design studio of the Fine Arts building with dimensions of 7.8 m by 24.5 m almost totalling an area of 200 m². Although it faces a western sun, lacks acoustic comfort and ease of control for the instructors, with 76 numbered drafting tables, studio embodies the largest groups of students. The studio also has a white board on the northern wall and ceiling fixed computer controlled equipment pro-

Figure 2. Studio B221 existing layout. ITU A|Z • Vol 12 No 3 • November 2015 • E. Edgü


jecting on this white board as well. Therefore, in this context, any comparison between different studio layouts with differing drafting table organisation is impossible to explore. However, behaviour patterns of two different groups of students of two following years are compared in a longitudinal study. In this aspect it is also important to mention that in this research, the instructor group delivering the course and the studio remained the same while students changed. The data concerning the seating preference of the students gathered from weekly photographs taken throughout the first few weeks of the basic design studio courses in the 2009-2010 and 2010-2011 fall semesters. Photos from the 3rd, 6th, 9th and the 12th week of the semester are matched with the drafting tables the students preferred to sit and the grades that they have for that specific weekâ&#x20AC;&#x2122;s studio assignment. The selection of these weeks based on the exclusion of initial and final weeks to ensure attendance and midterm exam weeks because of a different time schedule. Keeping the distance of three weeks apart between the photos also made it possible for students to forget about the photo shooting and select their seats in a more randomly manner. The sample groups were all students of architecture and interior architecture departments however, cohort size of 2009-2010 was twice larger than the following year. This is due to the academic decision of separation of lectures into groups for a more flexible weekly schedule. When the number of students enrolled is smaller than the number of available seats, their scope of preference widens, and it would be possible to differentiate the logic behind seat selection. However when the cohort size is just barely equal to the number of seats available, then the first come first served rule applies, as it was

seen in the case of 2009-2010 fall semester studentsâ&#x20AC;&#x2122; seat selection. Basic design studios introduce a totally new world for the student with its own values and behaviours. It is important for students to see at the beginning of their education that different variations may occur in designs and different results can be achieved. As the studio is conducted with three to five instructors depending on the number of students, instructors take turns on attending to each student individually and students can have the opportunity to receive different design opinions. Thus, receiving critics from different instructors consolidates what the instructors have been pointing out. Therefore, studentsâ&#x20AC;&#x2122; interaction and visual contact within the studio space, both with peers and instructors were the crucial aspect of the research. There were two policies of the researched basic design studio; one of them was to integrate basic design with space using short-term and daily studies that would create a design identity on an individual basis and the second one was to plan longer-term projects of team work that would create a sense of belonging, shown in Table 1. Therefore, it is assumed that for the daily assignments students would seek social interaction with the instructors by means of table critiques. On the other hand, the assumption is opposite for the short term group studies. The students select seats within close vicinity of the groupmembers to bond with them, while they disregard social interaction with the instructors. However, in this research only the results of daily assignments are explored. Students who have failed in attendance and the ones who had not submitted more than one of the assignments of the observed week were excluded from the sample set. A total of 72 student grades from 2009-2010 fall semester and 36 student grades

Table 1. Basic Design Studio conception. Design Methods Studio Aims

Individual Work


Development of a Designer Identity

Development of Sense of Belonging

Success in Basic Design Studios: Can seat selection be an advantage?


from 2010-2011 fall semester are comWWpared on the basis of their daily individual applications, and the syntactic values of the seat positions that they have selected with regression analysis and Spearman’s rho correlation test in SPSS. 3. Space syntax methodology and analyses Space syntax is defined as the set of rules that generate different spatial arrangements (Hillier and Leaman, 1974; Hillier et al., 1987). Space syntax is also used as a theory and a method in order to define the structural environment. According to this theory, there are relations among the exterior forces and the social forces, which generate the forms. As for the architectural point of view, space syntax helps to understand the interaction of design objectives and characteristics with social restrictions and formal possibilities. The essential concept of syntactic approach assumes that the interior and exterior geometry of spaces are shaped according to certain cultural considerations and these forms also affect social relations in one way or another. According to Hanson & Conroy Dalton (2007), space syntax is built on three distinct spatial units, each having a different representation. These are the axial lines, convex spaces and visual fields called as isovists. Axial lines denote movement as movement is essentially a linear activity. Social interaction on the other hand, necessitates a convex space in which all points of space can be seen from all other points, or users. Using convex shapes, and axial lines, space syntax data can be calculated mathematically in order to represent, quantify and interpret spatial

configuration and visual perception. The University of Michigan registered software, Syntax 2D is used in for the analyses of the mentioned syntactic properties. In this research integration values along with mean depth data are used to explore the socially active and passive sections of the studio layout. In an architectural layout, integration denotes the socio-petal aspects, whereas the depth denote the opposite, almost hidden sections of the layout. On the other hand, visual scope that describes the visual area and the visual boundary of the users is another determinant to be considered. An isovist is the directly visible area within the space and the visual field changes when people move around in spaces. Therefore, both the visual scope of the instructors if seated on the assigned seat and the visual scope of the students on the preferred seats reveal the seen/unseen sections of the layout. When working with syntactic aspects, the initial concern was the movement of instructors and the accessibility of drafting tables by peers or instructors. The assumption was that the instructors can give table critiques or the student may stand up and go to the instructor or any other peer’s table for interaction. In this scenario, the position of the unmovable tables within the rows and columns layout was important. The drafting tables acted as blocking walls and they can only be reached by moving the assigned stools in front of them. Therefore, the integration analysis of the studio layout is calculated according to the blockage of the drafting tables (Figure 3). The tables just adjacent to circulation path in the centre and the ones with a room in

Figure 3. Studio B221 integration analysis. ITU A|Z • Vol 12 No 3 • November 2015 • E. Edgü


Figure 4. Integration analysis with people blocking the view.

Figure 5. Visual scope of each seat assigned for students.

front of them for an instructor to stop by and comment are assumed to be more accessible and therefore, should be initially preferred by students who seek interaction with the instructors or peers through movement. The second concern on the other hand, was the visual scope of each stool, i.e., the students themselves, in

relation to instructors or peers (Figures 4 and 5). In this scenario, the sitting mode was taken into consideration. Thus the positions of tables are neglected, as if the floor was raised to table height, while the position of the stools and so the students sitting on them, gained importance. The assumption here is again the students who seek interaction especially with the instructors would select the seats with wide visual range or other seats to keep an eye on the peers in case they come up with something interesting or such. Although the isovists have the capability of showing a visual scope of 360°, selected isovist nodes are all positioned to face the board, thus the instructors. In order to maintain this, the stools are considered as the blocking objects with students sitting on them. Therefore studentâ&#x20AC;&#x2122;s default visual field is set to be towards the front to communicate with the instructors and sideways to communicate with peers. The last concern here was the actual visual scope of the instructor in sitting mode (Figure 6). Although this specific position provides a single datum, it was considered important especially for the social interaction between the instructor and the students selecting seats from the front rows. Therefore, this analysis is conducted solely with the thirty four seats that are within the visual scope of the instructor. However, it is also interesting to see that as the students sitting next to the corridor seats at the back of the studio can still keep their eye contact with the instructors as well as their peers (Figure 5) consistent with the high integration values of these seats and longer isovist perimeters. On the other hand, students on

Success in Basic Design Studios: Can seat selection be an advantage?


Figure 6. Visual scope of the instructor.

the back seats are completely hidden from the instructor’s view while their large isovist area enables them to keep visual contact with their peers. 4. Conclusions and discussion The integration results show the accessibility of the tables, while the isovist parameters show visual scope of the students, as well as the instructor. Correlation results from the daily assignment average on the predetermined weeks vs. the related seat’s syntactic values are conducted separately for the fall semesters of both academic years. In all of the analyses, the students’ grades are considered as dependent variables, while the syntactic values are independent. The results of the

regression analysis are shown in Table 2. Regression analysis is investigated with the R values with significance between -1 and +1. It is assumed that the third week results would indicate a rather random range owing to lack of lesser prior experience, while following weeks would fall into a better range of correlations. Therefore, according to Table 2, integration level and success relationship is only seen on the relatively crowded group’s early settlement. The isovist area values of the seats present no correlation with grades, while isovist perimeters, i.e. the farthest distance that can be seen while working on the table are worth noting. Although the values shown in the table can be regarded

Table 2. Fall semester regression analyses from both academic years with df=75. integration

isovist area

isovist perimeter


R=0.225 (p= 0,05=0,05)

R=0.131 (p= 0,26>0,05)

R=0.101 (p= 0,384>0,05)


R=0.146 (p= 0,208>0,05)

R=0.093 (p= 0,422>0,05)

R=0.189 (p= 0,101>0,05)


R=0.017 (p= 0,887>0,05)

R=0.062 (p= 0,595>0,05)

R=0.221 (p= 0,055>0,05)


R=0.143 (p= 0,218>0,05)

R=0.034 (p= 0,770>0,05)

R=0.335 (p= 0,03<0,05)


R=0.07 (p= 0,551>0,05)

R=0.001 (p= 0,996>0,05)

R=0.093 (p= 0,423>0,05)


R=0.134 (p= 0,248>0,05)

R=0.02 (p= 0,863>0,05)

R=0.144 (p= 0,215>0,05)


R=0.201 (p= 0,082>0,05)

R=0.073 (p= 0,529>0,05)

R=0.306 (p= 0,007<0,05)


R=0.159 (p= 0,169>0,05)

R=0.062 (p= 0,596>0,05)

R=0.305 (p= 0,007<0,05)

3rd week grades

6th week grades

9th week grades

12th week grades

ITU A|Z • Vol 12 No 3 • November 2015 • E. Edgü


as mild correlations, the significance of isovist perimeter values versus the grades of sixth and twelfth weeks of both years is interestingly striking. When we compare the outcomes in the Spearman’s rho, the correlation between the 3rd week grades of the 2009-2010 fall semester with integration values shows a high significance with r(76) = 0.428, p < 0.01, complying with the regression analysis results of Table 2. However, integration vs the grades of this year are striking. We see correlation between the integration and 6th week grades of the 2009-2010 fall semester as r(76) = 0.284, p < 0.05, whereas 9th week grades and integration correlation is r(76) = 0.319, p < 0.01, and lastly comparing 12th week grades, a strong correlation appears as r(76) = 0.456, p < 0.01. This situation indicates that when the medium is crowded and early seat selection is crucial for interaction with instructors, then the position of the tables located alongside of the circulation path gains importance. On the other hand, for 2010-2011 fall semester, where the seat selection options were more diverse than the previous year, there appears to be a strong negative correlation with isovist perimeter, in the 6th week grades r(76) = -0.359, p < 0.01, indicating that the students have selected seats on the front rows and mainly next to wall or window. While the 3rd and 9th week results don’t show significance again complying with Table 2, there appears to be another inverse correlation for isovist perimeter in 12th week as r(76) = -0.369, p < 0.01 indicating a similar seat selection with the 6th week. The similar situation is also seen in the comparison with grades of the 9th week of 2009-2010 fall semester; where it gives us a negative correlation of r(76) = -0.250, p < 0.05. Inverse correlation means that there’s a relation between the isovist perimeters and the failure of the students instead of success. The instructor’s visual scope as shown in Figure 6, however, has not presented the expected correlations. Spearman correlations between the syntactic properties of the 34 seats that fall within the scope of the instructor’s visual field and the actual grades of the

students who have selected these seats initially showed that visual field of the students with respect to their proximity to seated instructors had no impact on the grades. While isovist area and isovist perimeter presented no connection to the obtained grades, isovist circularity showed a negative correlation of r (34) = -0.353, p < 0.05, from the 6th week of 2010-2011 fall semester. Benedikt (1979) describes isovist circularity as another measure of compactness or complexity of the visual field like area and perimeter which don’t change according to vantage point. This result may imply that if given a variety of seat choice, the students prefer to have a small amount of visual contact with the instructors rather than a full scope or none. While the extent of this visual contact is more important than the width of visual range, it still does not give any valid information about the success level of the student. Although the unequal size of cohort may necessitate cautious interpretations, there are still some interesting results to be discussed. The results of the research imply a relationship between the seat selection and grades, in terms of physical and visual accessibility. The integration based correlations are seen mainly in the 2009-2010 fall semester where the student group is large, and sitting on the preferred table is a matter of coincidence, unless the student intentionally comes to the studio earlier. In the case of 2010-2011 fall semester however, since the number of students are almost half of the number of seats, the students of this group have a wider range of selection. It is seen that these students prefer to take first rows for a higher level of social interaction with the instructors, seats alongside the circulation axis for easy access and seats alongside the wall or window for longer visual scope. It is also discerned that, different cohort sizes also seem to affect the success of the basic design education. The ideal ratio of design studio lecturer per student changes between 8-15 according to semester and level of design complexity, however, it is seen that meeting the quantity requirements does not automatically satisfy the desired design quality. Having a group with cohort

Success in Basic Design Studios: Can seat selection be an advantage?


size not exceeding 35 opposed to a larger group also verifies the test results as well. When the students have an opportunity to select seats from a variety of tables, they prefer to prioritize their visual scope rather than accessibility; however this visual scope is mostly related with peer vision or general panorama of either studio or exterior space. While in design studios successful students show no significance in seat selection unlike theoretical lecture halls, average and upper average students prefer easy access to instructors’ circulation paths and instructors’ visual field by selecting front rows. It was also assumed that the value of isovist area would be important as it denotes the width of the visual scope; however the results showed no significance. This would have been more important maybe in a lecture hall, where a clear view of the board or stage would be prioritized. However, the nature of any design studio also involves the movement of students as well as the instructors. Since usually, it allowed eating and drinking during the studio hours, the students select their tables for a longer period than any lecture based course. Therefore average or unsuccessful students seem to attach importance not to the easy accessibility of their tables either by the instructors or their peers but instead they prefer to have a longer visual axis, so as to control the instructors or their peers. That is why, for example if there is someone important for them, who is getting a critique from the instructors, they can easily come to listen as well, or check if someone is using a different material or having a better model. This situation also confirms the importance of information sharing and peer to peer learning through social interaction especially in design as denoted by Wang, et al. (2010); McCormick, (2004); Chiu and Shih, (2005). It may also be concluded that, regarding the student interaction thus desired peer to peer collaboration, traditional seating pattern with accessible movement routes, without walls and column like barriers that hinders visual scope can still be safely used in a studio layout.

References Altman, I., Chemers, M. (1980). Culture and Environment. Monterey, CA: Brooks/Cole. Benedikt, M. (1979). To Take the Hold of Space: Isovists and Isovist Fields, Environment and Planning B: Planning and Design 6: 47-65. Benedict, ME., Hoag, J. (2004). Seating location in large lectures: Are seating preferences or location related to course performance? The Journal of Economic Education, 35(3), 215–231 Chiu, S.H., Shih, S.G. (2005). Initiating and sustaining cooperation in an architectural design studio: An exploratory study with a Design Scope Model, National Science Council of ROC Taiwan, under the project number NSC 93-2211-E-011-033. Costa, M. (2012). Territorial Behavior in Public Settings, Environment and Behavior, 44(5), 713-721. Dobriner, W.M. (1969). Social Structures and Systems a Sociological Overview, Pacific Palisades, Ca: Goodyear Publishing Company, Inc. Dutton, T.A. (1991). Voices in architectural education: Cultural politics and pedagogy. New York, London: Bergin and Garvey. Georgiadou, Z. (2003). Question of social potential in space use. Proceedings of 4th International Space Syntax Symposium :69, London, United Kingdom. Guyot, G.W., Byrd, G. R., Caudle, R. (1980). Classroom seating: An expression of situational territoriality in humans. Small Group Behavior, 11: 120-128. Hanson, J., Dalton, R. C. (2007). Feeling good and feeling safe in the landscape: a syntactic approach. Proceedings Open Space-People Space: Innovative Approaches to Research Excellence in Landscape and Health, Edinburgh. Hillier, B., Leaman, A. (1974). How is Design Possible?, Journal of Architectural Research and Teaching 3, 4-11 Hillier, B., Burdett, R., Peponis, J., Penn, A. (1987). Creating Life: or, Does Architecture Determine Anything, Architecture and Behaviour, Vol.3, No:3, 233-250. Kaya, N., Burgess, B. (2007). Territoriality: Seat preferences in different ITU A|Z • Vol 12 No 3 • November 2015 • E. Edgü


types of classroom arrangements. Environment and Behavior, (2007): 39, 859-876. McCormick, R. (2004). Collaboration: The Challenge of ICT, International Journal of Technology and Design Education, (2004):14, 159-176. Miura, M., Sugihara, T., (2011). Effect of Students’ Seat Location on Programming Course Achievement, Knowledge-Based and Intelligent Information and Engineering Systems Lecture Notes in Computer Science, (2011): 6883, 539-547. Pedersen, D.M. (1994). Privacy preferences and classroom seat location. Social Behavior and Personality, 22, 393-398. Perkins, K.K., Wieman, C.E. (2005). The surprising impact of seat location on student performance. The Physics Teacher (2005): 43, 30–33. Pomales-Garcia, C., Carlo, H.J., Ramos-Ortiz, T.M., Figueroa-Santiago, I.M., Garcia-Ortiz, S. (2009). Non-traditional exam seat arrangements. Computers & Industrial Engineering 57(1), 188–195 Collaborative e-Work Networks in Industrial Engineering. Sommer, R. (1969). Personal space:

The behavioral basis of design. Englewood Cliffs, NJ: Prentice Hall. Ünlü, A., Özener., O.Ö., Özden, T., Edgü, E. (2001). An Evaluation of Social Interactive Spaces in a University Building, Proceedings, 3rd International Symposium on Space Syntax: 46, Eds. Peponis, J., Wineman, J., Bafna, S., College of Architecture and Urban Planning, University of Michigan, Ann Arbor. Ünlü, A., Edgü, E., Cimşit, F., Şalgamcıoğlu, M.E., Garip, E., Mansouri, A. (2009). Interface of Indoor and Outdoor Spaces in Buildings: A Syntactic Comparison of Architectural Schools in Istanbul, 7th International Space Syntax Symposium Proceedings: 132. Wang, W.L., Shih, S.G., Chien, S.F. (2010). A Knowledge Trading Game for Collaborative Design Learning in an Architectural Design Studio, International Journal of Technology and Design Education, 20(4) 433-451, Springer. Webster, H. (2008). Architectural Education after Schön: Cracks, Blurs, Boundaries and Beyond, Journal for Education in the Built Environment, Vol. 3, Issue 2: 63-74.

Temel Tasarım Stüdyolarında başarı: Yer seçimi bir avantaj olabilir mi?

ri tarafından farkedilmesi zaman alır. Öğrencilerin üniversite binalarındaki sosyal davranışları, birbirleriyle olan sosyal etkileşimleri ve toplanma alanları özellikle tasarım programları açısından son derece önemlidir. Ancak, Miura ve Sugihara’nın (2011) öğrencilerin arkadaşları ya da eğitmenleri ile sürekli olarak etkileşimde oldukları bir yer olarak tanımladıkları stüdyoların oturma düzenleri de araştırılmaya değerdir. Sosyal etkileşim masaüstü eleştirilere ve yüz yüze tartışmaya dayanan tasarım eğitiminin temel özelliklerindendir, bu nedenle stüdyolar istenen etkileşimi sağlayabilmelidirler. Öğrencilerin çeşitli sınıf düzenlerindeki oturma tercihlerini görüş alanından çok savunma odaklı egemenlik alanı davranışına bağlayan çalışmalar bulunmaktadır. Bu çalışmalarda büyük ölçekli sınıflarda öğrenci ile eğitmen arasındaki uzaklık arttıkça öğrenmenin güçleştiği, ön sıralarda oturan öğrencilerin arka sıralarda oturanlara oranla daha yüksek notlar aldığı,

1. Stüdyolarda sosyal etkileşim Tüm tasarım programlarının esasını tasarım probleminin verileri, çıktıları ile sorunlarının incelendiği stüdyo dersleri oluşturur. Öğrencilerin eğitimlerinin başlarından itibaren tasarımda farklı çeşitlemelerin ve sonuçların olabileceğini görebilmeleri gerekir, çünkü öğrencilerin özgün tasarım, çizim ve sunum yapması esastır. Bu nedenle, tasarım stüdyolarında öğrencilerin özgün tasarım süreçlerini tek tek ele alacak, projenin gelişimini denetleyecek ve öğrencilerin de kendilerinden farklı tasarım görüşlerini alabilecekleri birden çok eğitmen bulunabilir. Bu tür bir eğitim yöntemi yüz yüze oturma düzenleriyle dışadönük davranışı olduğu kadar, eğitmenler ile öğrenciler arasında dengeli bir iletişim kurulmasını da gerektirir. Ancak çoğu zaman, bu özgün eğitim yönteminin avantajlarının birinci sınıf öğrencile-

Success in Basic Design Studios: Can seat selection be an advantage?


yer seçimi alışkanlığının devama ve performansa etki ettiği gibi sonuçlara varılmıştır. Ancak bu çalışmalar her öğrenciye kolçaklı bir sandalye verilen ve bireyselliği hedefleyen kuramsal içerikli derslerin verildiği geleneksel dersliklerde yürütülmüştür. Tasarım stüdyolarında ise öğrencilerin gerek arkadaşlarıyla gerekse eğitmenlerle farklı ilişkileri vardır. Mimarlık okullarında stüdyo düzenlerinin hareketli çizim masaları, bilgisayar donanımları, maket tezgâhları ve referans kütüphaneleriyle öğrenciye örnek olması beklenmekle birlikte, özellikle tasarım eğitimine yeni başlayan kurumlardaki bazı kısıtlar daha yararcı çözümleri gerektirebilir. Aynı stüdyonun kuramsal dersler için de kullanıldığı geleneksel sıra düzeninde oluşturulmuş mekânlarda odak noktası eğitmendir ve eğitmenin masa dizileri arasında dolaşırken öğrenciler ile kuracağı göz teması önemlidir. Fiziksel mekândaki oturma düzeni, sosyal etkileşim, derse katılma ve kullanıcının başarısı ile ilişkilidir. Pencere veya koridor yanındaki oturma elemanları, ön, orta ve arka sıraların her birinin bütünleşme değerleri farklıdır. Mahremiyet odaklı bir çalışmada Pedersen (1994) arka sıralarda oturan öğrencilerin görüş alanından ve diğerleriyle etkileşimden uzak olmayı tercih ettiklerini, Pomales-Garcia ve diğerleri (2009) ise, herkesin birbirinden farklı noktalara baktığı merkezi dikdörtgen düzenlerin kopya çekmeyi engellediğini öne sürerler. Oysa, bu tür bir düzen etkileşimi azalttığı ya da engellediği için tasarım stüdyosunda istenenin tam tersidir. Diğer yandan, geleneksel sıra dizisiyle oluşturulmuş dersliklerin U düzenle karşılaştırıldığı çalışmalarda (Sommer, 1969; Kaya ve Burgess, 2007), ön ve orta sıralardaki öğrencilerin derse katılımının daha yüksek düzeyde olduğu ancak bu tür ortamlarda odak noktasının eğitmen olduğu, U düzende ise eğitmenin tam karşısındaki öğrenciler daha katılımcıyken ortamın kendisinin tartışmayı ve sosyal etkileşimi özendirdiği savunulur. Tasarım stüdyosunda gerek duyulan bilginin dinamik ve karmaşık yapısı nedeniyle öğrenciler arasında bilgi ve kaynak paylaşımının önemi artmakta (Wang ve diğ., 2010; McCormick,

2004) ve akranlar arası öğrenme ile mesleğin özündeki disiplinler arası çalışma alışkanlığı oluşmaktadır. Bu nedenle, gerek eğitmenler gerekse arkadaşlar ile iletişim kurmayı sağlayacak yer seçimi iyi bir proje ortaya koyabilmek açısından önemlidir. Ancak birinci sınıfın ilk haftalarında sosyal etkileşim henüz tam olarak kurulmamıştır. Bu araştırmanın hipotezi henüz belirli yerlere karşı oluşmamış olan alışkanlık, kurulmamış sosyal ve arkadaşlık bağları nedeniyle öğrencilerin yer seçiminde daha bağımsız olacakları ve sabit düzendeki bir stüdyoda kuracakları sosyal etkileşimin başarılarını etkileyeceğidir. 2. Araştırma alanı ve kısıtlar Kolayca denetlenebilen sınırlı ortamlardaki küçük gruplar arasında sosyal etkileşimin oluşması daha kolaydır ve tasarım stüdyoları açısından da bu istenen bir özelliktir. Ancak bu araştırma için seçilen alan tüm bu ideal şartların dışında kalan bir stüdyodur. Geleneksel sıra düzeninde 76 çizim masası alabilen 200 m²’lik bu stüdyo, özel bir üniversitenin birinci sınıf mimarlık ve iç mimarlık öğrencileri ile ardışık iki yıl yapılan temel tasarım dersinde kullanılan ve kısıtlı fiziksel kaynaklar nedeniyle kuramsal derslerin de aynı yerde yürütüldüğü bir mekândır. Eğitmen grubunun aynı kaldığı, 2009-2010 ve 2010-2011 güz yarıyıllarını içeren çalışmada, 3, 6, 9 ve 12. haftalarda çekilen fotoğraflarla öğrencilerin yer seçimleri belirlenerek bu haftalara ait uygulamalarda aldıkları notlarla karşılaştırılmıştır. İncelenen iki yarıyıldaki önemli bir fark da 72 ve 36 öğrenci ile ikinci yıla ait öğrenci sayısının ilk yılın yarısı kadar olduğudur. Yer seçimi havuzunu etkileyen bu fark sonuçlara da yansımıştır. Öğrencilerin eğitmenlerle ve arkadaşlarıyla farklı sosyal etkileşim alanını kullanacağı varsayılan günlük uygulamalar, stüdyo oturma düzeni, eğitmenlerin dolaşım aksı ile görüş açılarının dizimsel değerleri regresyon analizi ve Spearman korelasyonları ile karşılaştırılmıştır. 3. Mekânsal dizim yöntemi ve analizleri Mekânsal dizim çeşitli mekânsal düzenleri üreten kurallar dizisi olarak taITU A|Z • Vol 12 No 3 • November 2015 • E. Edgü


nımlanabilir (Hillier ve Leaman, 1974; Hillier ve diğ., 1987). Hanson ve Conroy Dalton’a (2007), göre mekânsal dizimde farklı gösterim biçimleri olan üç birim bulunur; bunlar hareket belirten aks çizgileri, sosyal etkileşim belirten dışbükey mekânlar ve görsel kapsamı belirten eşgörüş alanlarıdır. Mimari düzende bütünleşme değerleri dışadönüklüğü, derinlik ise gizli alanları belirtir. Bu çalışmada çizim masalarının dolaşıma ve erişime olanak sağlayan düzenlerini araştırmak için bütünleşme değerleri, sosyal etkileşim için ise gerek eğitmenlerin gerekse öğrencilerin oturur durumdaki eşgörüş alanları dikkate alınmıştır. Çizim masalarının erişilebilirliği ve eğitmenlerin masalar arasındaki dolaşım aksının bütünleşme değerleri için masalar sabit kabul edilmiş, her bir taburenin görüş alanı için ise masaların konumu gözardı edilmiş ve en geniş görüş açısı araştırılmıştır. Eğitmenin görüş açısına giren masalar ayrıca analiz edilmiş ve özellikle stüdyonun arkalarında yer alan masalardaki öğrencilerin eğitmenin görüş açısından tamamen gizlenmelerine rağmen arkadaşlarıyla göz kontağı kurmaya devam ettikleri için bu masaları tercih ettikleri görülmüştür. 4. Sonuçlar ve tartışma Yapılan analizlerde öğrencilerin notları bağımlı değişken, dizimsel ve-

riler ise bağımsız değişken olarak ele alınmıştır. Tablo 2’de verilen sonuçlara göre, başarı ve bütünleşme değeri ilişkisi ancak kalabalık grubun ilk haftalarında görülmüştür; buna rağmen öğrenci sayısının neredeyse masa sayısına eşit olduğu bu dönemde sonuçlar rastlantısaldır. Bir sonraki yılda ise, seçenek çok daha fazlayken, orta düzeyin üstündeki öğrencilerin eğitmenlerle etkileşim için öndeki masaları, kolay erişim için dolaşım hattı boyundaki masaları, geniş görüş açısı için ise pencere veya duvar kenarlarındaki masaları tercih ettikleri görülmüştür. Öğrencilerin seçenekleri fazla olduğu zaman fiziksel erişim yerine geniş görüş alanını tercih ettikleri; ancak bu geniş görüş alanının önceliğinin arkadaş ya da manzara olduğu, eğitmen etkileşimiyle ilişkili olmadığı görülmüştür. Eğitmenlerin eşgörüş alanlarının öğrenci yer seçimi ve başarısıyla belirgin bir ilişkisi görülmezken, stüdyo kuramsal derslere oranla içinde daha çok zaman geçirilen bir yer olduğu için, eşgörüş çevresi her iki yıl sonuçlarında da özellikle orta ve düşük düzeyde başarılı öğrencilerin tercihleri açısından anlamlıdır. Bu araştırma tasarım süreci açısından ideal olmamakla birlikte, erişim kolaylığı sağlanabilen geleneksel masa düzeninde oluşturulmuş stüdyoların da hâlâ kullanışlı olmaya devam ettiğini göstermektedir.

Success in Basic Design Studios: Can seat selection be an advantage?

ITU A|Z • Vol 12 No 3 • November 2015 • 55-70

Tracing a biennial layout: Experiencing an exhibition layout through the syntactic analysis of Antrepo No. 3 at the 2013 Istanbul Biennial Mehmet Emin ŞALGAMCIOĞLU1, Fitnat CİMŞİT KOŞ2, Ervin GARİP3 salgamcioglu@itu.edu.tr • Department of Architecture, Faculty of Architecture, Istanbul Technical University, Istanbul, Turkey 2 fitnatcimsit@gmail.com • Department of Architecture, Faculty of Architecture, Beykent University, Istanbul, Turkey 3 ervingarip@gmail.com • Interior Design and Environmental Design Department, Faculty of Art and Design, Istanbul Kültür University, Istanbul, Turkey


Received: May 2015

Final Acceptance: October 2015

Abstract The design of an exhibition gallery and its curatorial intension for either a temporary exhibition or a permanent museum installation requires understanding how its morphology influences the use of space, as well as the spatial experiences of visitors. The morphology of a gallery in terms of its shape and configuration may affect the display of the artwork, visitors’ activities and their movement through the space. This paper examines the layout of Antrepo No. 3, the main exhibition gallery of the 2013 Istanbul Biennial. This research explores: 1) how museum design influences integrated or segregated locations, as well as visitors’ use of the space during their visits; 2) how spatial layouts influence visitors’ explorations in gallery spaces; 3) which spaces are more or less visited; and 4) what the predominate path is depending on the number of visitors during a specific period of time. Answers to these questions are crucial for this study to understand the impact of the morphology of space on museum visitors. In this sense, syntactic correlations are key to grasping the idea of morphology and visitor experience relations in a curated space in exhibition design. For this investigation, gate counts within the exhibition space and snapshots showing the number of people and their patterns of interaction with the exhibition are correlated with syntactic parameters. Visitors’ spatial experiences and the use of the overall layout depending on the number of visitors during a certain period of time in a specific convex space are taken into account. Keywords Building morphology, Exhibition architecture, Space syntax, Spatial layout, Syntactic analysis.


1. Introduction Built space is composed of patterns that are interrelated through different syntactic and semantic layers, and museum space layouts are no exception. According to Peponis and Wineman (2003), built space is to be understood as a relational pattern supporting that situation: “A pattern of distinctions, separations, interfaces, and connections, a pattern that integrates, segregates, or differentiates its parts in relation to each other” (Peponis and Wineman, 2003). The built environment has a “social logic” (Hillier and Hanson, 1984) that relates to the layout, the pattern of the space that hosts the activities of daily life or special occasions such as temporary or permanent curated museum exhibitions. Spaces of the built environment such as museum spaces also structure social relationships such that society and culture become intelligible through their spatial form (Peponis and Wineman, 2003). 2. Antrepo No. 3 and the conceptual framework of the 2013 Istanbul Biennial Antrepo is a complex made up of 4 buildings previously used as warehouses at the Salıpazarı Harbour in the district of Tophane, Istanbul (Figure 1). The buildings have been used as venues for various Istanbul Biennials, and Antrepo 3 was in fact honored as the best venue of the Istanbul Biennials. Antrepo 3 has previously housed major artworks by contemporary artists such as Renée Green, Hung-Chih Peng, Yan Pei Ming, Ivan Grubanov, and Michael Rakowitz. The venue also houses art fairs as well as temporary exhibitions other than the Istanbul Biennials. The neighboring building, Antrepo No. 4, houses the Istanbul Modern Art Museum. The conceptual framework of the 13th Istanbul Biennial is based on the main theme, “Mom, Am I a Barbarian?” Curator Fulya Erdemci declared that, “The notion of the public domain as a political forum will be the focal point of the 13th Istanbul Biennial. This highly contested concept will serve as a matrix to generate ideas and develop practices that question contemporary

forms of democracy, challenge current models of spatio-economic politics, problematize the given concepts of civilization and barbarity as standardized positions and languages and, above all, unfold the role of contemporary art as an agent that both makes and unmakes what is considered public.” (Anon, 2015) This idea is grounded in diverse historic, philosophical, theoretical and geo-political ideas. Questions of democracy, equality, civic rights and political debate are interpreted through the works in the Biennial largely in the main venue, Antrepo No 3. “From the existence of an artwork to the freedom of social media and the designation of urban spaces as public, the notion of public domain can cover a vast area where social engagement and political public debate are possible. It is this potentiality of public domain discourse that the exhibition aims to articulate.” (Anon, 2015) The curated artwork in the exhibition is organized by several themes: fragility: “Am I Not A Citizen?”; spatio-economic justice: “How Is It Possible To Be ‘Rich’ In A World That Is Steadily Growing Poorer?”; agoraphobia: “Istanbul Is Ready, Target 2023”; the public domain as a battle-ground: “Conflict Or Consensus?”; and ‘Between Agency And Action’. Reflections of these themes are also seen in the works presented in Antrepo No. 3 (Figure 3); a list of exhibiting artists

Figure 1. 2013 map of Istanbul Biennial spaces, highlighting Antrepo No. 3 in relation to the Bosphorus, the Golden Horn and the historic peninsula of Istanbul.

ITU A|Z • Vol 12 No 3 • November 2015 • M. E. Şalgamcıoğlu, F. Cimşit Koş, E. Garip


is provided in this paper according to the locations of their artwork in the building (Figure 2). The syntactic values, such as the integration of visual and accessibility concerns in locating the artists’ work, the circularity values of the convex spaces in which the artworks exist, and the visitor frequencies depending on the convex spaces and in relation to the works of these artists are also considered in this paper. 3. Research goals and relationship to existing theory and previous studies This research seeks to explore the main venue of the 2013 Istanbul Biennial, Antrepo No. 3, through syntactic analysis and semantic explorations. This study includes the interpretations about the morphology of exhibition spaces’ in terms of building morphology issues and explores the relations between the frequency of people and syntactic values of spaces. The relation between the geometry of the spaces and the perception issues in relation with the movement of visitors in those spaces are important in this study. In that sense, the term “building morphology” includes the issues about the shape of convex spaces that people move in or through while they are visiting the exhibition space. The physical pattern and the overall structure, layout of the exhibition venue, in terms of geometrical measures and shape characteristics of its convex spaces are important building morphology issues in relation with the syntactic measures such as “circularity” and “integration”. Experiencing the temporary exhibition layout of Antrepo No. 3 (Figure 2 & Figure 3) is crucial to understanding the syntactic and semantic patterns that complement the social and physical patterns exposed in this study. Understanding the nature of Antrepo No. 3’s layout also requires grasping the idea of “the theory of natural movement” (Hillier, Penn, Hanson, Grajewski, & Xu, 1993), that the distribution of movement is a function of spatial configuration. The theory of “virtual community” (Hillier, 1989) is also a key to this understanding and “brings focus to a particular form of community that is based on the pattern of coawareness and copresence arising

as a by-product of movement” (Peponis and Wineman, 2003). From this perspective, museum spaces hosting temporary or permanent exhibitions have the opportunity to attract people from various communities and act as a gathering space and a space of information and burgeoning intellectual values. Here, the Istanbul Biennial also acts around these values and its main venue should be investigated with this awareness. In this study, Perception and movement related issues should also be grasped in order to understand the morphology of exhibition layouts through their syntactic and semantic dimensions. Kuipers et al.’s (2003) study on the cognitive maps of movement of the people describes how visual perception and cognition plays a key role in the processes of navigation, movement, and wayfinding. In relation with the museum and exhibition environments, visual perception and accessibility of the spaces also play a significant role in movement and wayfinding. Dalton’s (Zimring & Dalton, 2003) approach to decisions of people in terms of visual perception of the space that is similar to Kuipers et al.’s (2003) approach, Zimring & Dalton (2003) was interested in decisions that people head to during their navigation and in route choice decisions that are made at path junctions. “She created an environment in which participants were presented with a variety of different junction types and then noted the sequence of decisions.” (Zimring & Dalton, 2003). Dalton (Zimring & Dalton, 2003) found that “Angles that deviated least from a continuous straight heading were preferable to sharp turns.”. Another interesting finding was “a strong evidence that participants tended to select routes that approximated a straight line and avoided routes that were particularly convoluted or meandering.” (Zimring & Dalton, 2003). The study of Wineman and Peponis (2010) point out the issues that construct spatial meaning through visitors’ movement in museum and exhibition spaces, “The ways in which visitors are encouraged to move through an exhibition, whether along a clearly defined path or more freely weaving a self-di-

Tracing a biennial layout: Experiencing an exhibition layout through the syntactic analysis of Antrepo No. 3 at the 2013 Istanbul Biennial


rected path, will structure the overall impression of the exhibition.” Wineman and Peponis (2010) argue these two polarized point of view and introduces the term, “spatially guided movement” and evolve it to “spatially dictated movement” and “spatially random movement” from a third point of view in between these two polarized views. “Spatially guided movement” kind of understanding makes the connection, interrelation of geometrical space with the perception and movement in space. The spatial order and form of space is always a parameter effecting the perception and movement issues. Perception and understanding of visitors in exhibition spaces are constructed through “patterns of accessibility through the space of the exhibition, connections or separations among spaces or exhibition elements, sequencing and grouping of elements” (Wineman and Peponis, 2010). The issues related with perception of space and movement of visitors become more clear when we think of spatial relations based on patterns of access and choices effected by visibility, where the curatorial message of the exhibition is also a parameter effecting the decisions. The experience of the visitors when they start to move through the exhibition space is also unfolding based on the artwork and the order of spaces in relation with the content (Wineman and Peponis, 2010). “As a by-product of the exploration of museum content, visitors are seeing and being seen by others. Thus, museums function to construct a sense of community arising according to patterns of otherwise random copresence” (Wineman and Peponis, 2010; Hillier, Peponis, & Simpson, 1982; Peponis & Hedin, 1982). Actually, the circulation pattern and visibility sequence of the artwork were mentioned as key issues earlier in terms of cultural function in the literature (Gilman, 1923; Levin, 1983; Montaner & Oliveras, 1986). In this study, examining the physical layout of exhibition spaces became possible using techniques of spatial analysis. Space syntax analysis is based on isovists (Benedikt, 1979; Hillier and Hanson, 1984); visual perception is taken into account and measures of

Figure 2. Location of the artists throughout the exhibition space of Antrepo No. 3 at the 2013 Istanbul Biennial.

Figure 3. Various views from the exhibition space of Antrepo No. 3 at the 2013 Istanbul Biennial (View from Artist 42-Upper left; Artist 3&4-Upper right; Artist 2-Lower right; Artist 42&43-Lower left in relation with Figure 2).

ITU A|Z • Vol 12 No 3 • November 2015 • M. E. Şalgamcıoğlu, F. Cimşit Koş, E. Garip


accessibility and movement are also considered in the analysis. The key definitions of space syntax theory and methodology, such as isovists and convex space (Hillier et al., 1987), should also be introduced for a better understanding of the concepts of space syntax and exhibition space in this study. An isovist is a concept of spatial recognition that defines any particular viewpoint in a space by its visibility field; the visibility field of a single viewpoint can also be called the isovist field. Understanding the presence of the artwork in the exhibition space and the visitor frequency at the venue in relation to the isovist fields is crucial to understanding whether there is a correlation between the physical characteristics of the space and visitor frequencies depending on different convex spaces in the exhibition venue. As Hillier et al. (1993) note in Figure 4, beyond the relationship between visitor frequency and the configuration of the space in the exhibition area, depending on the morphology of convex spaces as a whole, while attractors and movement may be mutually influential, the other two relationships are asymmetrical. The configuration may influence the location of attractors, but the location of attractors cannot influence configuration. Likewise, the configuration may influence movement, but movement cannot influence configuration. If strong correlations are found between movement and both configuration and attractors, the only logically possible lines of influence are from the configuration to both movement and attractors, with the latter two factors influencing each other. In this

study, the relationship between visitor frequency and configuration is analyzed in detail; the attractors, namely, artworks by various artists, are also considered in this relationship through certain critical counts within the exhibition space. Various techniques of spatial analysis have been used to discuss the functions of museums (Peponis & Hedin, 1982; Wineman & Choi, 1991). Choi (1999) has analyzed visitors’ paths and found that integration was significantly correlated with “tracking scores,” the number of people who reached each convex space, and the correlation of tracking scores with “tracking frequencies” was investigated. “Spatial variables play an important role in structuring exploration even where the purpose of exploration is not to comprehend the layout itself but to view the displays in it. Choi also studied the distribution of people present in the museum, using normal behavioral mapping techniques” (Peponis and Wineman, 2003). 4. Method of analysis in relation to the aim of the study The space syntax method will provide significant data in terms of the method of analysis and is an important theory used to define the structural environment. Used as a syntactic measurement method in the space syntax field, the Syntax 2D software developed by the University of Michigan makes its calculations starting from a logical ground built over vision fields we term “isovists” (Benedikt, 1979; Batty, 2001; Conroy, 2001; Edgü et al., 2012). As for the concept of convex space, ap-

Figure 4. Attraction, configuration and movement (Hillier, et. all., 1993). Tracing a biennial layout: Experiencing an exhibition layout through the syntactic analysis of Antrepo No. 3 at the 2013 Istanbul Biennial


proached by scrutinizing interspatial relations within space syntax theory, it reduces the differently sized plans, spaces whose relationships will be examined as cellular spaces. In space syntax analyses, studies within the framework of a base logic that progresses by examining the relationships among these cells, or convex spaces, are performed. Syntax 2D handles the analyses based on isovists. Within the definition of an isovist, the walls, furniture, exhibition systems, artwork and other systems obstructing our sight in the space are handled as walls and affect the determination of the visual field (Benedikt, 1979; Turner and Penn, 1999; Batty, 2001; Turner et al., 2001; Conroy, 2001; Ünlü et al., 2009; Edgü et al., 2012; Salgamcioglu and Unlu, 2013). For this study, it is key to specify an analysis method in Syntax 2D that will allow us to examine the relationships among the convex spaces of Antrepo No. 3, depending on the exhibition venue and visitor frequencies counted separately for the specified convex spaces on weekdays and weekends, which will be described in the next data analysis section of this study. Syntax 2D works by creating a grid fragmentation. A plan proportional to the actual size of the site was drawn digitally using AutoCAD (.dwg) and transferred to the Syntax 2D program. This enabled us to compare the plan integration and depth comparison values through different convex spaces within the same plan. Visible, perceived field (on the plan platform) analyses were performed within this context. This research explores how integrated or segregated locations in a museum influence the installation of artwork, museum design and visitors’ use of space during their visits; how spatial layouts influence visitors’ explorations of a gallery space; how the integration value of a space affects the number of visitors to a specific gallery in the museum; the impact of visiting time (weekday or weekend) on the number of visitors to the museum during a specific period of time; which spaces are more or less visited and which artworks more or less viewed; the predominate path, depending on the number of visitors

on this path during a specific period of time; and whether visitor frequencies in the convex spaces of Antrepo No. 3 are correlated with the syntactic values of the spaces. Of the data generated as a result of the analyses, the data utilized for every convex space were: • Mean depth • Mean integration • Mean circularity These three data points are three of the primary concepts addressed in space syntax theory. These data were calculated separately for every convex space. Subsequently, the values at the active grids of the exhibition plan were separated into these three data groups and added on three different charts, and a mean data value was obtained for the three concepts (see Tables 1 and 2). Contingently, calculations were made via the arithmetic averaging of the grid values for every convex space. To understand the impact of the morphology of space on visitors, gate counts and snapshot analyses were undertaken to understand visibility relations, the regions described and isovists. Gate counts for 6 gates (see Figure 6) in the exhibition gallery, which provide access to the exhibition and circulation areas of the exhibition gallery, are taken into consideration on a designated route (see Figure 6) for several time periods. During these gate counts, snapshots are also used to analyze the visitor frequencies in each convex space shown in Figure 5. Snapshots are created using the observations for one weekday and one weekend day in Antrepo No. 3. For both days, observations for the snapshots and gate counts were repeated 8 times per day, starting at 11:00 am and repeating hourly until 7:00 pm.. At the beginning of every hour, the researcher walked the route shown in Figure 6 and counted the number of visitors in each convex space (Figure 5). After completing the visitor count observations by walking the route, gate counts were taken for a period of 5 minutes at each gate, starting from gate 1 and ending at gate 6, to find the number of visitors passing through the gates; this is also shown in Figure 6. These gate count values provide information about the movement

ITU A|Z • Vol 12 No 3 • November 2015 • M. E. Şalgamcıoğlu, F. Cimşit Koş, E. Garip


Snapshot 14.00 pm weekday

Snapshot 15.00 pm weekday

Snapshot 16.00 pm weekday

Snapshot 17.00 pm weekday

Snapshot 18.00 pm weekday

2.22 2.24 2.77 2.73 2.98 2.85 2.56 3.18 2.48 2.36 2.65 2.55 2.57 2.70 4.59 2.72 2.75 2.25 2.55 5.63 2.65 3.51 2.67 3.98 2.77 2.22 2.53 2.56 2.99 2.55 3.98 2.55 2.53 2.41 4.12 3.64 2.36 2.27 2.28 3.03 4.97

Snapshot 13.00 pm weekday

Mean circularity

Mean depth 161.52 151.65 62.14 56.12 29.67 92.59 82.03 52.66 180.15 221.84 82.88 141.57 120.83 93.32 19.07 127.12 124.98 277.37 178.87 20.85 118.42 48.11 150.67 23.50 87.65 204.53 203.25 133.46 46.70 115.98 27.83 141.29 101.73 202.07 45.52 35.28 129.52 180.00 159.37 4447.00 20.68

Snapshot 12.00 pm weekday

485785.00 461897.00 207877.67 177923.00 89785.00 42111.00 221267.00 25659.00 265665.00 302639.00 195888.00 307059.00 259832.00 190268.50 10535.00 194992.00 183702.33 334739.00 209709.50 2680.00 158177.00 39099.00 127174.00 22777.00 52023.00 333477.00 148304.00 136294.00 65502.00 244243.75 3377.00 160970.00 237413.00 312821.00 61925.00 10759.00 351161.00 380872.50 4474.86 33192.00 10412.00

Snapshot 11.00 am weekday

CS1 CS2 CS3 CS4 CS5 CS6 CS7 CS8 CS9 CS10 CS11 CS12 CS13 CS14 CS15 CS16 CS17 CS18 CS19 CS20 CS21 CS22 CS23 CS24 CS25 CS26 CS27 CS28 CS29 CS30 CS31 CS32 CS33 CS34 CS35 CS36 CS37 CS38 CS39 CS40 CS41

Mean integration

Convex space number

Table 1. Syntactic scores and frequencies (number of visitors) according to the counts made on a weekday at hourly intervals for all the convex spaces specified in Figure 5.

8 10 10 2 3 20 4 1 4 1 4 10 28 2 4 1 2 0 17 10 0 12 5 0 0 0 3 8 0 19 0 22 3 0 8 0 0 1 5 3 3

5 12 3 5 0 0 20 1 3 6 6 0 17 0 2 3 5 0 14 0 1 12 2 0 0 7 0 0 0 10 3 12 16 3 7 1 0 8 1 8 0

10 4 5 1 0 7 10 1 2 2 4 10 9 5 2 9 7 2 7 11 3 2 2 0 1 1 0 5 2 9 1 3 1 2 10 2 1 20 4 7 3

6 24 12 15 5 11 4 2 1 3 4 6 9 0 5 5 6 0 11 6 1 3 5 0 0 0 3 4 3 5 0 0 11 8 5 6 0 5 6 4 0

21 12 10 7 4 13 2 1 1 7 8 3 39 5 9 5 3 0 8 5 6 26 3 4 1 0 3 4 4 7 12 8 10 0 9 0 1 14 3 3 0

20 3 6 8 1 6 6 3 5 8 7 4 14 5 2 5 5 0 5 1 5 3 5 0 4 3 2 4 9 5 1 5 4 3 12 1 0 7 2 14 2

9 4 12 1 0 4 8 3 0 1 1 6 18 2 2 5 3 0 2 10 2 4 1 0 6 1 1 1 0 20 0 5 7 0 5 1 0 9 5 6 1

4 5 5 3 0 7 1 4 3 2 1 1 11 3 2 3 9 0 4 9 5 1 1 0 0 0 0 13 0 3 0 2 7 0 8 1 0 1 3 1 2

of visitors in the exhibition venue and will help us understand the movement between different groups of convex spaces depending on the syntactic and curational issues. The number of people passing through the gates is also

examined by analyzing the space syntax with Syntax 2D: gate counts are performed at the 6 gates, the visitor frequency is noted and the syntactic scores of the gates are also used to understand the relationship between the

Tracing a biennial layout: Experiencing an exhibition layout through the syntactic analysis of Antrepo No. 3 at the 2013 Istanbul Biennial


frequency of visitors and the syntactic measures such as integration, circularity and mean depth. To investigate the relationship between the number of visitors and the syntactic values of the exhibition space

Snapshot 14.00 pm weekend

Snapshot 15.00 pm weekend

Snapshot 16.00 pm weekend

Snapshot 17.00 pm weekend

Snapshot 18.00 pm weekend

2.22 2.24 2.77 2.73 2.98 2.85 2.56 3.18 2.48 2.36 2.65 2.55 2.57 2.70 4.59 2.72 2.75 2.25 2.55 5.63 2.65 3.51 2.67 3.98 2.77 2.22 2.53 2.56 2.99 2.55 3.98 2.55 2.53 2.41 4.12 3.64 2.36 2.27 2.28 3.03 4.97

Snapshot 13.00 pm weekend

Mean circularity

Mean depth 161.52 151.65 62.14 56.12 29.67 92.59 82.03 52.66 180.15 221.84 82.88 141.57 120.83 93.32 19.07 127.12 124.98 277.37 178.87 20.85 118.42 48.11 150.67 23.50 87.65 204.53 203.25 133.46 46.70 115.98 27.83 141.29 101.73 202.07 45.52 35.28 129.52 180.00 159.37 4447.00 20.68

Snapshot 12.00 pm weekend

485785.00 461897.00 207877.67 177923.00 89785.00 42111.00 221267.00 25659.00 265665.00 302639.00 195888.00 307059.00 259832.00 190268.50 10535.00 194992.00 183702.33 334739.00 209709.50 2680.00 158177.00 39099.00 127174.00 22777.00 52023.00 333477.00 148304.00 136294.00 65502.00 244243.75 3377.00 160970.00 237413.00 312821.00 61925.00 10759.00 351161.00 380872.50 4474.86 33192.00 10412.00

Snapshot 11.00 am weekend

CS1 CS2 CS3 CS4 CS5 CS6 CS7 CS8 CS9 CS10 CS11 CS12 CS13 CS14 CS15 CS16 CS17 CS18 CS19 CS20 CS21 CS22 CS23 CS24 CS25 CS26 CS27 CS28 CS29 CS30 CS31 CS32 CS33 CS34 CS35 CS36 CS37 CS38 CS39 CS40 CS41

Mean integration

Convex space number

Table 2. Syntactic scores and frequencies (number of visitors) according to the counts made on a weekend day at hourly intervals for all the convex spaces specified in Figure 5.

7 8 7 2 0 5 0 2 3 10 8 12 19 0 3 2 0 1 4 6 6 0 3 0 0 1 1 2 8 13 1 1 7 2 9 2 0 5 0 3 0

0 10 9 6 0 14 8 0 2 5 0 8 16 3 5 5 5 1 4 0 1 0 3 0 1 2 0 2 3 1 0 0 3 1 9 1 4 15 9 3 0

7 15 18 12 0 13 5 4 6 14 4 22 36 0 6 2 4 0 9 0 5 7 4 2 1 4 1 1 2 5 0 7 14 1 7 1 3 14 0 9 0

13 13 33 17 0 7 15 1 8 17 2 4 33 2 7 9 20 0 8 0 7 1 0 0 9 3 2 3 2 11 2 9 9 3 8 10 2 7 4 3 1

45 28 15 12 2 16 20 10 11 21 4 12 47 8 15 6 9 4 16 13 11 10 6 2 8 7 4 8 20 16 3 14 16 9 21 8 2 28 3 6 5

11 32 35 21 0 19 14 6 14 30 12 15 53 16 13 18 17 2 22 12 9 3 8 5 8 12 5 8 6 31 5 7 30 4 28 2 7 19 24 15 4

10 33 32 10 0 16 20 9 6 17 8 9 46 10 7 11 16 8 4 12 26 21 8 15 7 6 0 5 13 14 6 18 18 3 41 0 1 1 1 13 4

9 12 24 7 0 13 15 5 8 15 5 13 27 5 16 13 8 2 20 8 11 7 6 4 3 5 6 9 13 16 0 6 28 5 31 4 0 10 10 23 6

depending on the convex spaces, gate counts within the exhibition space and snapshots showing the number of people are used. Visitors’ patterns of interaction within the exhibition are correlated with syntactic parameters

ITU A|Z • Vol 12 No 3 • November 2015 • M. E. Şalgamcıoğlu, F. Cimşit Koş, E. Garip


and the results are discussed. During this investigation, visitorsâ&#x20AC;&#x2122; spatial experiences, their contact with exhibition content and the use of the overall layout is also considered to gain a better understanding of the relationship of syntactic measures and visitor frequency in the biennial venue. Finally, the statistical relationships between the number of people present during a certain period of time in a specific convex space and the syntactic measures of these spaces such as mean integration, mean depth and circularity are scrutinized and the correlations assessed using the Statistical Package for the Social Sciences (SPSS) program. 5. Data analysis: Comparison of syntactic values and visitor frequency As noted above, in this study, it is important to specify an analysis method in Syntax 2D to examine the relationships among convex spaces shown in Figure 5 depending on the exhibition venue, i.e., Antrepo No. 3, a schematic of which is shown in Figure 6. The visitor frequencies are shown for a weekday and a weekend day in Tables 1 and 2, respectively, counted at hourly intervals for the specified convex spaces. In addition to the convex space analysis, which depends on the visual field of the visitors and the accessibility to the spaces where artwork is exhibited, the gate count (see Table 3) analysis notes the visitor frequency (Table 4) and syntactic scores (Figure 7) for 6 gates to understand the relationship between visitor frequency and syntactic measures such as integration, circularity and mean depth. The number of visitors passing through a gate in the space of 5 minutes per gate at the selected hourly intervals in the daytime is counted to find the frequency data for gate counts. The gate count data are used to investigate the relationship between the syntactic values of the gates and the movement and circulation of people throughout the venue. In the process of this investigation, the statistical relationships between the number of people and the syntactic measures are examined and the addressed correlations further investigated using SPSS. In the first step of the

Figure 5. Convex spaces in the exhibition space in Antrepo No. 3 during the 2013 Istanbul Biennial according to visitor frequencies, shown in Tables 1 and 2.

regression analysis for Antrepo No. 3, the syntactic parameters of mean integration, mean circularity, and mean depth of the gates and the gate count frequency parameter (i.e., the number of visitors passing through the selected gate for 5 minutes; see Table 3) are considered. The dependent variable is always the number of people passing through the gates at the hourly observation points (see Table 3) and the independent variables are the syntactic measures such as mean integration, circularity and depth of the gates. When the most integrated gate in the system, Gate 4 (see Table 3), has the highest frequency (see Table 3) at 14:00 pm on the selected weekend day, the regression analysis of the frequency value and integration value shows a positive regression result of R=0.525 (p=0.285>0.05) that is mildly close to significance. This type of relationship can be interpreted as meaning that when the number of visitors passing through a more integrated gate rises in the Antrepo No. 3 configuration of the Istanbul 2013 Biennial, the number of people passing through all gates in the

Tracing a biennial layout: Experiencing an exhibition layout through the syntactic analysis of Antrepo No. 3 at the 2013 Istanbul Biennial


Table 3. Frequencies (number of visitors passing through a gate for 5 minutes 8 hourly intervals) according to the counts made on a weekend day (WE) and a weekday (WD) for various hours in the gates shown in Figure 6. Gate Number

Gate Counts / Weekday - Sunny WD 11.00am

WD 12.00pm

WD 13.00pm

WD 14.00pm

WD 15.00pm

WD 16.00pm

WD 17.00pm

WD 18.00pm

G1 G2 G3 G4 G5 G6

24 17 18 15 47 53

28 11 35 34 46 107

28 16 24 34 28 66

33 24 20 21 31 52

32 28 33 26 30 36

34 37 16 37 18 40

20 9 41 43 34 47

23 19 14 28 17 13










Gate Number

Gate Counts / Weekend Day - Sunny-Partly Cloudy WE 11.00am

WE 12.00pm

WE 13.00pm

WE 14.00pm

WE 15.00pm

WE 16.00pm

WE 17.00pm

WE 18.00pm










G2 G3 G4 G5 G6 Total

21 32 27 25 15 132

27 29 49 18 27 182

22 19 34 32 52 181

32 33 117 56 125 433

34 62 107 82 190 564

70 73 89 71 195 564

62 64 50 63 96 390

48 32 64 85 76 334

system also rises in accordance with the gates’ integration values, which has a slight significant correlation with the mean integration values of these gates in the system. One of the regression analysis results that supports this observation comes from the 11:00 am weekend count at Gate 4. At 11:00 am, the regression analysis between the parameters of the frequency value and the integration value in Gate 4 present no significance, with values of R=0.065 (p=0.903>0.05), because the number of visitors passing through Gate 4 is one of the lowest counts, which affects all the frequency values in the venue. These types of results, which are taken into account by the change in the frequency of visitors, show that when the number of visitors passing through a highly integrated gate in the venue increases, the number of people passing through all the other gates in the system display a mildly significant correlation with the mean integration value scores of these gates. When the number of visitors passing through Gate 4, which is the most integrated gate, decreases, the number of vis-

Table 4. The syntactic values of the 6 gate count nodes shown in Figure 6. Gate number G1 G2 G3 G4 G5 G6

Mean integration 353866.50 293775.00 247349.33 506976.50 453497.00 328015.60

Mean circularity 164.45 201.71 146.21 166.48 163.09 135.41

itors in the entire system may be still increasing—exemplified by the weekend visitor counts at 3:00 pm and 4:00 pm—but there are no significant regression results linking the integration values of the gates and visitor frequency at these hours. Visitors’ tendency to leave the venue also increases by using the only entrance to the venue at these hours, Gate 6. Therefore, the entrance gate, Gate 6, shows high visitor frequency at those times of day. The results for weekday counts are somewhat different. A negative regression result of R=-0.574 (p=0.233>0.05) that is mildly close to significance is seen at 3:00 pm for the frequency and

ITU A|Z • Vol 12 No 3 • November 2015 • M. E. Şalgamcıoğlu, F. Cimşit Koş, E. Garip

Mean depth 2.43 2.42 2.41 2.14 2.27 2.47


Figure 6. Antrepo No. 3 Plan of 2013 Istanbul Biennial showing the gate count nodes and selected route for snapshot counts on the left and plain plan of the venue on the right.

Figure 7. Image of the circularity (left) and integration (right) analysis made using Syntax 2D to find the syntactic scores for gate count nodes and convex spaces. Tracing a biennial layout: Experiencing an exhibition layout through the syntactic analysis of Antrepo No. 3 at the 2013 Istanbul Biennial


integration value numbers of the gates. On weekdays, the number of people passing through a less integrated gate increases at one of the peak venue hours in terms of the total number of visitors; in other words, the less integrated gates attract more visitors, but the most integrated gates attract fewer visitors. This interesting result is not seen at any other hours of the weekday regression analysis, but examining the results for 3:00 pm, it may be argued that the weekday visitors tend to visit the less integrated spaces of the venue. This may be due to the artwork exhibited in the venue. On weekdays, visitors may spend more time looking at the artwork presented in less integrated spaces by passing through the gates that are less integrated in the system. As for the second part of regression analysis, the mean integration values of all the convex spaces and the number of visitors counted during the daytime hourly snapshots on both weekends and weekdays serve as parameters. The syntactic measure is independent and frequency is the dependent variable. At 12:00 pm, R=0.325 (p=0.038<0.05) is positively significant; at 1:00 pm, R=0.278 (p=0.079>0.05) is mildly positively significant; at 4:00 pm, R=0.279 (p=0.078>0.05) is also mildly positively significant. Similarly, on the weekend, the results for 11:00 am, R=0.327 (p=0.037<0.05); 1:00 pm, R=0.377 (p=0.015<0.05); 2:00 pm, R=0.312 (p=0.047<0.05); 3:00 pm, R=0.472 (p=0.002<0.05); and 4:00 pm, R=0.313 (p=0.046<0.05) all present positive, significant values. The result for 12:00 pm, R=0.245 (p=0.123>0.05) presents positive regression values that is mildly close to significance. These findings support the effect of configuration on the movement and frequency of visitors in the convex spaces. The regression of the mean depth and frequency parameters also show positive significance at 4:00 pm on a weekday with the value R=0.357 (p=0.022<0.05) and positive values that is mildly close to significance at 6:00 pm on the weekend with the value R=0.256 (p=0.106>0.05). The regression of the mean circularity parameter and frequency parameter of the convex spaces presents

negative regression results that are also mildly close to significance at 12:00 pm and 4:00 pm on a weekday with the value R=-0.255 (p=0.108>0.05) and R=-0.238 (p=0.133>0.05), respectively; 12:00 pm on the weekend with the value R=-0.245 (p=0.123>0.05); 1:00 pm on the weekend with the value R=-0.265 (p=0.094>0.05); 2:00 pm on the weekend with the value R=0.254 (p=0.109>0.05); and 4:00 pm on the weekend with the value R=-0.236 (p=0.138>0.05). The correlation of circularity and frequency, which is mildly close to significance, shows the tendency that the number of visitors increases when the convex spaces are more linearly shaped. As a result of the test between the mean circularity parameter and the frequency parameter of the convex spaces as explained above, all have the probability value (p) above 0.05, but it is interpreted that the close numerical findings to 0.05 shows a statistically mild significance, a tendency, which is close to significant results. A decrease in the value of mean circularity is an indicator of gradual differentiation throughout a convex space of the general averages of the mean dimensions. Briefly, this value is accepted as an indicator of dimensional tightening with the increasing differences between the width and the length of the convex space or a non-differentiating, a tendency to dimensional equilibrium of width and length of the convex space. In that sense, the statistical evaluation carried out over different hourly intervals of some weekday hours like 12:00 pm and 4:00 pm and of some weekend hours like 12:00 pm, 1:00 pm, 2:00 pm and 4:00 pm shows that the number of visitors increases when the mean circularity values of convex spaces decreased. This shows the relation between the geometry of the space and the number of people visiting the space including their visual perception and movement. It could be inferred from this situation that when the differences among the lengths of the convex space morphologies in two dimension (i.e., width and length) began to decrease and that the plan center began to shift to the center of the convex space, the

ITU A|Z • Vol 12 No 3 • November 2015 • M. E. Şalgamcıoğlu, F. Cimşit Koş, E. Garip


people circulating in that specific convex space has the tendency to rise. Alternatively, with the findings of the increasing circularity values that is related with the geometry of the convex spaces, the convex spaces with different integration values throughout the plan began to overlap in a single narrow field with a high integration value and were forced to use this field as a connection field due to tightening of the interconnecting areas among the convex spaces. This kind of results should be interpreted and discussed particularly around the circularity value. As explained, at some weekday hours like 12:00 pm and 4:00 pm and at some weekend hours like 12:00 pm, 1:00 pm, 2:00 pm and 4:00 pm an increase is seen on the number of people where a morphological structure similar to a circle was observed due to the decrease in the mean circularity value. It was also observed that the general tendency toward a decrease in the mean circularity value was due to the divergence of areas connected with narrow interconnecting zones in the plan organization or more linearly shaped convex spaces in terms of the geometry of spaces. The mean circularity has a tendency to decrease in a structure in which the interconnecting areas among the areas are tightened, narrowed and concentrated on a single field with a high mean integration. The remaining relationships among the various syntactic value parameters and the remaining periods not described above are not significantly correlated with the frequency parameter. 6. Conclusion In summation, this paper traces the layout of Antrepo No. 3 at the 2013 Istanbul Biennial; as urban events, Biennial exhibitions can be discussed in terms of the interrelation of venues as well as that of art products and discussions. There are typically many exhibition venues at a biennial, and preferences are defined by pragmatic or thematic decisions. The goal here is to understand the performance of the main exhibition galleryâ&#x20AC;&#x2122;s layout syntactically to understand whether these layout decisions influenced the movement and number of visitors to the

convex spaces of Antrepo No. 3. Ultimately, it appears that the attractors and movement were mutually influential in Antrepo No. 3, but the configuration of the venue had the strongest impact on the number of visitors. For example, the mild significant correlation of circularity and frequency at some weekday hours like 12:00 pm and 4:00 pm and at some weekend hours like at 12:00 pm, 1:00 pm, 2:00 pm and 4:00 pm shows that the number of visitors increases when the convex spaces are more linearly shaped in terms of the geometry of spaces. Such findings show that the geometry of the spaces has some strong impact on the perception of the spaces, the frequency and the movement through those spaces. When the total number of visitors to the venue rises dramatically at certain peak weekend hours, the visitors may be looking more to the artwork in the more integrated space in the venue, passing through the highly integrated gates such as Gate 4. On the other hand, when the number of total visitors in the system decreases on weekdays, visitors are more likely to look at the artworks in less integrated spaces, passing through a less integrated gate, as shown by the results for 3:00 pm on a weekday. This result shows that the increasing density of people circulating in the venue decreases the number of visitors circulating in the system and passing through less integrated gates. This further shows the situation depending on certain selected hours, but it is important to see that the impact of the configuration is independent from the installation of the artwork in relation to the number of people circulating in the venue. This kind of finding is also directly related with the geometry of the space and shows the highly strong impact of architectural design on the circulation of people in such exhibition spaces. We may also argue that the configuration influences the location of attractors here, but that the location of attractors cannot influence configuration. When we examine most of the selected hourly counts of visitor numbers in the convex spaces, we find a mild significant correlation between frequency and the integration values

Tracing a biennial layout: Experiencing an exhibition layout through the syntactic analysis of Antrepo No. 3 at the 2013 Istanbul Biennial


of each convex space. Similarly, mean depth and circulation show mildly significant correlations with visitor frequencies. Finally, configuration may influence movement, but movement cannot influence configuration. We see the reflections of that relationship in the considerably significant correlations of gate counts with the integration values of these gates. There are some mildly significant findings pertaining to this exhibition space that contradict results from previous studies by Choi (1999) and Peponis and Wineman (2003), where no correlation between scores or frequencies was found, but the present study also shows the parallel idea that spatial variables play an important role in structuring exploration. Acknowledgements First of all we would like to send our sincere appreciation to Prof.Dr.Alper Ünlü for preparing such a dossier dealing with architectural design and space syntax issues. Next, we would like to thank some of the MSc students at Istanbul Technical University (ITU) from several programmes but mostly from “Architectural Design MSc Programme”, Fulya Menderes, Barış Ateş, Şebnem Çakaloğulları and, Ecem Çalışkan, who have collected the data used in this paper during long hours in the field voluntarily. Lastly, in addition to our volunteers from ITU, we would like to thank Industrial Product Designer Bedii Engin Koş, who was also a volunteer, for his support and work for data collection in the field. References Anon, (2015), http://13b.iksv.org/en Batty, M. (2001). ‘Exploring isovist fields: space and shape in architectural and urban morphology.’ Environment and planning b: planning and design, 28, 123‐150. Benedikt, M. (1979). ‘To take the hold of space: isovists and isovist fields’. Environment and Planning (B): Planning and Design, 6, 47‐65. Choi, Y. K. (1999). ‘The morphology of exploration and encounter in museum layouts.’ Environment and Planning (B): Planning and Design, 26, 241–250. Conroy Dalton, R. (2001). ‘Om-

nivista: An Application for İsovist Field and Path Analysis.’ Proceedings of The 3rd International Space Syntax Symposium. Atlanta: Georgia Institute of Technology. Edgü, E., Ünlü, A., Salgamcioglu, M. E., Mansouri, A. (2012), ‘Traditional Shopping: A Syntactic Comparison of Commercial Spaces in Iran and Turkey.’ In M. Greene, J. Reyes and A. Castro (Eds.), Proceedings: Eighth International Space Syntax Symposium, Chile: Santiago de Chile, January 3-6 Gilman, B. (1923). Museum ideals of purpose and method. Cambridge, MA: Harvard University Press. Hillier, B. ve Hanson, J. (1984), The Social Logic of Space. Cambridge: Cambridge University Press. Hillier, B., Burdett, R., Peponis, J., Penn, A. (1987). ‘Creating life, or, does architecture determine anything?’ Architecture and behavior/Architecture et comportement, 3, 233-250. Hillier, B. (1989). ‘The architecture of the urban object.’ Ekistics, 56(334/335), 5–21. Hillier, B., Penn, A., Hanson, J., Grajewski, T., Xu, J., (1993), ‘Natural movement: or, configuration and attraction in urban pedestrian movement.’ Environment and Planning B: Planning and Design, Volume 20, p. 29-66 Hillier, B., Peponis, J., & Simpson, J. (1982). National Gallery schemes analyzed. Architects’ Journal, 27, 38-40. Kuipers, B., Tecuci, D. G., & Stankiewicz, B. J. (2003). The skeleton in the cognitive map: A computational and empirical exploration. Environment and Behavior, 35, 81-106. Levin, M. (1983). The modern museum: Temple or showroom. Jerusalem: Dvir Publishing House. Montaner, J., & Oliveras, J. (1986). The museum of the last generation. London: Academy Editions. Peponis, J., & Hedin, J. (1982), The layout of theories in the Natural History Museum. 9H, 3, 12–25. Peponis, J. and Wineman, J. (2003), ‘Spatial Structure of Environment and Behavior’, in Bechtel, Robert B., and Arza Churchman, eds. Handbook of Environmental Psychology. John Wiley & Sons. Turner, A. ve Penn, A. (1999). ‘Making Isovists Syntactic.’ 2nd Internation-

ITU A|Z • Vol 12 No 3 • November 2015 • M. E. Şalgamcıoğlu, F. Cimşit Koş, E. Garip


al Symposium on Space Syntax. Brazil: Universidad de Brasilia. Turner, A., Doxa, M., O’Sullivan, D., Penn, A. (2001). From isovist to visibility graphs: a methodology for the analysis of architectural space. Environment and planning b: planning and design, 28, 103-121. Ünlü, A., Edgü, E., Cimşit, F., Salgamcioglu, M. E., Garip, E., Mansouri, A. (2009), ‘Interface of Indoor And Outdoor Spaces in Buildings: A Syntactic Comparison of Architectural Schools in Istanbul.’ In D. Koch, L. Marcus, J. Steen (Eds.), Proceedings of 7th International Space Syntax Symposium, (p.132). Stockholm, Sweden: KTH Royal Institute of Technology. Salgamcioglu, M. E., Ünlü, A. (2013), ‘Examining Space Transformation in

Apartment-based Housing Units in Istanbul using Space Syntax Parameters.’ Young Ook Kim, Hoon Tae Park, Kyung Wook Seo (Eds.), in Proceedings of Ninth International Space Syntax Symposium, Sejong University Press, Seoul, South Korea. Wineman, J., & Choi, Y. (1991, December). ‘Spatial/visual properties of zoo exhibition.’ Curator, 34(4), 304– 315. Wineman, J. D., Peponis, J. (2010). Constructing spatial meaning: spatial affordances in museum design. Environment and Behavior, 42: 1, 86-109. Zimring, C., & Dalton, R. C. (2003). Linking Objective Measures Of Space To Cognition And Action. Environment and Behavior, 35: 3.

Bir bienal yapısının izini sürmek: 2013 İstanbul Bienali Antrepo No.3 yapısının sentaktik analizleri üzerinden sergi mekanının deneyimlenmesi Sergi galerilerinin mimari tasarımlarını ve küratörler tarafından belirlenen içeriklerini anlamak, bu galerilerin geçici veya kalıcı sergi mekanları olmalarına bağlı olmaksızın, morfolojilerinin mekan kullanımlarını nasıl etkilediğini ve ziyaretçilerin mekana ait deneyimlerini de anlamayı gerektirmektedir. Bir sergi galerisinin morfolojisi, bu galerinin konfigürasyonu ve formu gözönüne alındığı zaman, bu alanda sergilenen sanat yapıtlarının sergilenme biçimlerini, ziyaretçilerin aktivitelerini ve bu ziyaretçilerin mekan içerisindeki hareketlerini etkileyebilmektedir. Bu araştırma insan hareketleri ve sergilenen sanat eserleriyle ilişki bağlamında 2013 İstanbul Bienali’nin ana sergi mekanı olarak kullanılan Antrepo No.3’ün kurgusunun ve mekansal konfigürasyon özelliklerinin irdelenmesini içermektedir. Bu araştırmanın amaçları: 1) Müze veya sergi mekanlarının mimari tasarımının, bu mekanlardaki sığ veya derin alanların oluşumunu nasıl etkilediğinin ve ziyaretçilerin mekan kullanımlarının anlaşılması; 2) Mekana ait planlamanın, mekanların bir araya geliş ilişkilerinin ve düzeninin ziyaretçilerin galeri mekanlarındaki keşfetme süreç-

lerini nasıl etkilediğinin anlaşılması; 3) Bir sergi galerisinde hangi mekanların daha çok, hangilerinin daha az ziyaret edildiğinin anlaşılması; 4) Belirli bir zaman aralığı içerisinde ziyaretçi sayılarına, frekansa bağlı olarak sergi mekanlarında baskın olan kullanım ve dolaşım rotalarının, mekanlar arası geçiş alanlarının özelliklerinin irdelenmesi olarak ifade edilebilir. Bu araştırma sorularına ve irdeleme alanlarına verilebilecek cevaplar, mekan morfolojisinin müze ve sergi ziyaretçileri üzerindeki etkilerini anlamak için de önem taşımaktadır. Bu bağlamda, mekan dizimi yöntemi kullanılarak yapılacak olan sentaktik analizler ve bu analizler sonucunda ortaya çıkacak olan sonuçlar son derece önemlidir. Mekanın tasarımından kaynaklanan, fiziksel durumuna ait veriler ile ziyaretçiler tarafından mekanın kullanımına ait frekans verileri arasındaki istatistiksel ilişkilerin ortaya koyacağı sonuçlar araştırılmıştır. Sentaktik analizler, mekan morfolojisi ile tasarımın ilişkisinin anlaşılması ve küratörler tarafından oluşturulmuş mekanlarda ziyaretçi deneyiminin nasıl olduğunun irdelenmesi için anahtar bir rol oynamaktadır. Bu araştırmada, Antrepo No.3 sergi alanı içerisindeki konveks mekanlar arasında geçiş oluşturan 6 önemli noktada ziyaretçilerin hareket frekansına ait sayımlar (gate counts) yapılmıştır. Buna ek olarak, 41 konveks mekan içe-

Tracing a biennial layout: Experiencing an exhibition layout through the syntactic analysis of Antrepo No. 3 at the 2013 Istanbul Biennial


risinde anlık olarak ziyaretçi sayısını belirlemeye yönelik sayımlar (snapshots) belirli bir rota üzerinde ve farklı günlerdeki farklı zaman aralıklarında sistematik olarak yapılmıştır. Sayımlar için hafta içine ve hafta sonuna ait birer gün seçilmiştir ve bu günler içerisinde 8 farklı zaman aralığında sayımlar yapılarak, ziyaretçi frekansları tespit edilmiştir. Mekanlara ve mekanlar arası geçişlere ait bu frekans değerleri ile bütünleşme (integration) ve merkezilik (circularity) gibi mekanın sentaktik değerleri arasındaki ilişkiler istatistiksel olarak, SPSS (Statistical Package for the Social Sciences) programı ile araştırılmış, sonuç olarak ortaya çıkan veriler yorumlanmıştır. Regresyon analizleri ile ilişkilerin ne derece kuvvetli veya zayıf oldukları sorgulanmıştır. Araştırma sonucunda Antrepo No.3 özelinde ortaya çıkan sonuçlar, daha önceki çalışmalarla ilişkileri bağlamında değerlendirilmiştir. Müze veya sergi mekanlarında mekanın morfolojisine ait veriler ile mekanın ziyaretçiler tarafından kullanım verilerinin sanat eserlerinin mekan içerisindeki konumları ile ne derece ilişkili olduğu sorgulanmıştır. İnsanların hareketine ilişkin teorik altyapı ile Antrepo No.3 ziyaretçilerinin mekanlardaki hareketlerinin ilişkileri tartışılmıştır. Bu tartışmada mekanın fiziksel tasarımı, plan kurgusu ve düzeni ile mekan içerisindeki eserlerin konumlanmasının ilişkileri de irdelenmeye çalışılmıştır. Bu ilişkilerin araştırılması aşamasında ziyaretçi frekanslarındaki artış veya azalmanın

mekanın geometrisi, mekanlar içerisindeki hareket ve kullanım ile nasıl bir ilişkide olduğu da yorumlanmıştır. Antrepo No.3 içerisindeki farklı konumlarda çekici noktalar oluşturacağı öngörülebilecek birtakım eserlerin esasen mekanın kendi morfolojisinden bağımsız çekim noktaları oluşturamadıkları bulunmuştur. Araştırmanın yapıldığı belirli zaman dilimlerinde Antrepo No.3 içerisinde dolaşan ziyaretçi sayısı arttıkça daha derin geçiş noktalarında ve konveks mekanlardaki ziyaretçi sayılarının düştüğü görülmüştür. Buna karşılık, antrepodaki ziyaretçi sayısının düşmesi ile mekanın tümünde dolaşıma giren ve daha derin mekanları kullanan ziyaretçi sayılarında artışlar olduğu gözlenmiştir. Bu bulgular ışığında, mekan içerisindeki kurgudan dolayı derinleşen konveks mekanlarda insanların frekansındaki artışla birlikte keşfetme ve ziyaret düzeyi düşmekte, ziyaretçiler daha sığ ve görsel algısı daha yüksek konveks mekanlarda daha yoğun olarak dolaşmaktadırlar. Ziyaretçiler, toplam ziyaretçi sayısının artmasıyla Antrepo No.3’ü oluşturan 41 konveks mekanda ve bunları bağlayan 6 ana geçiş noktasında daha sığ fiziksel özelliklere sahip, daha dar bir dolaşım alanında kalmaktadırlar. Bu araştırmada ele alınan bütün ilişkilerin yorumlanması sonucu ortaya çıkan veriler, sadece Antrepo No.3 sergi alanı için değil, aynı zamanda diğer müze veya sergi mekanı tasarım araştırmaları için de sorgulayıcı ve yol gösterici niteliktedir.

ITU A|Z • Vol 12 No 3 • November 2015 • M. E. Şalgamcıoğlu, F. Cimşit Koş, E. Garip

ITU A|Z • Vol 12 No 3 • November 2015 • 71-87

Thinking and designing with the idea of network in architecture

Nilüfer KOZİKOĞLU1, Pelin DURSUN ÇEBİ2 1 nilufer.kozikoglu@izmirekonomi.edu.tr • Department of Architecture, Faculty of of Fine Arts and Design, Izmir University of Economics, Izmir, Turkey 2 dursunpe@itu.edu.tr • Department of Architecture, Faculty of Architecture, Istanbul Technical University, Istanbul, Turkey

Received: May 2015

Final Acceptance: October 2015

Abstract A spatial setup is designed considering the network of interrelations between its constituent units. This is a network significant for architectural discourse as it maps the interactions and social relations between users, defines the functional and latent routes, and indicates spatial proximities. Although design is subjective, design tools and methods provide objective criteria to interpret and iterate. Common tools of network thinking allow us to invoke scenarios that will lead us to visualize and exchange ideas about architecture, extrapolate up to date functional ratios, define ranges of proximities to bring forth spatial and potentialities of architectural program and test them within criteria. This study focuses on the idea of networks in architectural design and discusses the use of graph theory based tools in the design process. It presents the possibilities of systematic mapping of relations among spatial elements through their neighboring and attracting qualities in the initial phase whereby the relational network is still dynamic and non-hierarchical. The topic will be expressed by presenting two examples, one from an academic setting, the other elicited from practice. The first describes a workshop on systems thinking demonstrated with a game called “İkidebir”. The second is an iterative hospital campus design scheme in which functional and site specific relationships are modeled and animated with network modeling and assessment tools. Network-based thinking, graphs measurements, and the diagrammatic assessment of relationships between spatial organizations as a design exercise are valuable both for those who are in practice and in the education of architectural design. Keywords Networks, Architectural design, Relational thinking, Space syntax.


1. Introduction: Space as a networked artifact Today, networks, which can be described as structural and organizational models, are pervasive in every aspect of our lives and range from genes to power systems and from social communities to transport routes. These networks are concerned with the structure of relations between things and are informative as they allow us to uncover those inherent principles and behaviors that regulate a variety of natural and artificial systems (Lima, 2011; Wigley, 2007). In the field of architecture, the study of networks has emerged as an inspiring concept in the description of built environments. After all, the design of a spatial setting inherently implies the network of interrelated spatial units, and so we can view the practice of architecture as mainly involved in the creation of the specific configuration of this network. In other words, the outcome of an architectural design process is essentially a configuration (Nourian et al., 2013). Network relationships are thus tools that the architect utilizes to propose his/her perceptions. These relationships once regarded as a mutable also constitute the potentials of encounters for the users through connections and borders, including even new ranges and thresholds. Thus they make up the base for the interactions and social relations between users, defining both functional and latent routes, and indicating spatial proximities and neighbors. According to Dovey and Dickson (2002), the spatial dispositions of buildings constitute social organizations. They are not formal types or archetypes, but, rather, clusters of spatial segments structured in certain formations with syntactic rules of sequence and adjacency. Lawson develops this view by defining architectural and urban spaces as containers that accommodate, separate, structure and organize, facilitate, heighten, and even celebrate spatial behavior. He says that space creates settings that organize our lives, activities and relationships (Lawson, 2005). Hillier suggests that buildings carry social ideas within their spatial forms (Hillier, 1996) and spatial formations can be seen as visual

symbols of societies. We read the space and anticipate a life-style (Hillier and Hanson, 1984). To date, most of the research studies that set out to reveal the potentials of network systems have utilized graph theory, a theory that relies on the conversion of information into a network diagram that can be mathematically analyzed to determine the relative depth or significance of the nodes or edges that make up the network (Ostwald and Dawes, 2013). Architectural applications of this method have also been developed by several researchers (Alexander, 1964; March and Steadman, 1971; March, 1976; Steadman, 1983; Hillier and Hanson, 1984; Hillier, 1996). Generally speaking, these works discuss some of the concepts of mathematics and diagramming or graph theory based tools that have potential value in understanding architectural forms and spatial organizations. They primarily present the architectural designer with some mathematical methods of conceiving and manipulating the spatial configurations. An analysis of utilizations of graph theory based tools in architecture suggest there are in three different modes: (1) to analyze existing spatial formation (Hillier et al., 1987; March and Steadman, 1971), (2) to generate spatial form, (Mitchell et al., 1976; Steadman, 1983), and (3) to evaluate architectural design (March, 1976; Hillier, 1998; Space Syntax, 2002). The first of these types of utilizations begins by exploring the intrinsic nature of the existing built environment and then decoding the underlying principles and meanings. The second group uses a series of predefined rules in a computerized, automated process to search for a desired spatial product. The last group provides tools that architects may use to evaluate their design proposals and also gives them opportunities to argue for the best performing proposals. The criticism leveled against these approaches mostly stems from the following questions: To what degree does an architect become involved in this cognitive process and how does he/she evaluate their designs considering desirable social implications

ITU A|Z • Vol 12 No 3 • November 2015 • N. Kozikoğlu, P. Dursun Çebi


rather than focusing on an automated evolutionary process? (Nourian et al., 2013). In the last decade an analytical approach, space syntax theory and its applications, has made great strides in showing architects the possible effects of their design solutions and have enabled them to learn from their design solutions (Dursun, 2007, 2012). In this way such utilizations constitute evidence-based design processes (Hanson, 2001). Space Syntax theory is constituted on two hypotheses (Dursun, 2012): 1. The built environment functions as a spatial / social network. In this network the main interest is about relational characteristics of spaces rather than individual ones. Space is experienced through this spatial networks or relations. 2. Spatial networks create potentials of movement and describe a living pattern. Movement is the key element to decode man-space / manman relationship. Based on this network structure spatial configurations embody social or cultural meanings and generate or inhibit social interactions, movement patterns in built environments. Such analysis tools guide in the comprehension and depiction of the relational structures however there has been complicacies in the transfer of this research-based knowledge directly to design process. The design process is conventionally perceptive, experiential and subjective. The method to reference research based knowledge to the design process is a typically a matter of concern for most. Dovey tries to explain this contradiction by focusing on relation between phenomenological philosophy and Cartesian world. He describes these poles “lived space” (the realm of personal feelings, emotions and particulars) and “geometric space” (the space of plans, forms and universals) (Dovey, 1993). According to Dovey, geometric space is a representation of lived space with the meanings and values extracted. For him, geometric space is a universal language of spatial representation that has predictive value. How can one creatively externalize the spatial knowledge in a measurable, visible manner for evaluation for assessment and improvement even from

the initial stages of the design process? Design is a complex cognitive process that continuously engenders both problems and solutions (Lawson, 2003). It is a kind of experimental process that is largely learned and practiced through “making” (Schön, 1987; Al-Sayed, 2012). Rather than searching for optimal solutions (Simon, 1996), design is about experimenting and probing. Experiments lead architects to discover something, and then these help them to redefine their underlying concepts (Dursun, 2007). In network thinking the investigation focuses on systematically mapping relations among spatial elements through their shared and relative characteristics, in other words, neighboring and attracting qualities in rule-based dynamic network models. The “relations of the relations” and “the protocol between the rules,” which refer to the order and the scale that the rules will be enacted during the design process, are of prime importance in these models. By observing the effects, the creative process can be interpreted as a kind of choreography, one in which “pace” is also interrogated for the elements of the parametric model. It is possible to deduct that relational qualities that suggests life inside a spatial construct i.e. social interactions and the movement (form prone to flow patterns) and proximities are built up by formal qualities defined by rule sets i.e. distance close or far, vertical positions, below or over, and whether clustering or disparate. Dynamic network models suppose that spatial entities are in constant motion during the design process. Their exact positions are yet ambiguous, they hang in air, and sway, or jump from one location to another; they start to presume specific locations and concretize as their relationships among each other become more and more defined. This study aims to explore following question: How we can use the idea of network in architectural design? By focusing on the experimental and intellectual characteristics of the design activity the study tries to examine how this kind of thinking can be used as a creative and informative tool in design process. In the scope of the study first,

Thinking and designing with the idea of network in architecture


the main question is opened for discussion conceptually with architectural students by the help of a game, İkidebir. Secondly, the authors try to explore how this kind of thinking can be utilized in the design process by focusing on an iterative hospital campus design scheme from practice. Cross indicates that design has its own distinct intellectual culture and has its own ways of knowing, thinking, acting (Cross, 2007). Based on this idea this experimental study aims to open a discussion about how, a scientific and graphic tool, network thinking and modeling, could feed the design thinking and making. 2. Playing the systems game – “İkidebir” Played by architecture students as a component of the Architectural Morphology class at the ITU Faculty of Architecture in 2014, “Ikidebir,” is a game in which simple rules make up a network where the nodes are in motion until they asymptotically settle into a configuration that satisfies the rule for each individual. This game engenders a dynamic system in a given space, and has the following rules: (1) Players initially announce an avatar, a spatial entity in this case, they selected for themselves and represent it as a node; (2) each player then selects two other announced nodes in order to follow in discrete this time. (3) All players randomly position themselves in the confined space (game area). (4) Hearing the start signal the players try to stand at equal distance to the two nodes whom they have picked to follow (Figure 1). The students first write their selected spatial entities and later draw the relations that form on the board, (Figure 2). Then the system is opened for discussion with the students. After introducing some analytical tools, space syntax and other dynamic network models such as cytoscape to decode this relational structure, the authors re-evaluated the process by the feedbacks of the students. Network can be described as a structure that is constituted by the links between nodes. These nodes can represent different entities such as individual person, object, space or concept. Both countable and non-countable entities

can be interrelated. For example in the first series of the workshops for this game in the Architectural Morphology class between 2008 and 2014, the students selected fictitious avatars and that had caused a more concentrated discussion on the nature of networks. However in the last workshop students selected to represent spatial units. Thus the composed networks lead the players to question the adopted relations that provide typical configurations. Recorded sessions are revealed at the blog: http://ikide1.wordpress.com. As soon as the game starts, players move in order to position themselves between their selected players. However as those players are also in movement, they continuously have to recalculate their target positions. This can be seen as a systemic flow, which sometimes accelerates and sometimes slows down. The simple rules create a dynamic set of nodes until the game settles into an arrangement that satisfies the rule for each player. Discussions with the students yielded the following key aspects:

Figure 1. Ikidebir Game – A Demonstration of Game Evolution.

Figure 2. Choice of Spaces and Relational Characteristics Visualized by Cytoscape Program.

ITU A|Z • Vol 12 No 3 • November 2015 • N. Kozikoğlu, P. Dursun Çebi


• The networked structure is informed by the choices made by the players. • The rule that the relation between spaces must be equidistant to the selected two spaces both triggers and organizes the motion. • The nodes, spatial entities in this case are fixed in terms of the links created whereas their geometric compositions are constantly changing. • By default each player selects two other spatial entities; therefore everyone is plotted into the network. • Only two spaces may be selected. This type of selection brings an important limitation for the interrelations among the nodes. Such that each space is connected by at least two spaces and remains linked with the whole. On the other hand some spaces are selected more than the others and this causes the system to lose homogeneity and leads it to a varied distribution. • The nodes selected by more nodes tend to be key elements in the system, in the given case “kitchen, entryway, and courtyard”. Their positions / or fluctuations affect the whole group causing both accelerations and decelerations. Therefore these nodes are latent to change the form of the system. • Some nodes – such as porch and sofa – are less significant to the system and thus they either may be selected by only a few players or even by none at all. Their actions do not create major changes. • However, even though they may be less-selected, some nodes – such as the winter garden – may prove to be effective, especially so when they are selected by a single player who is selected by many. By the introduction of graph the-

ory based tools such as space syntax and cytoscape to analyze the network structure the students tried to make this network legible and accessible to reading and assessing (Figure 2). Based on mathematical and graphical data, following questions are put into considerations: How do the selected nodes (avatars) behave in that particular system? How do they interact? How many connections do they have? What do they share? Are they interactive or are they inactive? Is there any key connection among them? Are there any groups or divisions (clusters) between them? The relational whole in the graphic and the calculated syntactic values, such as integration, connectivity, depth, choice, etc., rationally support the experience of the students’ perception of the choices (Figure 3). These explorations induce some valuable insights associated with the network structure: • In order to play students made random selections from spatial entities as avatars. The choices are mostly relevant in a residential setting, defining a quality or a program inherent to that space, like living room, kitchen, bathroom, WC, entryway, terrace, or nursery, or a few less common spaces like a cellar. The game also includes spaces more typical of traditional Turkish architecture, like the inner courtyard, iwan (vaulted hall) and sofa (connecting hall or egress space). • Hearing all the choices, students then selected two other two spatial avatars to be linked to from the available set in the group. These choices result in conventional relations such as sofa-courtyard, living room-kitchen, terrace-entryway, cellar-kitchen, kitchen-WC and some unusual relations such as liv-

Figure 3. The Relational Whole and Calculated Values Visualized by Space Syntax for Grasshopper - İkidebir. Thinking and designing with the idea of network in architecture


ing room-iwan, living room-bathroom, bathroom-kitchen, nursery room-WC. This allows the players to experiment on uncommon or secondary relationships. • These selections provide enough information to analyze and figure out the key entities in the network are “kitchen, entry way and courtyard”. These represent powerful nodes that have strong relations with the other nodes. The game also presented that these nodes initiated the motion and acted on the pace of the system. “Sofa and porch” tended to be inactive nodes. As they do not have strong relations with the other nodes, their effects on the spatial system are limited. Syntactic analyses clarify these characteristics. Integration values for the spaces reveal the following order: kitchen (2.636) > entryway = courtyard (2.197) > storehouse = winter garden = terrace = living room = WC (1.883) > bath (1.757) > iwan = stairs (1.647) > cellar (1.551) > foyer (1.318) > nursery = porch (1.255) > sofa (1.198). • Networks do not need to link nodes specifically of the same genre. Students’ selections included vague spatial entities like “entryway” as well as very defined ones like a “cellar”. • Networks by default defy physical dimension; however, discrete groupings suggest varying snap-

shots of spatial possibilities. Iterative playing out of the rule hints form possibilities including proportions, zones, interior and exterior build-up, etc. Specific network visualization layouts simulate part– to-part and part-to-whole relationships and spatialize the network in 2D (Figure 4). Visualizing the game with cytoscape, it is possible to visualize adjacencies and clustering possibilities, although the model is exempt of physical dimensions. In this workshop network thinking in architecture have been opened to discussion a. through students personal experience b. through graph theory related tools that analyze the demonstrated network. In other words abstract spatial network that emulates a spatial construct is experienced by the students participation and then examined in a cognitive scientific platform. The study imparts the following potentials network thinking in architectural design process: 1. The spatial whole can be described as the relations among its constituent parts rather than as a sum of disparate units. The manner in which these relations are constituted may infer diverse connotations and there may be quantifiable aspects of these relational patterns. 2. The rules that construct the network (one space must be selected by at least two other spaces) and the rules that enact on the form or the

Figure 4. Network from the Game Modeled in Network Visualization Program Cytoscape. ITU A|Z • Vol 12 No 3 • November 2015 • N. Kozikoğlu, P. Dursun Çebi


configuration (the relation between spaces must be equidistant to the selected two spaces) conform the flow and the proximities between nodes, is therefore constrained. Actual design processes include more complex and varied relationships and rules. However in both cases specified rules for distancing and clustering are indicative for the propositions of form. To understand the implications of these rule sets and their implementation is significant for the designer. 3. Certain relationships tend to be prevalent and affect change to the whole, whereas other clusters of relationships are not at all effective to the whole, yet are dynamic in their groupings. It will be argued that this phenomenon relies on the designer and the brief. The game is a demonstration of the relational make-up and the dynamic quality of these relations when they need to attain spatiality. This conceptual visualization or modeling enables the architect to consciously model, through play, the bonds and proximities of spatial units and the site. 4. Concentrating on the idea of network in architectural design, space syntax helps designer to develop spatial awareness by transforming relational spatial structures into graphical, mathematical and scientific forms. It explains what does these relations mean and how does the system works. By making non-discursive characteristics of space discursive, it presents a language for thinking and talking about space (Dursun, 2007). 5. While space syntax provides a useful tool for architects in deciphering and assessing the relationship among spatial entities in terms of spatial accessibility and human flow, other dynamic network models such as cytoscape and customized parametric modeling reveals possibilities regarding on geometric-formal characteristics of this relational whole. In other words, these models visualize the possible formal end products of applied rules. 6. Relevant graph theory concepts and criteria, diagrams, and produced

data sets based on effective representation of spatial systems lead to powerful instigation, management, and assessment of design phases. It is thus that, in contrast to convention, these tools have potentials to be tools with which we can think (Hillier and Hanson, 1997) during the morphological stages. These tools are creative and constitute an educational component within the research-based design. They also lead the designer to better understand the relationship between form and its use (function), while opening up new possibilities for design based on research results and generative principles (Schneider et al., 2013). 7. The experiment does not refer to the use of graph theory based tools including space syntax to extract potentials after the architectural form is solid rather during the initial stages of design. In this context it advances design thinking, enables interactive exploration of the effects of programmatic relations on form and suggests a method to structure correspondence of form and function. 3. Design Research: Method to design a campus The second example is taken from practice and deals with a conceptual design scheme for a campus on psychiatry and neurology. Hospitals have been the subject of a great deal of research in the architectural literature, especially in regard to their functional and organizational structures. Human flow and way finding issues appear key concepts of these researches (Ă&#x153;nlĂź et al., 2005, Setola, 2009, Khan, 2012, Peponis et al., 1990). The aim of our study is to impart potentials of network thinking explored in developing this master design scheme. In parallel to existing research, this scheme also focuses on the human flow in terms of vehicular and pedestrian pace between specific subunits. The programmatic and site relationships and relations to the varied qualities of the site are modeled and animated by the use of custom-made modeling tools based on network thinking. Peculiar qualities

Thinking and designing with the idea of network in architecture


of the site – such as emergency-prone segments along a major artery, or the more tranquil neighboring residential areas, and/or the security latent zones are represented as polar attractors. This design exercise incorporates interrelated positioning; programs are attracted (tied up) to specific zones or segments, and also to one another, as are the nodes connected to each other by the students’ choices in the case of the game. In the initial phase the separable programmatic units, regardless of their sizes, are scripted to move around an abstract container, pulling and pushing one another and the poles of the container in terms of their space/use related attributes. These disparate units are determined according to the administrative organization chart and the patient flow described by the clinical team. Attributed criteria to these units are urgency, security and privacy. Each program unit is specified with varying degrees of these attributes (Figure

5). “Urgency” pole attracts programs with emergency zones such as the emergency of the neurology hospital, privacy node attracted the acute psychiatric clinical program nodes, public pole pulled the outpatient nodes, and, finally, security node pulled forensic clinical nodes. By regarding these contained program units as a network, the script allows similar attribute grades to accumulate and the defined polarities to pull each other, and to move the groupings toward specified poles of the abstract container. The script also allows for negotiations among the varying degrees of these attributes. In the second iteration, shown in Figure 6, the group formations are clustered in the layout to allow propagation to the actual site. In this case the rules for propagation are parameterized by diffusion, overlap possibility, and size. The rule implementation follows a hierarchical order. Certain program units link to others like their satellite, and certain units have priori-

Figure 5. Conceptual Polarities Mapped in Relational Modeling among Program Units and Specific Attributes. ITU A|Z • Vol 12 No 3 • November 2015 • N. Kozikoğlu, P. Dursun Çebi


ty in maintaining proximity to the defined abstract polar zones. For example, the rehabilitation unit is a sub-unit orbiting the psychiatric clinics, bound by the everyday personnel and patient flow, which has been quantified as 300m walking distance. The emergency department has first priority to be in proximity to the main artery which is the main urgency pole (node). These interrelations and hierarchy are plotted and evaluated on a table matrix. Size is derived from the “List of Requirements” as well as the height and floor space limitations as a work area serving both a group of patients and a health team: 20-30 patients to one floor, as in the case of the neurology inpatient building – up 100 patients in total. The floors of the building floors are limited to eight in total, suggesting a footprint area of 1500sqm depicted with a circle with equal area (Figure 6). In the site implementation, the value sets and interrelated network are then mapped to the site directly referencing the preferred poles and axis to certain nodes. This ‘machinic’ diagrammatic exercise is modeled and run iteratively. The distances between units are defined in ranges proportional to time and the pace of pedestrian and vehicle reach (Figure 7). For example, the emergency pavilion for the three departments (psychiatry, neurology and neurosurgery) are located in the same spot; however, once a patient is to be transferred to an inpatient unit, the neurological unit is accessed via a flight of ramps and elevators, taking a total of ten minutes, whereas psychiatric patients are transferred by vehicle

to the psychiatric inpatient clinic. One is vertical in positioning whereas the other is horizontal. Each unit “behaves” and situates according to the specified rules, with emergency related units tending to prefer the artery neighboring zones, the inpatient units moving towards the residential borders, etc. The process is further rationalized with the use of a major axis for pedestrian and vehicular flow and its possible orientation on one hand and the variations provided by possible positions of a hypothetical center of the system on the other, certain units only following other units as satellites (Figure 8). This exercise is repeated in iterations for assessment of the resulting configurations. Units that are directly linked to site poles and units that have more links to other units have greater potentials in defining the working configuration. The position of an emergency plateau close to the major road is a straightforward design decision; however the role of the diagnosis department and its location to the other departments is one example where probing is necessary. The process enables fine-tuning and easy reassessments of multiple possibilities. The space syntax analysis also demonstrated that the diagnosis department is the key spatial unit in the network as it represents a powerful node that has strong relations with the other nodes. The rehabilitation block and inmate unit tend to be inactive spaces. Based on the syntactic analyses integration values for the spaces reveal the following order: diagnosis (4.435)

Figure 6. Matrix of Relations of Program and Site. Thinking and designing with the idea of network in architecture


Figure 7. Setup Order for the Abstract Programmatic Polarities Diagram.

Figure 8. Relational Site Model in Iterations.

> psychiatric inpatient = umatem = forensic (1.267) > neurology inpatient = amatem (1.109) > psychiatric outpatient (0.986) > rehabilitation block = inmate units (0.634) (Figure 9). The focus of these design research sessions is to be able to abstract and re-evaluate relationships regarding

the program, and the site, and reconstruct corresponding layout options with their interrelation degrees in reference to specific attraction criteria. These attractions and repulsions, in other words the polarized units, hint at building/structure-prone units by their capacity to conjoin and to cluster as

ITU A|Z • Vol 12 No 3 • November 2015 • N. Kozikoğlu, P. Dursun Çebi


buildings around a courtyard, for example. Through a series of assessment, multiple layout potentials are derived and compared. The process gradually narrows into a discursive scheme and potentials for the master plan are extrapolated. With the parametric configuration it was possible to convey the fact that the project model is only a snapshot of the possible set, and yet major decisions are more defined than others, and that there is room for development. It is thus a map for action (Figure 10). While the project was actualized in 2009, it is still under discussion as the stakeholders continue to bear a great burden of existing patients and economical strain; however, it is important to note that the project has remained viable, despite the passage of time and change of certain personnel. This is mainly due to the fact that the project is in itself a tool that allows evaluation of site and program conditions, and has the potential to change in accordance with modification of the site and evolving needs. The idea of network has been influential in the process o conceptual de-

sign scheme for this campus project. The clinical team asked for the project to correspond with the new understandings as well as the client required to evaluate all possible scenarios at the site. Both interests were met the project. The process imparts the following potentials of relational thinking in architectural design process: 1. Same as case one, here it is demonstrated that the spatial whole can be described as the relations among its constituent parts rather than as a sum of disparate units. The manner in which these relations are constituted may result is diverse consequences as to form and these relational patterns can be mapped in a quantifiable manner although they are based on concepts. 2. The rules that construct the network (common conceptual/spatial qualities that can refer to both the site and the functional units) and the rules that enact on the form or the configuration (the distances attributed between units as a function of pedestrian and vehicular motion) conform the flow and the proximities between nodes. Spec-

Figure 9. The Relational Whole and Calculated Values by Space Syntax for Grasshopper â&#x20AC;&#x201C; Hospital Campus.

Figure 10. Propagated Site Model. Thinking and designing with the idea of network in architecture


ified distances as well as clustering operations (orbiting, attraction to axis and positioning with hierarchy) are indicative for the propositions of form. To understand the implications of these rule sets and the order and the pace in which they are implemented are of value for the designer. 3. Certain nodes and links (relationships) are more effective to change the whole, whereas other clusters in the network are not at all effective to the whole, yet can be active in their local groupings. It is important to note that the designer takes his position rather than an automated generation of form in orchestrating the relational model. Also in relational models enable to demonstrate peculiar qualities of networks such as an overlooked unit linked to a major node has equal effect on the system, and therefore may act on the design discourse equally. 4. The project made use of custom-made parametric models that animated the network of both subjective (based on functional qualities) and objective (based on functional size, distance and orientation) relations. This enables to link attractive qualities with their corresponding spatial abstractions. This is made possible by the advance in the now ubiquitous digital tools that enable live change and tracking of parametric relational models. Thus it is possible to have variations as well as breeds of solutions to a brief. 5. This process requires a sustained assessment strategy for the variations arrived by the modeling. Space syntax or network visualization and assessment with software like cytoscape enable the assessment of the relationships among spatial entities in terms of spatial accessibility and human flow, as well as other network measurements like closest path, clustering, etc. 6. The process involves iteration: restructuring the initial relational setup, remodeling, and reformulating the physical ties (distances and ratios), reassessment of the order of rule enactment. It is crucial to the

process that the model is remade up after the initial run which serves more as a prototype to the machine-like dynamic model. 7. In this project case, the units were thought of as clustering similar attributes of spatial concepts like public/private together. However the pattern to distribute and propagate the units at the site could have been different then clustering the likes. The model only allows the designer to apply his design decisions in a prototypical manner that he can observe exceptions, derivatives, and possible modifications live on the model. 4. Conclusion Architectural design is ultimately about the configurations, connections, shape, and orientations of physical forms (Do and Gross, 2001). It deals with designing connections, borders, new ranges and thresholds in the space. Two case studies (one derived from architectural education and the other from architectural practice) are valuable both in terms of their effort to conceptualize the idea of network in design and to use this idea to trigger production of space in design process. Networks are dynamic forms in which relations are alive, in that they are in states of constant change. By exploiting this way of thinking in early stages of architectural design, it becomes possible to keep the negotiation alive, which is important for a creative process. This approach also provides informative tools for architects as it permits designers to see different potentials and possibilities in design and constitutes mediums for experimenting and probing. This study mainly concentrates on the idea that a critical understanding of the network in spatial constructs can inform, shape, and enhance the design. To exemplify the discussion, the authors first engage architecture students in a game designed to explore how a space paradigm can be conceptualized through a process of dynamic network rules. Secondly, the authors also try to explore how this kind of thinking can be utilized in the design process by focusing on hospital campus design

ITU A|Z • Vol 12 No 3 • November 2015 • N. Kozikoğlu, P. Dursun Çebi


scheme from practice. The first example aims to trigger the architecture students to develop the idea of spatial network in design by the help of a system thinking game in which all the students are actively involved. Network thinking in architecture is opened for discussion conceptually in order to decipher the potentials of the space and make its un-discursive, intangible characteristics discursive and tangible. This experiment constitutes a conceptual ground that permits the designer to understand the dynamic interaction among the design parameters, and also permits evaluation of the relationships and their meanings in the design. Here network thinking appears as a powerful tool in order to underline the notion that design activity is neither a closed box nor an automated process, but is rather an intellectual process in which the architect plays an active role as a spatial choreographer. The second sample concentrates on a design practice, one in which programmatic and site relationships are modeled and animated by customized modeling and assessed by graph theory based tools such as space syntax. The main aim here is to explore how relational thinking can be integrated into the architectural and urban design process. This example is important as it regards a need for a dynamic design instrument that can satisfy the changing needs in a long-term process. The architect can use the resulting parametric work and relational thinking to reveal and/or meet the requirements. In networks, nodes are not constituted from the same genre. They can be structured with different components including not only spaces, but also design criteria or concepts. This is an opportunity to link tangible with non-tangible qualities in a cognitive process. In terms of network thinking, the two experiments in this study are structured through three main stages: (1) Description of the relational structure, (2) Analysis of this structure and (3) Application of a rule-based design. The first process concentrates on achieving an understanding of how the networks are constituted and reveals the linking filter that organize these complex sets

of relationships. The second process deals with the analyzing or decoding potentials of the constituted networks. The third introduces a phase in which definite metric design rules are applied to the network of nodes. In this way, relational structures are transformed into spatial form from which the design proposals emerge (Figure 11). In the fırst stage of the game main determinant is the choices of the students. The constituted  network can be referred to  as a conceptual and nonhierarchical one in principal. In the example from practice however the spatial relations are structured by the clients  preferences and through data arrived from user questionnaires. Therefore in this case the network is not only a mental construct but also has physical impositions, yet they are also nonhierarchical in terms of their networking. In the following stages in both cases, the spatial potentials of the structured networks are expedited by network assessment and graph theory based tools that include space syntax. In the process space syntax imparts flow, transition, integration among spatial units whereas other dynamic network modeling whether analogue or digital set forth clustering, neighboring conditions and their meanings. Such graph theory based tools including space syntax appear as informative and creative tools to think, talk about and engage in space and spatial constitutions. In the third stage we can denote that form is designated by the enactment of the geometric rules. The operative rule is “to remain in the median axis of the other chosen two” in the game described in the initial sample, and in the next sample it is the distances designated for the units to satisfy in reference to one another. It is possible to say that design process is the iteration between these stages, i.e. the assessment of the “fixed” form and its consequences in the third stage are examined and tested with tools mentioned in the second stage. Therefore the process continues with the feedbacks of the second stage reconfiguring rule sets of the third stage and rerunning these relational metric rules. Network thinking equips architects with data regarding space and enhanc-

Thinking and designing with the idea of network in architecture


Figure 11. Talking about Two Experiments.

es their spatial awareness. Mathematical and graphical tools render previously invisible characteristics of space visible, measurable, and discursive. In respect to other generative tools for design network modeling in architecture can thus be transformed into a design tool with which the designer can freely think, play and model. References Alexander, C. (1964). Notes on the Synthesis of Form. Harvard University Press. Al_Sayed, K. (2012). Can Knowledge Inform Creativity? Paper presented in Design Creativity Workshop 2012, Texas A&M University, USA. Cross, N. (2007). Designerly Ways of Knowing. Birkhauser. Basel. Boston. Berlin. Dursun, P. (2007). Space Syntax in Architectural Design. Paper presented at the meeting of Sixth International Space Syntax Symposium, ITU Faculty of Architecture, 12-15 June 2007. Dursun, P. (2012). Dialog on Space,

Spatial Codes and Language of Space. AZ, Vol.9, 104-119. Do, E. Y., Gross, M. D. (2001). Thinking with Diagrams in Architectural Design. Artificial Intelligence Review, Vol.15, 2001 Kluwer Academic Publishers, 135-149. Dovey, K. (1993). Putting Geometry in its Place: Toward a Phenomenology of the Design Process. Building, Seeing and Designing: Toward a Phenomenological Ecology. Editor: David Seamon. Suny Series in Environmental and Architectural Phenomenology. Albany, NY: State University of New York Press, 247-270. Dovey, K., Dickson, S. (2002). Architecture and Freedom? Programmatic Innovation in the Work of Koolhaas/ OMA. Journal of Architectural Education, Vol. 56 (1), September 2002, 4-13. Hanson, J. (2001). Morphology and Design. Paper presented at the meeting of Third International Space Syntax Symposium, Georgia Institute of Technology, Atlanta. Hillier, B., Hanson, J. (1984). The So-

ITU A|Z • Vol 12 No 3 • November 2015 • N. Kozikoğlu, P. Dursun Çebi


cial Logic of Space. Cambridge: Cambridge University Press. Hillier, B., Hanson, J. (1997). The Reasoning Art: or, The Need for an Analytical Theory of Architecture. Paper presented at the meeting of First International Space Syntax Symposium, UCL, London. Hillier, B., Hanson, J., Graham, H. (1987). The Ideas are in Things: An Application of Space Syntax Method to Discover House Genotypes. Environment and Planning B, Planning & Design, Vol.14, 363-385. Hillier, B. (1996). Space is the Machine. Cambridge: Cambridge University Press. Hillier, B. (1998). From Research to Design. Urban Design Issue, Vol.68, October 1998, 35-37. Khan, N. (2012). Analyzing Patient Flow: Reviewing Literature to Understand the Contribution of Space Syntax to Improve Operational Efficiency in Healthcare Settings. Presented at the meeting of Eighth International Space Syntax Symposium, Pontificia Universidad Católica de Chile, Santiago de Chile, 2012. Lawson, B. (2003). How Designers Think. Architectural Press. Lawson, B. (2005). The Language of Space. Architectural Press, Oxford, UK. Lima, M. (2011). Visual Complexity: Mapping Patterns of Information. Princeton Architectural Press. March, L., Steadman, J. P. (1971). The Geometry of Environment, An Introduction to Spatial Organization in Design. London: Riba Publications Limited. March, L. (1976). The Architecture of the Form. Cambridge University Press. Mitchell, W.J., Steadman, J.P., Liggett, R.S. (1976). Synthesis an Optimization of Small Rectangular Floor Plans, Environment and Planning B, Planning & Design, Vol.3, No.1, 37-70. Nourian, P., Rezvani, S., Sariyildiz, S. (2013). A Syntactic Architectural Design Methodology. Paper presented at the meeting of Ninth International Space Syntax Symposium, Seoul, Korea.

Ostwald, M., J., Dawes, M. (2013). Differentiating between Line and Point Maps Using Spatial Experiences: Considering Richard Neutra’s Lovell House. Nexus Network Journal, Vol.15 (1), 63-81. Peponis, J., Zimring, C., Choi, Y.K. (1990). Finding the Building in Way Finding. Environment and Behaviour, Vol.22, No.5, September 1990, 555590. Schneider, S., Kuliga, S., Hölscher, C., Conroy-Dalton, R., Kunert, A., Kulik, A., Donath, D. (2013). Educating Architecture Students to Design Buildings From the Inside Out: Experiences from a Research-Based Design Studio. Paper presented at the meeting of Ninth International Space Syntax Symposium, Seoul, Korea. Schön, A.D. (1987). Educating the Reflective Practitioner. San Francisco: Jossey-Bass, A Wiley Imprint. Setola, N. (2009). A New Approach to the Flows System Analysis in the Teaching Hospitals. Paper presented at the meeting of Seventh International Space Syntax Symposium, School of Architecture and the Built Environment, KTH, Stockholm. Simon, H. A. (1996). The Sciences of the Artificial. MIT Press, Cambridge, Massachusetts, London, England. Space Syntax (2002). Tate Britain, Report on the Spatial Accessibility Study of the Proposed Layouts, Space Syntax Limited, July 2002. Steadman, J. P. (1983). Architectural Morphology. London: Pion. Ünlü, A., Ülken, G., Edgü, E. (2005). A Space Syntax Based Model in Evacuation of Hospitals. Paper presented at the meeting of Fifth International Space Syntax Symposium, Delft University of Technology, The Faculty of Architecture. Wigley, M. (2007). Architectural Brain. Network Practices, New Strategies in Architecture and Design, Editors: Burke, A. and Tierney, T., Princeton Architectural Press, 30-53.

Thinking and designing with the idea of network in architecture


Mimarlıkta ağ düşüncesi ile düşünmek ve tasarlamak Bir mekan kurgusunun tasarımı onun parçaları arasındaki ilişkiler ağının düzenlenmesi ile ilgilidir. Bu ağ yapısı mimarlık söyleminde kullanıcılar arasındaki sosyal ilişkileri, etkileşimleri resmettiği, mekanda fonksiyonel ve potansiyel rotaları deşifre ettiği, mekansal yakınlıkları gözler önüne serdiği için önemlidir. Mimari tasarım öznel bir süreç ise de kimi tasarım araçları ve metotları tasarımcıya tasarlananı değerlendirmek, öğrendikleriyle yeniden üretmek için nesnel kriterler sunar. Ağ düşüncesi içeren mekan kurma ve ölçme araçları mimarlığa ilişkin düşünceleri görselleştirme ve tartışmaya açmaya, verilerden mekansal ilişkilere dair yeni oranlara ulaşmaya, mimari programa yönelik potansiyelleri ortaya çıkaracak çeşitlilikleri tanımlamaya ve kriterlerle test edebilmek için senaryolar geliştirmeye olanak verir. Bu çalışma mimari tasarımda ağ düşüncesinin kullanılmasına odaklanır ve tasarım aktivitesinin deneysel ve zihinsel özelliklerine vurgu yaparak bu tür bir düşünme biçiminin tasarım sürecinde yaratıcı ve bilgilendirici bir araç olarak nasıl işlevselleşebileceğini araştırır. Çalışma, kural temelli dinamik ağ modelleri içindeki komşuluk ve çekim özellikleri yardımıyla mekanı oluşturan elemanların ilişkilerinin sistematik haritalanmasına yönelik araştırmalar sunar. Mimarlıkta ağ düşüncesine odaklanan çalışmalar incelendiğinde temelde üç amaçla kullanıldığı söylenebilir. (1) Var olan mimari biçimi anlama (Hillier ve diğerleri, 1987; March ve Steadman, 1971), (2) Mimari biçimi üretme (Mitchell ve diğerleri, 1976; Steadman, 1983), (3) Mimari biçimi değerlendirme (March, 1976; Hillier, 1998; Space Syntax, 2002). İlkinde varolan mekansal biçimlenmelerin kendilerini oluşturan dinamiklerin keşfi için analiz edilmesi hedeflenir. Tanımlayıcı ve açıklayıcı yönleri ön planda olan bu çalışmalar mimarın mekana ilişkin bilinç düzeyini arttırarak tasarım sürecini besleyecek bilgi birikimini çoğaltır. İkinci grup çoğunlukla bilgisayar odaklı, mekanik bir süreç içinde ve önceden belirlenmiş kurallar bütününde istenen mekansal biçimi

aramaya niyetlidir. Burada çoğunlukla üretilen biçimin nasıl bir yaşam biçimi kurguladığı sorgulanmadan tüm olasılıklar tasarımcının gözü önüne serilir. Son grup çalışmada ise tasarımcı üretilmiş mekansal kurgular arasında istenen kurallar, sınır şartlarına uygun en iyiyi seçme görevini üstlenir. Burada kritik olan ve çokça eleştirilen konu tasarımcının bu bilişsel sürece ne denli dahil olabildiği, mekanın belirli bir kural setini aramak ötesinde ürettiği olası yaşam senaryoları ile ne denli değerlendirilebildiğidir (Nourian ve diğerleri, 2013). Nitekim son donemde mekan dizimi çalışmaları tasarımcıya tasarladıkları mekansal kurguların nasıl yaşandığını göstererek, kendi tasarımından öğrenmesine, önerisini yeni düşüncelerle geliştirmesine olanak sağlamaya, bilgi temelli tasarım sürecinin de özünü biçimlemeye niyet etmiştir (Hanson, 2001; Dursun, 2007, 2012). Bu noktadan hareketle bu çalışma mimarlıkta ağ düşüncesinin tasarımcının birebir dahil olduğu bir interaktif araştırma süreci içinde tasarımın ilk evrelerinde, yaratıcı bir araç olarak nasıl kullanılabileceğine odaklanmaktadır. Yazıda bu olgu biri mimarlık eğitimi diğeri mimarlık pratiğinden seçilmiş iki deneyim üzerinden tartışılmıştır. Bunlardan ilki mimarlıkta ağ düşüncesinin kavramsal olarak sorgulandığı “ikidebir oyunu”dur. Bu atölye çalışmasında amaç, öğrencilerde ağ düşüncesine yönelik bir kavrayış ve farkındalık geliştirmektir. Mekana ilişkin oluşturulan karmaşık ağ yapısının ne tür potansiyeller ürettiğinin, ağın karakteristik özelliklerinin, bu ağ yapısının nasıl görünür, tartışılabilir ve de değerlendirilebilir kılındığının öğrencilerle birlikte irdelenmesi hedeflenmiştir. Sentaktik ve grafik-teorik araçlar oyunda kurgulanan ilişkiler ağını analiz etmek için kullanılır. Bu deneysel çalışmanın amacı mekan tasarımının belirli kurallar çerçevesinde parçalarının, parametrelerinin karşılıklı ilişkide olduğu bir sistem kurmak olduğunun soyut bir model üzerinden altını çizmektir. Yazıda tartışılan ikinci örnek ise yerleşim ve programa ilişkin kararlarının dinamik ağ modelleme araçları ile değerlendirildiği bir hastane kampüsü tasarımıdır. Bu deneyim söz konusu kavrayışın yani mimarlıkta ağ düşün-

ITU A|Z • Vol 12 No 3 • November 2015 • N. Kozikoğlu, P. Dursun Çebi


cesinin, mekanı kurarken, tasarlarken nasıl kullanılabileceği ile ilgilidir. Burada temsil edilen yerine deşifre edilerek aranan, potansiyelleri sınanarak geliştirilen bir mekansal kurgudan söz edilebilir. Benzer şekilde sentaktik ve grafik-teorik araçlar da mekanın potansiyellerini çözümlemek için kullanılır. Bu deneysel çalışma doğrudan,

üretilen bilgi ile sürecin beslendiği bir mekan yapma pratiği ile ilgilidir. Mimarlıkta ağ odaklı düşüncenin mekansal organizasyonların ölçülmesine ve bir tasarım araştırması olarak kullanılmasına yönelik olarak ortaya konan deneysel çalışmalar mimarlığı öğrenen ve gerçekleştirenler için değerli olacaktır.

Thinking and designing with the idea of network in architecture

ITU A|Z • Vol 12 No 3 • November 2015 • 89-103

Behavioral responses of the elderly regarding spatial configuration: An elderly care institution case study

Esra ÖZSÜT AKAN1, Alper ÜNLÜ2 esakan65@hotmail.com • Graduate School of Science, Engineering and Technology, Istanbul Technical University, Istanbul, Turkey 2 aunlu@itu.edu.tr • Department of Architecture, Faculty of Architecture, Istanbul Technical University, Istanbul, Turkey


Received: May 2015

Final Acceptance: October 2015

Abstract This article details a study presenting behavioral responses of elderly users regarding spatial configurations within an elderly care institution, using the “Space Syntax Method” within the framework of ”Environment Behavior Studies”. The case study reveals a variance of both adaptive and maladaptive behavioral responses towards spatial configurations regarding the elderly and their residential environment. A linear configuration is seen at the Maltepe Elderly Care & Rehabilitation Institution where behavioral responses are emphasized by analyses of how the design of elderly care institution corresponds directly with the spatial behaviors of elderly users’ overtime, utilizing the Observation Method. The linear configured elderly care institution shows an increase in behavioral responses in low level interaction (socio-fugal) areas whereas a decrease is seen in highly social interaction (socio-pedal) areas. Two different syntactic analyses are made with the inclusion of the garden area and without demonstrating the visual relationship with outer space. In doing such, it can be seen that both affective and behavioral needs of this particular age group into the design of elderly care institutions are a relevant parameter to be included as a conceptual framework within the architectural design process.

Keywords Behavioral responses, Elderly, Elderly care institution, Spatial configuration, Space syntax.


1. Introduction The aging of the population is an increasingly important issue in this century, especially considering how this issue affects all aspects of society including but not limited to health, social security, environment, architecture, socio-cultural activities and family. From the beginning of the 20th century, influenced by the age of industrialization, rapidly improving technologies, urbanization and an individualized life cycle there have been many changes in both family and social structures. The transition from extended family structures to immediate family which has been influenced by status and the decreased ability of function in the elderly has caused many of the elderly to become isolated. Factors such as the change in traditional family structure, death of spouse, decrease in income levels, not having social security, nor security in living alone have made the lives of the elderly living alone quite difficult. Therefore for these and other reasons, elderly care institutions have become the preferred living space for the elderly in recent times. However whereas trends may have changed in recent times, until now little importance has been given to studies on the spatial features of such elderly care institutions being sufficient in meeting the physical, social, psychological needs and expectations of the elderly. 2. Conceptual framework As part of “Environment Behavior Studies”, the conceptual framework includes the scope of “psycho-spatial” and “psycho-social” concepts as well as “architectural design” and “elderly user spatial behavior and responses” which due to their importance are being set forth and examined syntactically with the Space Syntax Method. Due to having a framework which holds multiple disciplines together, the importance of parameters regarding the study are being discussed with a transactional approach. 2.1. Environment behavior studies According to Altman (1975), the features of human-environment relations are classified in biological, physical,

psychological and socio-cultural levels. In research, the psycho-spatial processes concerning the elderly have multiple conceptual infrastructures. For this reason, separate identification of these concepts is important in terms of the behavioral spatial processes of the elderly users in order to understand and measure information concerning elderly care institutions. These concepts are the factors that form behavioral patterns concerning spaces and are discussed within the context of ecological harmony regarding the elderly and environmental stress, personal space, belonging, and social interaction, in order to form a space syntax relation. Osmond (1959) classified two kinds of conversational space; non-supportive socio-fugal and supportive socio-pedal. Being large, open, and expansive, with high ceilings and bright lighting, socio-fugal spaces tend to drive people apart and discourage social interaction. In opposition, he believed that smaller spaces with lower overhangs and close lighting, socio-pedal spaces tend to bring people together, encouraging conversation. Both these concepts are important regarding the elderly care institutions meeting their social needs. The literature suggests that older adults want to see rather than be seen, they sit in the areas surrounding open spaces, almost preferring the exterior edges of socio-pedal spaces (Sommer 1969). The primary target of analysis conducted is to understand the potential of physical spaces bringing the elderly users together dependent on their spatial behaviors. Within the scope of this article, the harmony of elder users’ spatial behaviors within spatial configurations form interactions at specific locations anticipated by the architect or reactions towards spatial configuration in which conversion or localizations of space in the different areas were seen. Within the context of Environment Behavioral Studies, Lawton and Nahemow (1973), as part of “Ecological Theory of Adaptation and Aging” evaluated the affective and behavioral conditions with personal efficiency, competence and the warnings received from the interaction between the physical and social environment or as a results of its relation with environmental

ITU A|Z • Vol 12 No 3 • November 2015 • E. Özsüt Akan, A. Ünlü


press. When a user reaching a certain level of competence shows behavior in an environment where there is also a certain level of press, the behavioral output is seen in the zone continuing between positive and negative zones. These zones are rated according to their level of behavior and influenced between competence and press is defined as the adaptation level. If environmental press reaches a high level, it means the competence level is also increased. Alternatively when there is a drop in environmental press, although behavioral output is generally positive if in fact it is of any significant measure the situation may lead to distress, sensory illness and behavioral anxiety, which is not organized (Ünlü, 1998). In that sense, elderly people in their institutional care environments show behavioral reactions against spatial configuration or harmony with their spatial behaviors such as when modifying the function of the space. Site Planning and Design for the Elderly, by Carstens (1993), did discuss issues relating to aging and presented recommendations for meeting elderly requirements and preferences (orientation and wayfinding, predictability and control, socializing, sensory stimulation, and environmental comprehension) and the practical requirements (safety and security, comfort, and physical, psychological and visual access).

Figure 1. Lawton & Nahemow’ s ecological model of aging (1973).

According to psychology, institutionalized elderly residing in a corporate environment are at a greater risk for depression. Long term care institutions recommend the elderly and their relatives to bring familiar and personal belongings in order to personalize the environment and alter the perception of the environment. Yet although significant, a negative impact is still seen (Eshelman et al., 2002). However, regarding interior decoration objects, wallpaper upholstery materials combined with a friendly environment similar to a home image accentuated with the use of lighting elements aids it giving a feeling of familiarity and safety (Kopec, 2012). According to Zeisel’s (2005) case study in the Alzheimer’s Assisted Living Treatment Residence, both the design and layout were modified for people living with Alzheimer’s disease and the architecture, landscape and interior were planned to augment memories and the ability of self or auto-functioning. By taxing the areas of residents’ brains which functioned well and relieving damaged areas, the entire individual was supported. Residents felt at home, competent and in control as much as their age allowed. 2.2. Social structure of space and space syntax based theories Space syntax is a theory and methodology used to define structural environments. The theoretical base was first set forth by Hiller and Hanson (1984) in a book entitled “Social Logic of Space” in which the thesis states there is a relation between outside factors that generate forms and social powers. According to Hiller and Hanson (1984) the biggest obstacle in creating better designs lies in the fact that the relation between social structure and spatial organization was not being fully understood. In order to achieve this more emphasis must be placed on the interdisciplinary literature of space and society. The Space Syntax theory is used as a parameter of spatial scheme defining behavioral changes, cultural differences and social functions. This theory is used within the context of elderly care institutions by measuring the design forms of buildings as well as

Behavioral responses of the elderly regarding spatial configuration: An elderly care institution case study


the interaction between elderly users, which aims to provide readability from an architect’s perspective. According to Seamon’s (1994) phenomenological approach, humans are in close relationship with the world and it is believed that the two create and reflect upon each other. For example, long, narrow roads are cold, give little feeling of space and are perceived with their one-dimensional axial shapes, movement flow and circulation area. On the other hand, large convex spaces are the places where elderly people rest, children play and district bazaars are set up. If axial spaces are mostly connected with gradual change and interaction amongst residential districts and neighborhoods, convex places are connected with the meaning of these spaces which provides an opportunity to read architectural plans with Space Syntax analysis (Edgü, 2003). In a study regarding the elderly conducted on the remodeling of a care institution, progressive privacy is turned towards access to public and common areas as much as possible such as outside sheltered areas and can be controlled entirely by the user dividing the space into subsections (Trotter et al., 1998). The Progressive Privacy Model with public, semi-public and private areas separated into different zones allows balance to be kept in the levels of access and control. Concerning Wojgani and Hanson (2007) their configuration in which they redesigned an elderly care and rehabilitation center, located general spaces around the main entrance where social interaction was strong, removing personal spaces that enabled privacy from the entrance. With that configuration, they identified the physical features of spatial configuration as well as determined the social interaction of the space. In this context, we see a move away from a home environment towards one that involves more experiences forcing users into a new environment for the elderly which carries the perimeter which includes psycho-spatial behaviors. In the thesis study that formed the basis for this article, the hypothesis was that spatial configuration in elderly

care institutions is a determining factor on spatial behaviors and perception of elderly users. Therefore, with that in mind elderly care institutions should be analyzed in terms of environmental perception combined with the influence of spatial configuration which is demonstrated in this case study. It has been shown that spatial configuration does in fact cause behavioral responses on elderly users and the relation of spatial configuration in elderly care institutions concerning behaviors of these users are set forth and examined at a syntactic level. 3. Case study In this case study, to what extend spatial configurations conflicted with spatial behaviors of the elderly and the level of relation between them have been emphasized as well as the adaptive or maladaptive behavioral responses imposed from the elderly. In the elderly care institution that was chosen for this case study, behavioral responses of elderly users were emphasized by analysis of the influence of spatial configuration on spatial behaviors and responses as well as the interactions between them. Behavioral patterns within the scope of the architectural program were determined by observational method and perceptional features were digitized within Space Syntax parameters including the interaction between space configuration and users responses. By interpreting usage frequency of the space with a syntactic value and socio-pedal or socio-fugal space characters, the developed adaptive or maladaptive behavioral responses of users towards spatial configurations were shown. The Maltepe Elderly Care & Rehabilitation Institution (MECRI) was designed by Yalçın Emiroğlu in 1975 and included in the complex was a residence for mobile and semi-mobile elderly with two units for eighty to one hundred people, a rehabilitation center, social services unit, central kitchen, laundry, infirmary and recreational facility. Each unit has three floors with gardens and sitting areas between them. The structural complex has a linear plan scheme on a horizontal settlement plan (Figure 2).

ITU A|Z • Vol 12 No 3 • November 2015 • E. Özsüt Akan, A. Ünlü


Figure 2. MERCI, layout and ground floor plan.

Figure 3. Views from the observation points on the ground floor.

Figure 4. Views from the observation points on the first floor.

With regards to the scope of this case study, Block A and C apartment units are discussed and have living units that generate a linear layout by lining up across along a corridor. All living units have a balcony and face south overlooking the landscape. The social block unit was planned separately from the living units, connected to the administration and in the entrance lies waiting, secretary, manager and meeting rooms. Further on the social area unit is accessible where in lies a multi-purpose hall, service area and cafeteria (Figures 3-4). 3.1. The identification of the method of case study The relation between numerical values obtained and social structure are objectively identified. During identification of this relation, outcomes were

obtained from two stages of analyses providing definition of the different point characteristics of space. Firstly, by Observation Method, data acquisition at the MECRI was conducted on a day from 10:00 a.m. to 5:00 p.m. Secondly, by using the Space Syntax Method to acquire numerical and graphical data: Integration, Mean Depth, Isovist Area, Isovist Perimeter Analyses were made. These analyses and obtained values were then input into a two-dimensional plan of the space in the University of Michigan licenses “Syntax 2D” program and floor plans with the Space Syntax Method. Data and Space Syntax data were overlapped with the “SPSS (Statistical Package for the Social Sciences)” statistical analysis program. During the first stage, correlation values were formed by overlapping elderly usage frequency data acquired with Observation Method and syntactic values acquired with the Space Syntax Method were then interpreted within the context of the hypothesis. In order for these studies to me conducted initially effective parameters on spatial behavior and perception of the elderly users in an elderly care institution were set forth and associated (Figure 5). Next, the necessary data within the context of the parameters and methods used to acquire data were determined. While specifying the method of study, concepts such as the elderly users, spatial configuration, spatial behaviors and perceptions were approached conceptually within the scope of the Space

Figure 5. Parameters of elderly user’s spatial behavior and perception in the elderly care institution.

Behavioral responses of the elderly regarding spatial configuration: An elderly care institution case study


Syntax. By observing how elderly user perceived the environment and how they were affected by it, their spatial behavior was evaluated together with their adaptive or maladaptive behavioral responses. The study conducted within this context is as follows: • The method of periodic observation was used to present the frequency and duration of interaction between elderly users and spatial configuration. • Analysis was performed regarding: integration, mean depth, isovist area and isovist perimeter utilizing the Space Syntax Method. • Data were overlapped and read using the SPSS to determine if the designs of elderly care institutions act in coincidence with elderly users as well as the behavioral responses which the designs created on the user. 3.2. The data from observation In the MECRI, interaction with spatial configuration and elderly users was conducted with the Observation Method. For elderly users to be observed in the institution, identification of the spaces where social interaction was most intense was necessary and such observations were conducted on the ground and first floors of the preferred spaces. It was observed that elderly users preferred areas such as corridors and block entry areas which involved activity rather than places which were originally planned to be the social interaction areas. All observed points are described as follows: • MA1 (A Block Entry Area), • MA2 (A Block Ground Floor Corridor Area), • MA3 (A Block First Floor Corridor Area), • MA4 (A Block Daily Resting Room), • M1 (Cafeteria), • M2 (Multi-Purpose Hall), • MC1 (C Block Entry Area), • MC2 (C Block Ground Floor Corridor Area), • MC3 (C Block First Floor Corridor Area),

• MC4 (C -Social Block Connection Area). The areas allocated in the architectural planning as social interaction areas and also the spaces preferred by elderly users which were transformed to social interaction areas over time as well as the paths they followed while reaching these spaces are marked below (Figure 6-7). In addition the behavioral patterns and the number of elderly users that were affected by the use of these spaces as well as their frequency have been shown below (Table1). Utilizing the Observation Method, spaces and how often elderly users in the institutions used them along with behavioral moods and patterns which were periodlically performed in this institution as well as social areas with high/low interaction levels are indicated in the below figures (Figures 8-9). The light colored areas are inter-grated, highly social interaction and socio-pedal areas, while the dark areas indicate depth and low social interac-

Figure 6. Ground floor observation points in MECRI.

Figure 7. First floor observation points in the MECRI.

ITU A|Z • Vol 12 No 3 • November 2015 • E. Özsüt Akan, A. Ünlü


Table 1. The usage and the frequency of the spaces in the MECRI.





  PERIOD (10:0013:00)   PERIOD (14:00-17:00)

M A1 (A Block Entry Field) M A2 (A Block Corridor Area On The Ground Floor)


Waiting, Looking Around; Drinking Tea or Coffee etc.

Watching TV






Reading Standing Newspaper Chat or Book etc.


USAGE FREQUENCY (Number of Total Users)

Sitting Chat

Playing Card Game













M A3 (A Block Corridor Area On The First Floor)










MA 4 (A Blok Day Resting Room)










M C1 (C Block Entry Field)


























M C2 (C Block Corridor Area On The Ground) M C3 (C Block Corridor Area On The First Floor)



M C4 (C Social Block Connection Area)










M 1 (Cafeteria)










M 2 (MultiPurpose Hall)










Behavioral responses of the elderly regarding spatial configuration: An elderly care institution case study


Figure 8. Ground floor social interaction character analysis in MECRI. Figure 10. The model of space syntax association with the elderly user’s spatial behavior and perception in the elderly care institutions.

Figure 9. First floor social interaction character analysis in the MECRI.

tion and socio-fugal areas. To what extent spatial configuration of the instiution was, to what effect on the intergration level of the elderly was made and to what extent as well as the behavioral responses of the elderly were determined. In the model below, set within the context of Person-Environment-Behavior, A- Environment Figures is quantitative figure for the MECRI, set forth by the Space Syntax Method. However, B- the Behavior in the Environment is qualitative and was acquired in the case study (Figure 10). 3.3. Space syntax data The functional structure of the observed elderly institution was defined mathematically by identification of space and relations among them as a network. The relation between the acquired numeric data and social structure was defined objectively and determined using the Space Syntax Method. For acquiring numerical and the graphical data four analyses were made; Integration, Mean Depth, Isovist Area, and Isovist Perimeter Analysis. Values were obtained by accessing the University of Michigan licensed ‘’Syntax 2D’’program in which separate analyses for the institution, the interaction between the spatial configuration and the elderly user behavior were questioned.

Proceeding from this, how much the spatial configurations of the institution conformed to elderly behaviors and their behavioral responses were discussed. The spatial configuration of the floor plan was questioned with regards to the design of the elderly institution. These values were converted to numerical data by the Space Syntax Method combined with the Observation Method, then overlapped by the SSPS statistical program. Therefore analysis regarding the design and the elderly user’s spatial behaviors in the elderly institution, as well as adaptive or maladaptive responses that elderly users formed against the spatial configuration was set forth. There is a possibility to compare different forms on the same quantitative basis with the Space Syntax analyses (Kim, 1999; Penn, 2003). The concepts being set forth by this method help the space, together with the physical and lexical parameters in the elderly care institutions to be understood. Space Syntax data analysis was calculated separately by inserting Ground Floor and First Floor plans into the Syntax 2D program with Isovist Area, Isovist Perimeter, Integration, and Mean Depth. Based on the Syntax 2D and isovist area, by defining the border of the plans to be analyzed and the walls inside it, the area to be analyzed was determined and the relations within the scope of this area were searched (Figure 11). The program working with a grid system, after determining a grid towards this influence area, was connected with identical field of view and the grid separations, calculating the physical space relations within the borders selected for analysis (Şalgamcıoğlu,

ITU A|Z • Vol 12 No 3 • November 2015 • E. Özsüt Akan, A. Ünlü


2013). The Isovist Area is; the value that gives the surface area of the isovist area. The perimeter values of the isovist area, differ from the Isovist Area, and do not provide the surface value of 360 degrees of the Isovist area on separate points but instead the calculation of the perimeter value of a two-dimensional polygon. This situation demonstrates about whether the sizes of the space perceived from each point are thin, long or not. When being discussed on the basis of the institution, it also reveals the differences between convex spaces having close sizes and convex space geometrics have different size values. The Perimeter of the Isovist Area and the size of the perceived space from the chosen point is show to be long or short. The Integration value comes first among the syntactic values, having the property of giving information. It enables interpretation of the information on which spaces are deep or shallow in overall general relation with the integration analysis. It is interpreted, that the mean depth value is contradictive of the integration value. In those spaces, this value is related with an accumulation amount in stage wise engagement within each other as shown in the Space Syntax Analysis

Figure 11. Visual field area from observation points with the garden.

Figure 12. Syntactic analyses of isovist area, isovist perimeter integration and mean depth with the garden.

in Figure 12. Where the garden was included, the Isovist Area of the front garden, Isovist Perimeter and Integration value came out to be highest. In this connection, the A block entrance faces the front garden, and although MA1 is a narrow space, it had the highest integration value (Figure 12). The aim of this study was to attempt to find the spatial behavioral responses towards spatial configuration in an elderly institution. For this reason, the syntactic data values were calculated separately for two plans, including the values of sectors differentiated as “with the garden”, “without the garden” to read adaptive and maladaptive behavioral responses on the spatial configuration. The mean syntactic values were obtained for the four concepts listed by dividing the total data value of institution plans on its own grid count (Table2; Table 3). The names applied for each data group for which a value was obtained are listed below; • Integration • Mean Depth • Isovist Area • Isovist Perimeter In the Space Syntax Analysis made; with the isovist fields, the space that users could see a complete 360 degrees, was scanned. A user sitting by the window could easily perceive both the outside and inside and hold the isovist area at the highest level. As a result when observed, these spaces had the characteristic of being the most preferred. When the plan without the garden was taken into account, the isovist area graphs from the observed points are as follows (Figure 13). When we look at configuration of the elderly institution without a garden, as the colors of living units are indicated by deep and dark blue colors, corridors, multipurpose hall and cafeteria are green and yellows color in a complimentary manner (Figure 14). 3.4. Correlations Finally, the statistical relationship of the different phases was also be evaluated, and the addressed comparisons examined using the “SPSS”. This was the comparison of Usage Frequency

Behavioral responses of the elderly regarding spatial configuration: An elderly care institution case study


Table 2. Space syntax values with the garden. SPACE SYNTAX VALUES SPACES





M A1 (A Block Entry Field)





M A2 (A Block Corridor Area On The Ground Floor





M A3 (A Block Corridor Area On The First Floor





MA 4 (A Blok Day Resting Room)





M C1 (C Block Entry Field)





M C2 (C Block Corridor Area On The Ground





M C3 (C Block Corridor Area On The First Foor





M C4 (C Block- Social Block Connection Area)





M 1 (Cafeteria)





M 2 (Multi-Purpose Hall)





Figure 13. Visual field area from observation points without the garden.

Figure 14. Syntactic analyses of isovist area, isovist perimeter, integration and mean depth without the garden.

and Syntactic Values. The case study attempts to ascertain the effect of space configuration on elderly users’ responses. Four relationships were examined in the network in the use of the parameters of the spatial frequency range. • Frequency- Integration • Frequency- Mean Depth • Frequency- Isovist Area • Frequency- Isovist Perimeter When we consider Syntactic Values with a garden and the Usage Frequency correlation; Usage Frequency- Integration (r = 0,730; p= 0,017), Usage Frequency-Isovist Area (r = 0,707; p= 0,022) and Usage Frequency- Isovist Area Perimeter (r = 0,795; p= 0,006) correlations are positively related, Usage Frequency- Mean Depth (r = -0,704; p= 0,023) correlation turned out to be negative (Figure 15). Elderly users preferred to settle in those spaces with a higher integration value, isovist area and perimeter. When we take into account the Syntactic Values and Usage Frequency correlation, without considering the

ITU A|Z • Vol 12 No 3 • November 2015 • E. Özsüt Akan, A. Ünlü


Table 3. Space syntax values without the garden. SPACE SYNTAX VALUES SPACES





M A1 (A Block Entry Field)





M A2 (A Block Corridor Area On The Ground Floor)W





M A3 (A Block Corridor Area On The First Floor)





MA 4 (A Blok Day Resting Room)





M C1 (C Block Entry Field)





M C2 (C Block Corridor Area On The Ground)





M C3 (C Block Corridor Area On The First Floor)





M C4 (C Block- Social Block Connection Area)





M 1 (Cafeteria)





M 2 (Multi-Purpose Hall)





Figure 15. Correlation analyses of usage frequency and syntactic values with the garden plan.

Figure 16. Correlation analyses of usage frequency and syntactic values without the garden plan.

garden; Usage Frequency is not related with an integration value (r = -0,463; p= 0,178). The Usage Frequency and the Isovist Area correlation is (r= -620; p= 0,056) nearly negatively related. Usage Frequency and the Isovist Perimeter correlation is also negatively related (r = -0,740; p= 0,014). However, the Usage Frequency and Mean Depth correlation is not related one another (r = 0,049; p= 0,893) (Figure 16). The garden as seen in the analysis of correlation between the external environment with the use of spatial frequencies ranges close to negative values without any visual interaction however a significant relationship was found in that elderly users did not move along with the spatial construction. 3.5. Results and discussion The methodology of this study was set forth and examined within the concept relation between the spatial configuration in the elderly care institutions with the elderly usersâ&#x20AC;&#x2122; behavioral moods and patterns at perceptional and behavioral levels syntactically.

Behavioral responses of the elderly regarding spatial configuration: An elderly care institution case study


With reference to this context, the case study conducted in MECRI focused on how much spatial configurations affected the spatial behaviors of elderly users, and how elderly people responded towards the configuration, as well as the mutual interaction and conversion between the spatial configuration in the elderly care institution and elderly users spatial behaviors. Within the scope of the case study conducted with the Observation Method, it was determined that elderly users are against spatial configuration instead localizing and interacting within spaces which were not the intended function. It was determined that the mostly used spaces in the institutions were the A Block entrance, connection corridors and the heads of corridors instead of the configured social interaction area. Although there was a weak interaction inside the linear order, elderly users moved social spaces to shallow regions. As Ünlü et.al (2001) stated, the average depth value of the spaces show that the social interaction is weak. The fact that the spaces such as the multi-purpose hall and cafeteria in the general spatial configuration, overlooking the backyard where the social interaction can be formed, by not being able to meet the sensory and affective needs of the elderly people, lowered their usage frequency. When the focusing on the correlation which was obtained with the “SPSS”; the Syntactic values such as Isovist Area, Isovist Perimeter, Integration and Mean Depth values, which have been obtained from the architectural plan and Usage Frequency from Observation in the MECRI were overlapped; the values found with the garden correlation analyses are as follows: r= 0,730; p= 0,017, r= 0,707; p= 0,022, r= 0, 795; p= 0,006, r= - 0,704; p= 0,023.Through these values, it is seen that the Usage Frequency forms a positive correlation with the Isovist Area, Isovist Perimeter and Integration values coming from Space Syntax. Also, there is a negative correlation between Usage Frequency and Mean Depth. These correlations demonstrate that the elderly users preferred shallow and integrated spaces that, included movement, high isovist field and sun-

shine. Isovist Area, Isovist Perimeter, Integration and Mean Depth Syntactic values found without the garden along with the Usage Frequency correlation analyses were as follows: r=-0,620; p=0,056, r=-0,740; p=0,014, r=-0463; p=0,178, r=-0,049; p=0,893. These correlation results demonstrate that the Usage Frequency and Space Syntax values did not have any correlation. Elderly users used social interaction spaces available that were not included in the planned layout of the garden. Rather they localized in the places having high integration and connectivity values as the social integration area in the general configuration. Configurations have adaptive and maladaptive varied effects within the scope of the elderly users’ sensory and affective capabilities on their spatial behavior responses. While the users also show maladaptive behaviors towards the configured space in the MECRI, they behaving improperly against the spatial configuration, perform the expected spatial behavior. The social areas are fictionalized in the architectural design of the socio-pedal areas and it is seen that they became socio-fugal areas as a result of behavioral responses. Simultaneously, areas design as socio-fugal areas were turned into socio-pedal areas using external functions of the elderly users’ response. 4. Conclusion Within the scope of the case study; spatial syntax properties and the effects of these properties on the elderly user’s spatial behaviors and the responses were examined. In the institution being discussed, the presence of the relation between the elderly user’s own behavior and the syntax values of the space and its quality, were set forth by means of the observation method, case study, and acquiring the space syntax values providing the combined analyses. When the correlation between Syntactic Values and Usage Frequency is considered the elderly people begin to behave rather adaptively to space configuration with the garden, whereas they begin to behave rather adversely developing maladaptive behavior to

ITU A|Z • Vol 12 No 3 • November 2015 • E. Özsüt Akan, A. Ünlü


space configuration without the garden. The elderly users reacted quiet uninterestedly in the space configuration without the garden. Additionally, they did not use the presently configured multi- purpose hall and cafeteria. Elderly users in the MECRI did not use those spaces or rarely used and responded behaviorally in an adversely effecting manner, which were built as social inaction area and configured into inactive spaces, or corridors needed to be walked or even those blocks reached though entrance from administration buildings. In conclusion of evaluation regarding this case study, it is put forth that the elderly care institutions in a country where new investments are made, should consider the physical and psycho-social and psycho-spatial features of this specific age group and contribute those features into future architectural designs of elderly care institutions, while simultaneously building a detailed database so as to develop a parameter within the architectural design process as a conceptual framework. Acknowledgments I am very much thankful to my advisor Prof. Dr. Alper Ünlü for his valuable guidance, encouragement at various stages through my dissertation period. References Altman, I. (1975). The environment and social behavior: Privacy, personal space, territory and crowding. Monterey, CA: Brooks/Cole Publishing. Carstens, D. Y. (1993). Site planning and design for elderly. New York, NY: Van Nostrand Reinhold. Edgü, E. (2003). Konut tercihlerinin mekansal dizin ve mekansal davranış parametreleri ile ilişkisi (relation of the house preferences with space syntax and spatial behavior parameters). (Unpublished doctoral dissertation). Istanbul Technical University, Graduate School of Science, Engineering and Technology, Istanbul. Eshelman, P. E., Evans, G. W. (2002). Home again: environmental predictors of place attachment and self-esteem for new retirement community residents.

Journal of Interior Design, 28(1), 3-9. Hillier, B., Hanson, J. (1984). The social logic of space. Cambridge: Cambridge University Press. Kim,Y. O. (1999). Spatial configuration, spatial cognition and spatial behaviour: the role of architectural intelligibility in shaping spatial experience. (Unpublished doctoral dissertation). University of London, UK. Kopec, D. (2012). Environmental psychology for design. New York: Radford University Fairchild Books. Lawton, M. P., Nahemow, L. (1973). Ecology and the aging process. In C. Eisdorfer & M. P. Lawton (Eds.), The psychology of adult development and aging (pp. 619-674). Washington D.C: American Psychological Association Press. Osmond, H. (1959). The relationship between Architect and Psychiatrist. In C. E. Goshen (Ed.), Psychiatric architecture: A review of contemporary developments in the architecture of mental hospitals, schools for the mentally retarded, andrelated facilities. Washington, DC: American Psychiatric Association. Penn, A. (2003). Space syntax and spatial cognition or why the axial line? Environment and Behavior, 35(1), 3065. Seamon, D. (1994). The life of the place. Nordic journal of architectural research, 7(1), 35-48. Sommer, R. (1969). Personal space: The behavioral basis of design. Englewood Cliffs, NJ: Prentice-Hall Inc. Salgamcioglu, M. E. (2003). İstanbul’da çoklu konut gelişiminin semantik ve sentaktik olarak irdelenmesi: 1930-1980 dönemi (Examining the development of urban apartment housing through semantic and syntactic consideration in istanbul: 1930-1980 period). (Unpublished doctoral dissertation). Istanbul Technical University, Graduate School of Science, Engineering and Technology, Istanbul. Trotter, E., Phillips, M., Watson, L. (1998). Remodelling sheltered housing. London: Housing 21. Unlu, A. (1998). Çevresel tasarımda ilk kavramlar. İstanbul: İTÜ Mimarlık Fakültesi Yayınları. Unlu, A., Ozener, O., Ozden, T., Edgu, E. (2001). An evaluation of so-

Behavioral responses of the elderly regarding spatial configuration: An elderly care institution case study


cial interactive spaces in a university building. Proceedings of The 3rd International Symposium on Space Syntax, Georgia Institute of Technology, Atlanta, US. Wojgani H., Hanson J. (2007). Extra

care housing: A paradigm shift. Proceedings of The 6th Space Syntax Symposium, Istanbul, Turkey. Zeisel, J. (2005). Inquiry by design. New York and London: W.W. Nortan & Company.

Yaşlıların davranışsal tepkilerinin mekansal kurgu üzerinden okunması; yaşlılık kurumu örneği Yaşlı birey ile çevresi arasında ki etkileşimde; kullanıcıda yaşlılık kurumunun mimari tasarımına bağlı olarak mekansal kurguya karşı uyumlu ya da uyumsuz davranışsal tepkiler oluşmaktadır. Makale; yaşlı kullanıcıların mekansal davranışları yoluyla yaşlılık kurumunun mekansal kurgusuna karşı geliştirdikleri tepkilerin varlığını ve niceliğini ortaya koymaya yönelik çalışmaları ve sonuçlarını içermektedir. Bu makalede sunulan çalışmalar ile yaşlılık kurumu tasarımı ve yaşlı kullanıcı davranışsal tepkileri arasındaki ilişki “Çevre Davranış Çalışmaları” ana çerçevesi içinde “Mekansal Dizim Yöntemi” ile ele alınmaktadır. Alan çalışması ile mekansal kurgunun, yaşlı kullanıcı davranışları üzerindeki etkileri incelenerek farklılaşan mekansal davranış tepkileri saptanmış ve mekanların dizimsel değerleri çıkartılmıştır. Lineer mekan kurgusuna sahip Maltepe Yaşlı Bakım ve Rehabilitasyon Kurumunda yapılan alan çalışması sırasında Gözlem Tekniği ile kullanıcıların hangi mekanları hangi sıklıkta kullandıkları periyodlar halinde gözlemlenerek yoğun ve düşük etkileşimli sosyal alanlar belirlenmiştir. Space Syntax Yöntemi kullanılarak ta mekanların Eşgörüş Alanı, Eşgörüş Alanı Çevresi, Bütünleşme ve Derinlik değerleri elde edilmiştir. Bu bağlamdan hareketle; yaşlılık kurumunun tasarım performansı, mekansal dizim değerleri ile yaşlı kullanıcıların davranışsal tepkileri çakıştırılarak tartışılmaktadır. Sonuçta; yaşlı kullanıcıların mimari tasarımdan gelen mekansal kurguya karşı davranışsal tepkiler geliştirdikleri ortaya konulmaktadır. Yaşlılık kurumu tasarımı yaşlı kullanıcıların duyuşsal ve davranışsal ihtiyaçlarına cevap verip vermemesine göre uyumlu ya da uyumsuz davranışsal tepkilere sebep olmaktadır. Lineer kurguya

sahip yaşlılık kurumunda davranışsal tepki artarak uyum azalmakta, yüksek etkileşim (socio-pedal) düzeyi düşmekte, bireysel düşük etkileşimli (socio-fugal) ilişki düzeyi artmaktadır. Maltepe Yaşlı Bakım ve Rehabilitasyon Kurumunda yaşlıların mekansal kurgu üzerinde sosyal etkileşim alanları olarak tasarlanmış Çok Amaçlı Salon ve Kafeterya gibi alanları kullanmadığı gözlemlenmiştir. Tasarımın sosyal mekan çözümünde bu yaş kuşağına yönelik olarak yeterli olmadığı sosyal alanların az ya da hiç kullanılmadığı görülmektedir. Bu durumun başlıca nedenleri; bu mekanların hareketi görmeyen, arka bahçeye bakan, uzun bir yol yürüyerek ulaşabilecekleri ya da idari bloktan geçilerek gidilebilen alanlarda kurgulanmış olmalarıdır. Yaşlılar hareketi gören, kolay ulaşabildikleri, idari bölümle ilişkisiz olan A Blok girişi, bağlantı holleri, koridor başları gibi sirkülasyon alanlarını fonksiyonları dışı değiştirip dönüştürerek sosyal etkileşim alanları gibi kullanmaktadırlar. Bu yolla yaşlı kullanıcılar duyuşsal ve davranışsal ihtiyaçlarına cevap verip vermemesine göre mekanın fonksiyonunu değiştirmekte, kullanarak ya da kullanmayarak tepki oluşturmaktadırlar. Mimari tasarımda sosyal etkileşim alanları olarak kurgulanan mekanlara, davranışsal tepkileri sonucu gitmeyerek düşük sosyal etkileşimli mekanlar haline dönüştürebildikleri gibi bunun tersini de yapabilmektedirler. Bu noktada yaşlı kullanıcı davranışsal tepkisi ile mekan kurgusu arasında nasıl bir ilişki kurulabileceği sorusu önem kazanmaktadır. Çalışma kapsamında mekansal kurgunun kullanıcı davranışına uyumu ya da uyumsuzluğu Gözlem Tekniği ve Mekansal Dizim Yöntemi arasındaki korelasyonlar ile ortaya konulmaktadır. Mekansal Dizim Yöntemi ile Bütünleşme, Derinlik, Eş Görüş Alanı ve Eş Görüş Alanı Çevresi Analizleri yoluyla mekansal kurgu üzerinde belirlenen noktaların dizimsel

ITU A|Z • Vol 12 No 3 • November 2015 • E. Özsüt Akan, A. Ünlü


değerleri ayrı ayrı çıkartılmıştır. Bahçe dahil edilerek ve edilmeyerek yapılan iki farklı dizimsel analiz sonuçlarına göre yaşlı kullanıcının mekan kullanımında dış mekanla kurduğu görsel ilişki önemli bir parametre olarak öne çıkmaktadır. Kullanıcı görsel etkenlere bağlı olarak mevcut sosyal alanları kullanmak yerine kendi sosyal alanlarını kurgudan bağımsız, dış ortamı gören blok girişlerine, koridor başlarına ve bağlantı koridorlarına taşımaktadır. Mekansal Dizim Yöntemi ile elde edilen dizimsel değerler ve gözlemlerden çıkartılan kullanım frekansı arasındaki korelasyonlar ile davranışsal tepkinin varlığı ve niteliği yorumlanmaktadır. Bahçeli ve bahçesiz korelasyon değerleri göstermektedir ki yaşlı kullanıcı lineer kurguya göre hareket etmeyerek davranışsal tepki vermektedir. Kurumun mevcut kurgu düzeni; uzun yol yürümeyi gerektirerek sosyal alanlara ulaşımı zorlaştırdığından ve hareketi görme ihtiyacına yönelik dış

ortamla görsel ilişkiyi azalttığından dolayı yaşlı kullanıcının duyuşsal ve davranışsal ihtiyaçlarına cevap verememektedir. Kullanıcı dış ortamla görsel ilişki kurduğu bahçeli plan kurgusunda; Eş Görüş Alanı, Eş Görüş Alanı Çevresi ve Bütünleşme değerleri yüksek olan yerleri fonksiyonu dışı değiştirip dönüştürerek sosyal etkileşim alanları olarak kullanmayı tercih etmektedir. Çalışmayla birlikte örnek yaşlılık kurumunun tasarımının yaşlı kullanıcıların mekansal davranışlarıyla ne kadar uyumlu oldukları irdelenerek davranışsal tepkileri ortaya konulmaya çalışılmaktadır. Bu suretle gelecekteki yaşlılık kurumları tasarımlarında bu yaş grubunun duyuşsal ve davranışsal ihtiyaçlarının da dikkate alınarak mimari program kapsamına bir parametre olarak sokulmasının gerekliliği ileri sürülerek kavramsal bir çerçeve oluşturulmaktadır.

Behavioral responses of the elderly regarding spatial configuration: An elderly care institution case study

ITU A|Z • Vol 12 No 3 • November 2015 • 105-113

The influence of architectural configuration on the pedestrian network in Büyük Beşiktaş market

Ervin GARİP1, Mehmet Emin ŞALGAMCIOĞLU2, Fitnat CİMŞİT KOŞ3 ervingarip@gmail.com • Department of Interior Design and Environmental Design, Istanbul Kultur University, Istanbul, Turkey 2 salgamcioglu@gmail.com • Department of Architecture, Faculty of Architecture, Istanbul Technical University, Istanbul, Turkey 3 fitnatcimsit@gmail.com • Department of Architecture, Faculty of Architecture, Beykent University, Istanbul, Turkey


Received: May 2015

Final Acceptance: October 2015

Abstract A study of trading can help us understand a wider range of organizations and building types. Many shopping centers have been built in Istanbul over the past twenty years. Although these privately owned places serve as social spaces and provide an area for many public activities, their size and close proximity separate them from the urban environment and choke off daily urban street life. Buyuk Besiktas Market, which is selected as the subject of the case study has a specific architectural form. Multiple entrances and a conductive interface converts the building and especially the ground floor into a common area and urban domain that hosts concerts, urban activities and special meetings. The study seeks to understand how does the architectural form and syntactic pattern of the outer layer play a role in changing the pedestrian network inside the building and how do basic architectural elements such as inner courtyards and open spaces affect pedestrian flow and preferences? The research procedure is based on two main steps. The first step comprises a gate count of the people passing through the gates and a density analysis. The second step is to analyze the architectural configuration using Syntax 2D program developed by scientists at the University of Michigan. The results of the study support the idea that particularly for multi-entrance buildings, the urban environment can be more dominant or at least effective in manipulating the natural movement in buildings. Independent from the structure of the building, the configuration which was set by urban dynamics is so dominant. Keywords Architectural layout, Pedestrian movement, Syntactical configuration.


1. Introduction Many shopping centers have been built in Istanbul over the past twenty years. Although these privately owned places serve as social spaces and provide an area for many public activities, their size and close proximity separate them from the urban environment and choke off daily urban street life. Yet, from the perspective of commercial land use patterns, shopping spaces should be able to integrate urban spaces in a continuous way. In this manner, this study examines the syntactic character of the market places that affect people’s movement, particularly the interface between architectural layout and urban form that affects that movement. Istanbul’s Buyuk Besiktas Market is selected as the subject of the case study. It has a specific architectural form that can be characterized as a semi-open, interior courtyard shopping building with multiple entry points that is located in a dense shopping district in Istanbul. The configuration of the building will be structurally analyzed considering the architectural layout, pedestrian flow and integration. The study seeks to answer the questions below: • How does the architectural form and syntactic pattern of the outer layer play a role in changing the pedestrian network inside the building? • How do basic architectural elements such as inner courtyards, open spaces and store allocation affect pedestrian flow and preferences? 2. Retail pattern and pedestrian movement A study of trading can help us understand a wider range of organizations and building types. Trading has a pervasive effect on urban form and land use patterning as well as building interiors and appears in one form or another in every society and in every period of history (Penn, 2005). As Hillier (2005) noted, buildings and cities exist for us in two ways: as the physical forms that we build and see and as the spaces that we use and move through. In a situation where movement, configuration, and attraction are all in

agreement, logic strongly suggests configuration as the primary ‘cause’ of movement. Logically, the presence of attractors can affect the presence of people; however, these attractors cannot affect the fixed configurational parameters that describe the spatial location. Similarly, configuration may affect movement; however, configurational parameters cannot be affected by movement; see Figure 1, (Hillier et al., 1993). Differences in layout affect movement independently from the attractors. As Hillier (1993) illustrated in Figure 2, attractors and movement may affect each other; however, the other two relations are asymmetrical. Configuration may affect the location of attractors, but the location of attractors cannot affect configuration. Likewise, configuration may affect movement, but movement cannot affect configuration. If strong correlations are observed between movement and both configuration and attractors, the only possible lines of influence are from configuration to both movement and attractors, with the latter two influencing each other. We can better understand how cities work if we draw a distinction between movement ‘to’ or ‘from’ spaces and movement ‘through’ spaces. Movement ‘to’ or ‘from’ spaces is primarily

Figure 1. (a) The more central segments of the ‘main street’ are likely to be the most frequently used. (b) The two most central vertical elements, one above and one below the ‘main street,’ would be on shorter routes than the more peripheral vertical elements (Hillier, et. all., 1993).

Figure 2. Attraction, configuration and movement (Hillier, et. all., 1993).

ITU A|Z • Vol 12 No 3 • November 2015 • E. Garip, M. E. Şalgamcıoğlu, F. Cimşit Koş


a function of land use, whereas movement ‘through’ spaces is primarily a function of configuration. More importantly, urban configuration creates an interface between those two types of movement. Thus, these two types of movement can be evenly balanced in some spaces and unevenly emphasized in others. Liveness, however, appears to require that both components be present and mutually supportive (Peponis, Ross, Rashidi, 1997). The actual patterns of agglomeration and differentiation of retail functions that we observe in urban property use patterns appear to be strongly related to both the geometry and network topology of the urban street system. Two theories have sought to account for this phenomenon from a space syntax perspective. The first theory is the theory of natural movement (Hillier et al., 1993), which proposes that the configuration of the street grid accounts for a substantial proportion of pedestrian movement in urban areas. Retail land use is demonstrated to affect movement patterns by acting as a multiplier, transforming a linear relationship between spatial integration and pedestrian flows in mono-functional residential areas into an exponential relationship in mixed-use areas. The thesis is that the primary factor is urban spatial configuration, which then causes a pattern of space use that makes certain locations more attractive than others for retail. Retail occupies these locations preferentially and then becomes an attractor of new trips in its own right. The result is a multiplier in which configurationally strategic through routes become dominant retail aggregations. The result is an emergent correlation between land use, pedestrian movement and configuration that demonstrates immense stability over time. The second theory is the theory of the movement economy (Hillier & Penn, 1992; Hillier, 1996; 1997), which proposes that as a by-product of every trip between an origin and a destination, one passes opportunities for interaction and transaction in spaces along the way. We propose that this phenomenon allows for multi-purpose trips and is the link between urban spatial configuration and movement flows

that provides logic for the disposition of land uses. An additional phenomenon exists, however, which is recognizable in many different city forms and cultures. This phenomenon involves how land use patterns remain roughly similar as one travels along a street but change radically as one turns a corner. The traditional ‘gravity’ model employed by shopping mall developers attempts to create an artificial ‘flow’ of pedestrian movement between two known attractors. Shopping malls generally work on the premise of the classic dumb-bell concept; the large competing ‘anchor’ stores at two ends working as ‘magnets’ spaced between a two sided mall of smaller multi-cellular units (Fong, 2003) . Although Hillier (1993, 1996) has argued that movement is determined mainly by the configuration of space, a case study that was done on seven super regional shopping centres (Fong, 2003) shows that variables of attraction could best predict movement distribution rather than variables of configuration. In cases where the functional attractors like big stores do not exist, the interface of the building must be examined by considering that some environmental relational or urban links may serve as urban attractors that can affect pedestrian flow. 3. Conductive interface Building surfaces play a significant role in the relationship between buildings and the urban environment. Especially at the ground floor level, building surfaces act as a “membrane” that serves as an interface between the building and the urban environment. As more of these surfaces enable the transition, the buildings become urban interiors and their gates become the nodes of the urban environment. The Buyuk Besiktas Market (BBM), which is examined in this study, is a three-story building with 184 stores selling accessories, shoes and clothing, in addition to other facilities like a post office and pay offices. Multiple entrances and a conductive interface converts the building and especially the ground floor into a common area and urban domain that hosts concerts, urban activities and special meetings

The influence of architectural configuration on the pedestrian network in Büyük Beşiktaş market


(Figure 3). This provides us with an opportunity to study the architectural characteristics affecting the pedestrian flow and preferences. The surface of the BBM can be described as a conductive interface with multiple entrances that connect the building to its urban environment. The entrances demonstrate several common attributes that can be categorized and gives us the opportunity to discuss how the pedestrian flow is affected by architectural characteristics and the urban interface. Although the entrances that connect the building to the urban grid are defined as “street based,” the entrances located on the periphery and that activate the surface of the building act as a “periphery.” Further, the connections with other shopping areas can be defined as “transition based” (Fig. 4). This predefinition of gates allows us to compare and discuss the role of the interface of the building in pedestrian movement and preference. 4. Method BBM is selected for the case study. The research procedure is based on two main steps. The first step comprises a gate count of the people passing through the gates and a density analysis, which provides the distribution of people inside the building. To understand the tendency of movement, people leaving or entering the market are calculated separately. These observations provide information about the preferences and distribution of pedestrians according to their choices of roles. In other words, this analysis will provide clues about how the conductive interface of the buildings affects public movement and how the existence of the inner courtyard manipu-

Figure 3. Image showing the open courtyard and ground floor level as urban use.

Figure 4. Representation of each gate.

lates that movement. The second step is to analyze the architectural configuration. This will help us understand the syntactic pattern of the building and its basic architectural characteristics such as galleries, corridors and shop layout. Furthermore, analyzing the configuration will provide significant data about integration, which is known to have an effect on natural movement. The space syntax method, a key theory used to define the structural environment, will provide significant data in terms of analysis.

Figure 5. Locations of entrances (A) and convex spaces (B). ITU A|Z • Vol 12 No 3 • November 2015 • E. Garip, M. E. Şalgamcıoğlu, F. Cimşit Koş


The Syntax 2D program developed by scientists at the University of Michigan provides data about the level of integration between convex spaces considering their spatial properties. 4.1. Analog method • Gate count: Recording the number of people passing through the gates (frequency) for 10 minutes synchronically on weekdays and weekends during the morning, midday, and afternoon. • Density: Mapping the distribution of pedestrians, considering the cells defining the convex space. 4.2. Syntactic analysis There are two basic architectural components characterizing the market: the gate composition that gives

Figure 6. Image of the integration analysis, completed using Syntax 2D to find the syntactic scores for gate count nodes and convex spaces.

Figure 7. Image of the mean depth analysis.

the building multiple entrances and the open inner garden that makes the configuration unique. Integration is evaluated by considering the inner garden as a connective space regulating the pedestrian movement around it. We therefore seek to understand whether the open garden serves as an edge that distributes movement or as a perceptive continuity that attracts pedestrians. Syntax 2D creates a grid fragmentation that enabled us to compare the plan integration and depth comparison values through different convex spaces. The mean integration (Figure 6) and mean depth data (Figure 7) are generated as a result of the analyses. These are two main concepts addressed in space syntax theory that can help us understand how some convex spaces are more integrated and shallow than others. For gates, the calculation uses the arithmetic average of entrance lines, whereas the calculation of convex space uses the arithmetic average of homogeneously divided equal areas. 5. Syntactic comparisons and analysis Matching the syntactic data (integration and depth value) with the gate count will help us to explore the influence of architectural configuration on the people passing through the gates. Additionally, the comparison of the syntactic data and people located in convex spaces helps us to explore how integrated or segregated locations influence pedestrian movement, and whether there is any correlation between the architectural configuration of the market and the distribution of movement. A schematic shown in Table 1 and Table 2 examines the relationship between the syntactic values (integration and mean depth) of the market gates and the frequency of people passing through the gates during weekdays and weekends, counted at regular hourly intervals. As previously defined in the space syntax literature, convex spaces have distinctive properties that characterize each space as unique and common (Hillier et al., 1987). The comparison shown below in Table 3 and Table 4 indicates the relationship between the

The influence of architectural configuration on the pedestrian network in Büyük Beşiktaş market


Table 1. Syntactic values and the number of people passing through each gate on weekdays.

Table 2. Syntactic values and the number of people passing through each gate on weekends.

Table 3. Syntactic values and frequencies of people in each convex space on weekdays.

Table 4. Syntactic values and frequencies of people in each convex space on weekends.

syntactic value of predefined convex spaces and the frequency of people passing or spending time in the following convex space. For consistency, the analysis is made over the same time intervals used for the gate count. The statistical analysis was performed using SPSS software considering the gate count numbers (people passing through the gates) and convex number (people located in convex space) as the dependent variables

and syntactic values (integration and depth) as the dependent variables. All regression analyses are shown in Figure 8. First, the regression analysis of the gate count and syntactic value using both the integration and mean depth was performed for each time interval and day shown in the figure. Interestingly, no significant relationship was found between the syntactic values of the ground floor and the people

ITU A|Z • Vol 12 No 3 • November 2015 • E. Garip, M. E. Şalgamcıoğlu, F. Cimşit Koş


passing through the gates. The same result was investigated both for weekdays and weekends. For instance at the 15.15 time interval, the regression between the integration and gate count yields no significant result (Weekday: R=0.145; p=0.732>0.05, Weekend: R=0.220; p=0.6>0.05); nevertheless, for the same time interval, no significant relationship was found between the mean depth and gate count (Weekday: R=0.145; p=0.732>0.05, Weekend: R=0.220; p=0.6>0.05). Almost parallel results were seen for every time interval, as shown in Figure 8. Second, a regression analysis was performed to examine the relationship between the syntactic values and convex space frequency. At the 17.00 time interval, no positive relationship was found between the integration value and convex space frequency (Weekday: R=0.507; p=0.680>0.05, Weekend: R=0.518; p=0.125>0.05). Although there was no significant relationship between the syntactic value and convex

space frequency (Figure 10), the values were higher than the results of the gate count analyses. The negative results for the comparison of both the gate count – syntactic value and convex space frequency – and the syntactic value invite a different perspective for the research. These findings encourage us to discuss the architectural layout of the market in terms of the urban context and immediate surroundings. 6. Discussion and conclusions Considering the trends in the number of people entering and leaving the market, every gate shows different patterns. For example, more people tend to enter than leave through Gates 3 and 6, whereas more people leave through Gates 4 and 5. At Gates 1, 2, 6 and 8, the number of people entering and leaving is almost equal. Referring to Hillier’s “through-to movement” theory and considering the data gathered from BBM, it is clear that street-based movement character-

Figure 8. Regression comparison results a. syntactic values-convex space (above), b. syntactic values-gate count (below). WD weekday; WE weekend. The influence of architectural configuration on the pedestrian network in Büyük Beşiktaş market


izes both through and to movement, where pedestrians move with intention and motivation (Gate 3, Gate 6). Transition-based movement characterizes “through movement,” when people act with intention (Gates 4 and 5). On the other hand, peripheral-based movement is identified as “to movement,” when people behave self oriented. By taking into account the people entering and leaving the marketplace, it is clear that every gate has a different character. It was surprising to see clearly that there was no relationship between the architectural layout configuration and the distribution of people passing through the gates. Although the most significant characteristic of the market place is its multiple entrances, the importance of gate location stems not from local integration but from the urban context, which can be described as urban gravity. Moreover, a case study done by Zhang et al. (2012), compares the multilevel and single level shopping buildings to predict the accesibility and pedestrian flow. Related study suggests that multi‐level commercial cases, measures of configuration can not explain the pedestrian flows well. Being a multilevel building, might be seen as another reason of uncorrelation between pedestrian flow and syntactic values of ground floor of Buyuk Besiktas Market. Independent from the structure of the building, the configuration which was set by urban dynamics is so dominant. The gates have different frequencies due to these dynamics. Whether the building have an introvert occupancy its gates creates a new syntactical discussion on urban interfaces. The peripheral, transitional and street based flows are the main attractors in this discussion. The differentiations of these gates are independent from inner syntactical configuration however defines the movement patterns through urban interfaces. The results of the study support the idea that particularly for multi-entrance buildings, the urban environment can be more dominant or at least effective in manipulating the natural

movement in buildings. In particular, Istanbul’s Büyük Beşiktaş Market has a very distinctive architectural layout that has the potential to connect the interior to its urban surroundings at almost every surface of the building. This design decision connects the building to its environment and makes the building indispensable to its environment. References Fong, P. (2003). What makes big dumb bells a mega shopping mall? Proceedings of The 4th International Space Syntax Symposium, London, UK. Hillier, B., Penn, A. (1992). Dense civilizations: The shape of cities in the 21st century. Applied energy 43(1), 4166. Hillier, B., Penn, A., Hanson, J., Grajewski, T., Xu, J. (1993). Natural movement: Or, configuration and attraction in urban pedestrian movement. Environment and planning b 20(1): 29-66. Hillier, B. (1997). Cities as movement economies. In P. Droege (Ed.), Intelligent environments: spatial aspects of the information revolution (pp. 295-342). Elsevier. Hillier, B. (2005). The art of place and the science of space. World Architecture, 185, 96-102. Penn, A. (2005). The complexity of the elementary interface: shopping space. In The Proceedings of 5th International Space Syntax Symposium, Delft, Netherlands. Peponis, J., Wineman, J. (2002). Spatial structure of environment and behavior. In R. B. Bechtel & A. Churchman (Eds.), Handbook of environmental psychology (pp. 271-291). New York, NY: John Wiley. Peponis, J., Ross, C., Rashid, M. (1997). The structure of urban space, movement and co-presence: The case of atlanta. Geoforum 28(3-4), 341-358. Zhang, L., Zhuang, Y., Dai, X. (2012). A configurational study of pedestrian flows in multi-level commercial space - case study Shanghai. Proceedings of The 8th International Space Syntax Symposium, Santiago, Chile.

ITU A|Z • Vol 12 No 3 • November 2015 • E. Garip, M. E. Şalgamcıoğlu, F. Cimşit Koş


Mimari konfigürasyonun Büyük Beşiktaş Çarşısı’ndaki yaya dolaşımı üzerindeki etkisinin irdelenmesi İstanbul şehri, son 20 yıl içerisinde alışveriş merkezlerinin hızla çoğalması ile farklı bir morfolojik düzene doğru evrilmektedir. Bu yapılar, kendi içlerinde sosyal bir yaşam ve kamusal kullanımlar önerseler de, ölçekleri ve genellikle dışa kapalı özellik sergilemeleri dolayısıyla bulundukları çevreden ve kentsel süreklilikten kopuk özellik göstermektedir. Bu bağlamda yapı çeperleri, mimari kurguları ile beraber önem kazanmaktadır. Sunulan çalışma, mimari kurgu – yapı çeperi – ve kent ilişkisini İstanbul-Beşiktaş’ta bulunan Büyük Beşiktaş Çarşısı üzerinden irdelemektedir. Söz konusu çarşı, yukarıda sözü edilen kapalı alışveriş merkezlerinden farklı olarak, çok girişli, kent içerisine yayılan, yarı açık ve açık galerileri ile özgün bir mimari kimliğe sahiptir. Sahip olduğu dış çeper, çok girişli ve geçirgen yapısıyla giriş katını kentsel bir arayüze dönüştürmekte, iç mekanın dışarıya, dış mekanın da içeriye sızdığı bir kurgusal düzen sergilemektedir. Bu tespitlerden hareketle çalışma kapsamında gerçekleştirilen alan çalışmasının araştırma soruları aşağıdaki gibidir; • Mimari form ve yapı çeperinin dizimsel özellikleri bina içindeki yaya akışı ve dağılımlarını etkiler mi? • İç avlu, açık mekan ve dükkan dizilimi gibi mimari özellikler yayaların rota seçimlerini nasıl etkiler? Büyük Beşiktaş Çarşısı’nın giriş noktaları, hem bina ile kurdukları ilişki hem de kent ile kurdukları ilişki bağlamında özgün özellikler sergilemektedir. Bu farklı karakteristikler, yapının dizimsel değerleri ve gözlemlerden elde edilen özellikler ile karşılaştırıldığında yayaların rota tercihleri, ve mekansal seçimleri hakkında bilgi verebilmektedir. Söz konusu analizler, yapı çeperinin kullanımı (frekans) ve mimari kurgu içindeki rota tercihlerini

ortaya koymaktadır. Araştırma prosedürü, iki aşamadan oluşmaktadır. İlk aşamada bir hafta sonu ve bir hafta içi olmak üzere gün içindeki 5 zaman diliminde ve 10 dakikalık aralıklarda giriş noktalarınındaki insan akışı gözlemlenmiş, bunun yanında aynı zaman aralıklarında mimari kurgu içerisinde tanımlanan konveks mekanlarda kullanıcı dağılımı not edilmiştir. İkinci aşamada mekanın dizimsel özellikleri Syntax 2D programı ile ortaya konmuş, binanın bütünsellik ve derinlik değerleri tanımlanmıştır. Giriş noktalarının dizimsel değerleri ile aynı noktalardaki insan akışının karşılaştırılması, bu akışın mimari kurgu kaynaklı olup olmdığı hakkında bilgi vermekte, iç mekan kurgusunun dizimsel değerleri ile konveks mekanlardaki insan dağılımlarının karşılaştırılması da bina içerisindeki yaya dağılımının sebeplerini ortaya koymaktadır. Verilerin analiz edilmesi sonucunda, giriş noktalarının dizimsel değerleri ile bu noktalardan geçen insan yoğunluğu arasında net bir şeklde anlamlı bir ilişki tespit edilmemiş, aynı şekilde mimari kurgunun dizimsel değerleri ile iç mekandaki insan dağılımı arasında da güçlü bir anlamsal ilişkiye rastlanmamıştır. Bu sonuçlar, bina içindeki insan akışının mimari kurgudan çok, kentsel dinamiklerden kaynaklandığını, yakın çevredeki sokak, açık pazar, meydan gibi katalizörlerin yaya akışını domine edici özellik sergileyebileceğini ortaya koymaktadır. Sonuç yargı olarak, Büyük Beşiktaş Çarşısı, kendi mimari strüktüründen bağımsız olarak kentsel dinamiklerden etkilenmektedir. Çalışma, kapalı ve çevresinden kopuk alışveriş merkezlerinden farklı olarak Beşiktaş Çarşısı’nın Kentsel dinamiklerden beslendiğini, kentin parçası haline geldiğini ve tekil olarak değil çevresi ile beraber bir bütün olarak var olduğunu ortaya koymakta, özellikle kamusal işlevler üstlenen yapılarda kentsel arayüzün önemini göstermektedir.

The influence of architectural configuration on the pedestrian network in Büyük Beşiktaş market

ITU A|Z • Vol 12 No 3 • November 2015 • 115-125

A syntactic analysis of social interfaces in Istanbul Biennial patterns in case of biennial buildings in 2013 Fitnat CİMŞİT KOŞ1, Mehmet Emin ŞALGAMCIOĞLU2, Ervin GARİP3 1 fitnatcimsit@gmail.com • Department of Architecture, Faculty of Engineering and Architecture, Beykent University, Istanbul, Turkey 2 salgamcioglu@gmail.com • Department of Architecture, Faculty of Architecture, Istanbul Technical University, Istanbul, Turkey 3 ervingarip@gmail.com • Department of Interior Design and Environmental Design, Istanbul Kultur University, Istanbul, Turkey Received: May 2015

Final Acceptance: October 2015

Abstract Biennial events in urban life can be discussed in terms of the interrelation of venues as well as art products and dialogues. There are a number of exhibition venues, where preferences are defined by pragmatic or thematic decisions, such as historical buildings, most common public spaces or contemporary popular places. These buildings are territorial markers of specific patterns and act as museum-like environments. This study aims to explore the potential and performative outcomes of these patterns in the Istanbul Biennial between 1995-2013 and aims to discuss the last biennial in 2013 with syntactical parameters and frequencies comparatively for each venue using interface activities and occupancy through the other biennials. For that purpose, this paper will try to answer the questions below: Do biennial space preferences have performative differences in their syntactical configurations through biennial history (between 1995-2013)? Is there a performative relationship between the syntactical values of the interfaces (for the 2013 biennial) and the frequency of each gate of the venues, considering the interface activities and moods? Comparison of the biennial patterns in Istanbul raises many questions in terms of spatial configuration, social network and functional hierarchy in addition to syntactic parameters such as the mean depth, integration or circularity. The territory of the 2013 biennial and its effects on frequencies will also be examined through biennial venues. To understand whether collective memory or accessibility is effective (dominant), audience frequencies are studied within the biennial pattern using gate counts and interface activities. Keywords Social interface, Urban performance, Syntactical configuration.


1. Introduction Biennial events in urban life can be discussed in terms of the interrelation of venues as well as art products and dialogues. There are a number of exhibition venues, where preferences are defined by pragmatic or thematic decisions, such as historical buildings, most common public spaces or contemporary popular places. This study aims to explore the performative outcomes of exhibition space patterns in the Istanbul Biennial between 1995 and 2013 and aims to discuss the last biennial in 2013 with syntactical parameters and frequencies comparatively for each venue using interface activities and occupancy throughout the other biennials. The performance activity of the Biennials should be examined through different patterns with different orientations of biennial venues. As Özpınar (2011) indicated, through the art, knowledge and criticism platform generated outside of the academic space, it is possible to argue the biennial’s public space quality as an operation where ideas, expressions and experiences are produced within a portion of the social life. The biennial is an event, generally organized by independent institutions, that aims to spread across the city with versatile activities but is often only realized in the city center. The difference between the biennial and classical museum exhibitions, international exhibitions or public art projects is not only the fact that the artifacts are not for sale but also that it is a local activity expected to spread across the city and communicate a message. Therefore, the positioning of the biennial within the city, its forms of exhibition, and the urban instruments and mediums it utilizes are extremely important. The choice of location and its past and future indications gain importance as exhibition practice. The interaction between the location and the installed artwork becomes a priority for this choice. The artwork can present itself as a contrast, a criticism, a compliment or an attraction to the location and the meanings it conveys (Özpınar, 2011). The Habermasian idea of the public sphere points to spaces created by the community where ideas, expressions

and experiences are produced, explored, shared, spread and discussed. City streets can facilitate encounters, opportunities, and divergent identities. Because city streets are a place of socialization, the exhibits should have a permissive quality with open access to all sorts of social, human and individual performances. The utilization and presentation of the public space in the biennial is important in terms of reconstruction of the space, the city and the individual (Özpınar, 2011). For that purpose, this paper will try to answer the questions below: Do biennial space preferences have performative differences in their syntactical configurations through biennial history (between 1995 and 2013)? Is there a performative relationship between the syntactical values of the interfaces (for the 2013 biennial) and the frequency for each gate of the biennial venues, considering the interface activities and moods? 2. Territories and social interfaces The space preference for biennial venues is a special urban territory open to public interaction. Venues, boundaries and movement patterns work to create a walking-based context, which creates a performative area in the city. Each biennial has the potential to create a specific micro-environment for urban events and interactions. The biennial buildings are territorial markers during the biennial period. Therefore, this study aims to study both this micro-environment pattern and the building interfaces in order to explore the potential and performances. Territorial space and behaviors are the keys to understanding this interactive pattern. The study includes different levels of space organization. These are the biennial patterns as micro-urban environments and social interfaces of buildings as territorial markers. The hierarchy in these specific micro-environments and their markers can be discussed using Stea’s (1970) theory. The three scales of territory are units, clusters and structures. The scales are interactive and conceptually interrelated (Figure 1). The micro-environments and their potentials in urban performances cre-

ITU A|Z • Vol 12 No 3 • November 2015 • F. Cimşit Koş, M. E. Şalgamcıoğlu, E. Garip


Figure 1. Three territorial spaces (Stea, 1970).

ate spatial practices. Walking based routes and moving based configurations creates a site-specific experience in biennials. As Hillier (2005) mentioned, buildings and cities exist in two ways: as the physical forms that we build and see and as the space that we use and move through. That’s why this study tries to explore the differentiations and similarities of different biennial patterns and analyses urban performances through spatial configurations. Contreras (2006) derived the term ‘The Urban Performance’ from the ‘representational space’ concept, which is a part of the spatial triad defined by Lefebvre (1974) and reinterpreted by lain Borden (Borden, 2001). The assumption is that spatial practices and representations are the things people do and the patterns they physically create for disrupting abstract space. The ‘potential energies’ of groups act to transform and create new social spaces. Thus, the Istanbul biennial has the power to organize these new social spaces. According to Borden (2001), the supposition is that in the urban realm, the idea of an activity in space is the key to understanding the representations and experience of space, and because of these actions or performances, we become true objects in time and space and not simply users or experiencers of but produced by, and products of, the architecture around us. A multi-stranded contemplation of the notion of “knowing a place” includes both the existence and the possibilities

of architecture and the city. Whether for inhabitants or foreign audiences, the biennial pattern is a space for experience, coding or decoding. By movement through exhibition areas, cognitive maps are created by the mind and experienced by the public space itself. Through biennial patterns, movement and actions are simply dependent on markers. Thus, the ‘events’ and ‘activities’ are the bodies that experience the biennial in different ways, routes, frequencies, etc. “If movement is to be one of the generating factors of architecture, it will not take a single form or configuration. There is random movement, as experienced on a flat plane, free of any attraction or constriction. But there is also vectorized movement, which interact with static spaces, often activating them through the motion of bodies that populate them” (Tschumi, 2000). We also interact with the city conceptually as the events occur. In that specific time, with collective artifacts, the city is shared by the people. Thus, as Hillier (2009) indicated, we need a concept of distance which reflects not only the relationship of one part to another but of all parts to all others. In biennial patterns the question is, whether the venue is familiar or not, what are the parameters of performance? Even though people can move randomly, they need to orient themselves with the big picture of the biennial route. Exhibition points within the pattern should be seen as a path to reach a performative event. This is why movement possibilities inside the route have a relationship with the integration of exhibition points. The space syntax theory states that spatial configuration influences the distribution of movement within a network system and that when spaces are more directly connected to other spaces, they are likely to attract more movement (Peponis & Wineman 2002). Our concern with movement patterns over biennial patterns suggests a relationship between the legibility of physical space and the social occupation of that space. Beside the biennial patterns and their social possibilities, the exhibition buildings are also important markers

A syntactic analysis of social interfaces in Istanbul Biennial patterns in case of biennial buildings in 2013


with interfaces they have. When some markers are new, some others have a memory in biennials. The interfaces means interactive layouts both in memory and in physical environment. The audiences have a role to re-produce this spaces in each biennial. Thus, the pattern has a possibility to create new markers and a new spatial practice in the city. To Lefebvre (1974), all social space, at all scales of consideration, is produced. Lefebvre’s theory of space sets forth three principles or modes of production: ‘spatial practice’, ‘representation of space’ and ‘representational space’. (1) Spatial practice is what people - the enactors of social space - do. Though this may seem obvious, it is a considerable conceptual leap for those who assume that space (as a container) precedes activities in space. Spatial practice is ordered, and spaces take on order through (2) representations of space or the plans established by social bodies with the power to create blueprints for the world. Societies are thus said to inhabit (3) representational spaces that contain and are produced by spatial codes that change over time. The representational spaces of everyday life are produced by contemporary spatial codes, fragments of discarded codes, and echoes of revolutionary codes (Protevi, 2006). The biennial buildings are re-produced each time by audiences. The spatial experiences of audiences and how it has shared are the main concern to explore spatial characters of biennial buildings in each pattern. Thus, the second phase of analysis in this study aims to explore this spatial codes in interfaces. 3. Method Through theoretical point of view the method includes different scales of studies both in micro-environments of biennials and building interfaces. The interactivity of these scales will be discussed with an introduction to the comparative analysis of biennial patterns through history. Comparison of biennial patterns in Istanbul raises questions in terms of spatial configuration, social networks and functional hierarchy. The similarities and the differences in the biennial layouts are the basis

of a discussion on the elements and possibilities of the spatial configuration. The anticipated path movement of the pedestrian may differ according to context. Therefore, the relations of typologies, the venues, clusters and interconnections are the concerns of the research. The territory of the 2013 biennial and its effects on frequencies will be examined through the biennial venues. Biennial venues are different types of buildings where people both visit exhibitions and socialize. The pattern of the biennial is a designed integration that is a part of the collective memory. The number of times each building has been occupied throughout biennial history will also be considered. In order to understand whether collective memory or accessibility is effective (dominant), the audience frequencies are studied within the biennial pattern using gate counts and interface activities. The research for the 2013 pattern focuses on parameters such as the comparison of the frequency of audiences at the gates of biennial buildings, collective memory of these buildings throughout biennial history, the modes of the audiences in interfaces, and the syntactic values of the pattern and the gates of the venues. The social interfaces are performative spaces with social, individual and movement-based modes. These interfaces are the spaces where people wait, gather or pass through before exhibition in biennial buildings. At the end of the research, these modes will be examined and the interrelation between the syntactical values and the gate frequencies will be discussed. The following methods are applied for the biennial patterns in this study: 3.1. Analog method • Gate count: Simultaneous recordings of the frequencies at the gates of biennial buildings for 10 minutes both during weekdays and weekends. • Interface frequencies: Simultaneous 2-minute camera shots in each biennial buildings interfaces. • Interface moods: The interface moods are individual (waiting, sitting, etc), social (gathering, talking,

ITU A|Z • Vol 12 No 3 • November 2015 • F. Cimşit Koş, M. E. Şalgamcıoğlu, E. Garip


etc.) and active (movement based) behaviours. in 2 minutes • Collective memory of buildings: How many times has these buildings occupied throughout the history of biennials 3.2. Syntactic analysis • Visual space analysis including integration, circularity and mean depth for each biennial are comparatively studied for each biennial pattern (the active grid numbers for each 100 m2 were the same for all patterns, which are different in scale) • Syntactical analysis of the 2013 biennial pattern, including integration, circularity and mean depth • Focus on building analysis in order to explore the social interfaces which are the spaces where people

Figure 2. The biennial venues.

Figure 3. The relationship between biennial venues and years / 2013 venues.

wait, gather or pass through before exhibition in biennial buildings. 3.3. Statistical analysis • In the statistical analysis conducted with SPSS software, significant relationships are researched through regression analysis. The method includes transforming the patterns and the last biennial building interfaces to syntactic data (with the help of a program named “Syntax 2D” licensed by the University of Michigan) including mean depth, circularity and integration. 4. Istanbul Biennial patterns as museum-Like environments and buildings as territorial markers The Istanbul Biennial is an international cultural network for local and international artists, curators and art critics showing new trends in contemporary art every two years. The Istanbul Biennial is an exhibition model that enables a dialogue between artists and the audience through the work of the artists, exhibitions, panel discussions, conferences and workshops. The Biennial is organized by the İstanbul Foundation for Culture and Arts (IKSV). The first two biennials were established under the general coordination of Beral Madra in 1987 and 1989. After 1989, a curator system was established. The Istanbul Biennial does not have a permanent location. Although certain structures were used more than once, each biennial has had various locations. The biennial venues show differences in historical and contemporary context in Istanbul (Figure 2). The study was done in five biennial contexts defined by context-based walking characters. The biennials that took place in both the Asian and European sides of Istanbul were dismissed because of the need to consider public transportation in these cases. The configurations of biennials differed, whether in the number of venues or in the place references related to for each year (Figure 3). As a consequence, the last biennial will be examined with pattern syntactical parameters and gate frequencies with the interface activities of biennial venues in that pattern. The venues will also be discussed with

A syntactic analysis of social interfaces in Istanbul Biennial patterns in case of biennial buildings in 2013


reference to the collective memory through all biennials. 5. Syntactic analysis and discussion The space syntax helps us analyze the patterns of connection, differentiation and centrality that characterize urban systems and the relationship between the parts and the whole that they engender. The axial map comprises the fewest and longest lines that are necessary to cover all parts of the urban fabric. The number and length of the axial lines is a function of the degree to which other parts of the system are directly accessible and visible from each point. The intersections between axial lines are treated as the elementary relations between spaces. The key property of axial maps is integration. Integration measures the relationship of each line to the network as a whole (Peponis, Ross, Rashidi, 1997). The computation of integration values for the space of a given area is, of course, affected by the location of the area boundary. As Peponis et al. (1997) noted, an integration core that not only links all the parts together but that also relates the center to the periphery seems to encourage the diffusion of movement and the opportunities for exchange and interaction. The area boundaries and accessibility patterns for analysis have been chosen to indicate the possible movement routes between each venue (Figure 4).

Figure 4. Selected biennial pattern layouts.

1st step: Descriptive discussion of biennial patterns from 1995 to 2013 due to urban performance: The locations of choice for the 4th Biennial (1995) titled ‘Orient-ation’ were Antrepo (1), AKM art galleries (2), Aya Irini (3), and the Basilica Cistern (14). The locations situated the biennial in the historical peninsula and the modern front of Istanbul, with the two areas connected by the Galata Bridge (Figure 4). The pattern can be described as a bridging quality with two different loops. The locations of choice for the 6th Biennial (1999) titled ‘The Passion and the Wave’ were Aya Irini (3), the Basilica Cistern (14) and Dolmabahce Palace (7). The locations situated the biennial in the historical peninsula with continuity through the Bosphorus connected by the Galata Bridge (Figure 4). The pattern can be described as a continuous linear quality with one loop. The locations of choice for the 8th Biennial (2003) titled ‘Poetic Justice’ were Antrepo (1), Hagia Sofia (4), the Garanti Platform (9), Tophane-i Amire (12), and the Basilica Cistern (14). The locations situated the biennial in the historical peninsula and the modern front of Istanbul connected by the Galata Bridge (Figure 4). The pattern can be described as a bridging quality with two different loops extended through Istiklal Street, a main pedestrian street with high traffic. This biennial strived to join modern city life and addressed not to only specific target audiences but to everyone with access to the public space (Ozpınar, 2001). The locations of choice for the 9th Biennial (2005) titled ‘Istanbul’ were Antrepo (1), Bilsar (5), Deniz Palas (6), the Garanti Platform (9), the Garanti Bank (10), Garibaldi (11), and the Tobacco Warehouse (13) (Figure 4). This biennial represented a modernist transformation of the city with the chosen locations giving reference to daily life through the use of the most crowded and integrated streets of Beyoglu. The biennial moved away from the historical peninsula. The pattern can be described as having a compact quality with one loop. The locations of choice for the 11th Biennial (2009) titled ‘What Keeps

ITU A|Z • Vol 12 No 3 • November 2015 • F. Cimşit Koş, M. E. Şalgamcıoğlu, E. Garip


Mankind Alive?’ were Antrepo (1), Feriköy Greek School (8) and the Tobacco Warehouse (13) (Figure 4). The venue was limited, with only three locations. The route had more indirect possibilities for passing or landing. The pattern can be described as a continuous linear path with two ends without a loop. The locations of choice for the 13th Biennial (2013) titled ‘Mom, Am I Barbarian?’ were Antrepo (1), the Garanti Platform (SALT Beyoğlu) (9), Arter (15), and Galata Greek School (16) (Figure 4). This biennial also represented the modernist transformation of the city, and the chosen locations gave reference to daily life in the context of the most crowded and integrated streets of Beyoglu with two new venues. The pattern can be described as having a compact quality with one loop through Istiklal Street, a main pedestrian street with high pedestrian traffic. The biennial pattern typologies and their syntactical values can be summarized as follows (Figure 4): 1995, 4 venues, bridging typology with two different loops, 963 mean integration. 1999, 3 venues, continuous linear typology with one loop, 1468 mean integration

2003, 5 venues, bridging typology with two loops extended through a main pedestrian street with high frequency, 1121 mean integration 2005, 7 venues, compact typology with one loop, 1768 mean integration 2009, 3 venues, continuous linear path typology with two ends without a loop, 5080 mean integration 2013, 4 venues, compact typology with one loop extended through a main pedestrian street with high frequency, 1974 mean integration Table 1 summarizes the results of the syntactical data for selected biennial venues between 1995 and 2013. On the basis of this data set, ‘mean integration’ values were produced to investigate the extent to which street connectivity and land use density explain the distribution of movement per street segment with the different typologies indicated above. As mentioned in previous figures, these patterns show differences in the form and character of the paths: The results of this analysis show that pattern configurations have a direct effect on syntactical values. The Golden Horn acts as a boundary in 1995 and 1999, whereas bridging typology has the lowest integration (table 1). The linear configuration without a loop has the highest integration in

Table 1. The syntactical data of biennial venues.

A syntactic analysis of social interfaces in Istanbul Biennial patterns in case of biennial buildings in 2013


Figure 5. 2013 Istanbul biennial pattern.

2009, with the least number of venues. The compact typologies have similar results in 2005 and 2013 with a different number of venues (table 1). Thus, the configuration seems more effective than the number of venues. The interrelation between venues, boundaries, paths and movement has a strong effect on syntactical outcomes. As mentioned above, the configuration may influence the location of attractors, but the location of attractors and their positions cannot influence configuration. In other words, patterns may influence movement, but movement cannot influence the configuration of the patterns. The syntactical values of configurations can orient a through movement in venue that is independent from the distance. The configuration has an effect on movement independent from specific attractors or choices. As Hillier et al. (1993) mentioned, layout differences have effects on movement independent from the attractors. The results of our analysis show that Antrepo, which is one of the most important and most occupied venues through the biennials, has the highest integration value and the lowest pattern integration value in 1995 Table 1). 2nd step: Syntactical analysis of biennial pattern in 2013 due to social interfaces: The configuration of the 13th Biennial (2013) titled ‘Mom, Am I Barbarian?’ has four venues: Antrepo, the Garanti Platform (SALT Beyoğlu), Arter,

and Galata Greek School. This biennial also represented a transformation from previous pattern choices with two new venues, Arter and Galata Greek School, and gave reference to daily life in the context of the most crowded and integrated streets of Beyoglu. This pattern was described above as having a compact quality with one loop (Figure 5). The research for the 2013 pattern focuses on parameters such as the comparison of the frequency of audiences at the gates of biennial venues , the occupancy of the venues through biennial history, the interface activities and modes and the syntactic values of the pattern and the gates (Table 2, Figure 6-7). The outcomes can be summarized as follows: Even though they are both new venues in 2013, Galata Greek School has a higher gate count than Arter, and Galata Greek School has the highest integration value.

Figure 6. 2013 Istanbul biennial buildings and interfaces.

Table 2. The analogue and syntactical data of 2013 biennial buildings.

ITU A|Z • Vol 12 No 3 • November 2015 • F. Cimşit Koş, M. E. Şalgamcıoğlu, E. Garip


Figure 7. 2013 Istanbul biennial building interfaces and syntactical analysis.

Even though Arter is on the main pedestrian street in Istanbul, Galata Greek School has a higher frequency due to its highest integration value in the biennial configuration. As mentioned above, this is because the data reflect the propensity of spaces to be passed through on the way from all origins to all destinations. This shows that new venue preferences for the biennial should be considered with syntactical values of the biennial pattern instead of the urban pattern. Antrepo is the most occupied and the main venue through all biennials. Therefore, it has the highest gate frequency even though it has the lowest integration value. Antrepo and SALT Beyoglu are the most occupied venues during all biennials, with significant differences between weekdays and the weekend. Even though Arter and SALT Beyoglu are on the same street and have similar integration values, the gate frequencies are quite different. This means that attractors have less effect on traffic than the syntactical values of the venue itself. SALT Beyoglu, even as a well-known biennial space, could not affect the frequency of Arter. The interface of Antrepo is the most social and most integrated venue. The integration value of the interface is the highest. Arter is a relatively new venue and has a lower integration value. This affects the interface moods, and the total activity seems lowest despite the fact that the venue is on an active pedestrian street, Istiklal Street. The most active venue is SALT Beyoglu, which has three times the occupancy of other venues in biennial history, and the location is on an active pedestrian street, Istiklal Street.

3rd Step: Statistical analysis: In the statistical analysis conducted with SPSS software, we observed a significant relationship when we used gate count numbers (the number of people passing through a gate) as the dependent variable and when we used the integration values of these gates as the independent variable on weekends. The regression analysis between the integration values of the selected gates of biennial venues and the number of people (frequency) using these buildings (gates) shows a tendency of 80,7% and a mildly significant negative relationship, with R=-0,807. We observed another significant relationship when we used interface moods as the dependent variable and the integration values of these interfaces as the independent variable on weekends. These interfaces are the spaces where people wait, gather or pass through before exhibition in biennial buildings. The interface moods are individual (waiting, sitting, etc), social (gathering, talking, etc.) and active (movement based) behaviors. The regression analysis between the integration values of the selected interfaces of biennial venues and the moods of people using these interfaces shows a significant tendency of 92,7% and a significant positive relationship, with R=0,927. 6. Conclusions The pattern of the biennial is a designed integration of venues that is also a part of the social interface. The biennialâ&#x20AC;&#x2122;s interactive structure provides potential for more effective social outcomes through the design of its location patterns and the use of buildings as territorial markers. The pattern configurations and their qualities have more effect than the metric distance related to certain directions. This is not just a matter of seeing buildings but is also about observing space. The visual distances, collective memories, and integrations on any level orient the movement more than the metric distances. The human relationship with space in cities is generated not only by movement but also by experience and interactions. These museum-like environment patterns define a street con-

A syntactic analysis of social interfaces in Istanbul Biennial patterns in case of biennial buildings in 2013


References Borden, I., Kerr, J., Rendell, J. Pivaro, A. (2001). The Unknown City, Cambridge, Massachusetts: MIT Press. Contreas, Christian, Beros. 2006.

Space, Events and Urban Performance, Proceedings of the 10th Iberoamerican Congress of Digital Graphics (pp. 333336). Santiago de Chile. Hillier, B., Penn, A., Hanson, J., Grajewski, T., Xu, J. (1993). Natural movement: or, configuration and attraction in urban pedestrian movement, Environment and Planning B: Planning and Design (Volume 20, pp. 29-66). Hillier, B. (2005). The art of place and the science of space, World Architecture (11/2005 p. 185). Beijing. Hillier, B. (2009). Studying cities to learn about minds: some possible implications of space syntax for spatial cognition, Environment & Planning B: Planning and Design. Lefebvre, H. (1974, 2004). The Production of Space, United Kingdom: Blackwell Publishing. Tschumi, B. (2000). Event-Cities 2. Boston: MIT Press. Özpınar, C. (2011). Istanbul Biennial in the Contexts of the Individuals, the City and the Public Sphere, Kült-Refereed Cultural Studies Journal (pp. 263283). Istanbul: Istanbul Bilgi University Publications. Peponis, J., Ross, C., Rashid, M. (1997). The Structure of Urban Space, Movement and Co-presence: The Case of Atlanta. Geoforum (Vol.28, No.3-4, pp. 341-358). Peponis, J. & Wineman, J. (2002). Spatial structure of environment and behavior. In R.B.Bechtel & A. Churchman (eds.), Handbook of Environmental Psychology (pp. 271-291). New York, NY: John Wiley. Protevi, John (2006). A Dictionary of Continental Philosophy (pp. 356-357). Yale University Press. Stea, D. (1970). Space, Territory and Human Movements. In Proshansky, H., Ittelson, W., Rivlin, L. (eds.) Environmental Psychology: Man and His Physical Setting. New York: Holt, Rinehart and Winston, Inc.

2013 İstanbul Bienal mekanları üzerinden İstanbul Bienal dokusunun sosyal arayüzlerinin sentaktik analizi Kent yaşamı içinde bienaller, sanat ürünleri ve diyaloglar kadar, bienal mekanlarının ürettiği ilişkiler ile de

tartışılabilir. Tarihi binalar, popüler mekanlar, çağdaş mekan örnekleri gibi pragmatik ve veya tematik kararlar üzerinden tercih edilen bienal mekanları mevcuttur. Bu çalışma 1995 ve 2013 yılları arasındaki bienal mekan

figuration in a specific context based on walking, with buildings as markers. The use of buildings as territorial markers in these museum-like environments create social interfaces. These interfaces have an effect on social relations and gathering activities. Additionally, collective memory is important, and future decisions about biennial venues should consider its influence. The configuration of exhibition buildings is important to an active social network and performative territory. This study has analyzed venue preferences for specific public events such as biennials according to the interaction levels, integration and frequencies of interfaces related to a specific pattern for performative outcomes. The study’s aim has been to to discuss these multi-levels and to show that both scales should be discussed together. In this manner, this study will guide further research about performative pattern configuration choices and social interface outcomes for future biennials. Acknowledgments We would like to thank some of the MSc students at Istanbul Technical University (ITU) from several programmes but mostly from “Architectural Design MSc Programme”, Fulya Menderes, Barış Ateş, Şebnem Çakaloğulları and, Ecem Çalışkan, who have collected the data used in this paper during long hours in the field voluntarily. In addition to our volunteers from ITU, we would like to thank Industrial Product Designer Bedii Engin Koş, who was also a volunteer, for his support and work for data collection in the field.

ITU A|Z • Vol 12 No 3 • November 2015 • F. Cimşit Koş, M. E. Şalgamcıoğlu, E. Garip


örüntülerinin etkileşimli çıktılarını ve sonuç olarak 2013’te gerçekleşen son bienal çerçevesinde karşılaştırmalı olarak mekanların sentaktik parametreleri ve giriş frekanslarını, binaların oluşturduğu sosyal ara yüzleri ve bu ara yüzlerdeki eylem biçimlerini de dikkate alarak tartışmayı amaçlamaktadır. Bu amaçla, bu makale aşağıdaki sorulara cevap aramaktadır; 1995 ve 2013 tarihleri arasında bienal tarihi boyunca mekan tercihlerinin sentaktik konfigürasyonları doğrultusunda etkin farklılıkları var mıdır? 2013 bienali kapsamında mekanların yarattığı sosyal ara yüzler olan ön mekanlarının sentaktik değerleri ve frekansları arasında aktivite ve biçimleri de göz önüne alarak etkin bir ilişki var mıdır? İstanbul Bienal örüntülerinin karşılaştırması, sentaktik parametrelerin yanında sosyal ağ, fonksiyonel hiyerarşi, sosyal ara yüzler gibi bir çok soruyu da tartışmaya dahil etmektedir. 2013 bienalinin etki alanı ve frekanslara etkisi bienal mekanları kapsamında tartışılmıştır. Bienal mekansal örüntüsü aynı zamanda ortak belleğin de bir parçası, tasarlanmış bir konfigürasyon

ve etkileşim ağıdır. Bu nedenle her bir mekanın bienal tarihi boyunca kaç kez kullanıldığı ve bellekteki yeri de çalışma parametreleri arasındadır. Belleğin mi yoksa sentaktik değerlerin mi etkin olduğunun araştırılması için, kullanıcı frekansları hem kapıda yapılan sayımlarla hem de mekanın ürettiği ara yüzdeki aktivite ve biçimleri ile araştırılmıştır. 2013 Bienali kapsamında bu tartışmalar bienal noktalarındaki kullanıcı frekansları, mekanın bienal tarihi boyunca kaç kez kullanıldığı, ara yüz aktiviteleri, hem bienal mekan örüntüsünün hem de mekanların sentaktik değerleri gibi parametrelere odaklanmaktır. Bu sosyal ara yüzler, sosyal ve bireysel eylem biçimleri ve hareket modları ile etkin mekanlar olarak değerlendirilir. Metot kapsamında Michigan Üniversitesi lisanslı Syntax 2D programı ile hem mekan örüntülerini hem bina ara yüzlerini sentaktik değerlere dönüştürerek, SPSS yardımı ile belirleyicilerin istatistik çalışması yapılmıştır. Çalışmanın sonucunda gelecek bienaller için daha etkin örüntü ve mekansal organizasyon öngörülerinin tartışmaya açılması amaçlanmaktadır.

A syntactic analysis of social interfaces in Istanbul Biennial patterns in case of biennial buildings in 2013

ITU A|Z • Vol 12 No 3 • November 2015 • 127-140

Can cognitive maps of children be analysed by space syntax?

Nevşet Gül ÇANAKCIOĞLU gulcanakcioglu@gmail.com • Graduate School of Science Engineering and Technology, Istanbul Technical University, Istanbul, Turkey

Received: May 2015

Final Acceptance: October 2015

Abstract According to environment and behavior theories, individuals in their first years of cognitive development enter a world full of environmental stimuli motivating them to perceive and learn. As a result of many perceptual processes, people convert and transfer perceptual information to their cognitive schemata. The prominent researcher Piaget (1955), who analyzed the perceptual processes of children through cognitive development stages based on age, mainly studied the way that a child perceives the environment within a constructive approach. One of the methods of revealing constructed and stored data in the memory is through the analysis of cognitive maps that children have drawn as they each uniquely perceive the environment and construct specific cognitive schema. Additionally, Lynch (1960) contributes to the theory of perception with the idea that if an urban part has a strong imageable character in terms of paths, edges, districts, nodes and landmarks, one can orient oneself easily when influenced by the synthesis of perceptual processes. Some other scholars have also been investigating how cognitive maps can be analyzed within the theory of space syntax (Zheng & Weimin, 2010; Zimring & Dalton, 2003; Haq & Girotto, 2003; Kim & Penn, 2004). In this sense, this article aims to contribute to both the methodology through the analysis of cognitive maps by using justified permeability graphs within the theory of space syntax and to the understanding of how the perception of children differs depending on gender and socioeconomic status. This includes a case study of children aged 11 who have drawn their home and nearby surroundings as part of Çanakçıoğlu’s research (2011). Keywords Children’s cognition, Cognitive maps, Environmental perception, Space syntax.


1. Introduction Children learn their environments through their actions in physical settings and several schemata that they develop through their reciprocal relationship with physical environmental stimuli. What is learned by this motor-sensory correspondence is reflected in certain kinds of behaviour in children. This interactive process depends on both the developmental stage of the child and the characteristics of the physical environment. J. J. Gibson, who tackles environment and behaviour theories through an ecological approach, stresses the importance of individual movement in the extraction of environmental knowledge and defines this process as a perceptual system in which the senses, behaviour of the perceiver and the physical environment are inseparable parameters (Wohlwill and Heft, 1987). 2. The development of perceptual process in children The reciprocal relationship between the physical setting and children, in the scope of developmental psychology, is studied by Jean Piaget (1955), who made great research contributions to how cognitive development is developed in children through a constructive approach. He argues that the cognitive development of children involves conditioning their behaviour by intuitions when they first meet the world; eventually as they grow up through ongoing experiences; they develop organized patterns of behaviours dependent on certain schemata. He argues that these organized patterns of behaviour eventually integrate children with the ongoing life around them and impact how they think through in a more questionable manner gradually depending on their age. Piaget classifies the perceptual system dependent on the cognitive development of children in four different stages depending on age (Cüceloğlu, 2009): (1) Sensorimotor Stage (0-2 years) in which children interact with the outside world with their intuitions and senses (Wilson, 1995); (2) Preoperational Stage (2-7 year-olds) through which children, behaving in an egocentric manner, start to represent the

objects and occasions with symbols and words and classify the objects in groups (Hart and Moore, 1973); (3) Concrete Operational Stage (7-12 yearolds) at which children form the ability to turn intuitions into operations, and can add, subtract, classify and put objects in order depending on their personal experiences; start to understand and feel empathy for others’ thoughts; and show a more adaptive behaviour to their environment (Piaget and Inhelder, 1967); (4) Formal Operational Stage (12+ year-olds), which is defined as the completion of the constructive approach of cognitive development, when children develop such a systematic attitude that they can use cognitive operations to tackle problems so that they can use conceptions and symbols without residing in concrete issues (Hart and Moore, 1973). So, it can be summarized that while senses and intuitions are dominant in a child’s life in the first two stages, through interaction and stimuli caused by the environment, the child shows the ability to think and question the issues in the last two stages of their cognitive development (Cüceloğlu, 2009). 2.1. Development of spatial perception in children According to Piaget and Inhelder (1967), the children define their positions in space depending on cues in the environment and later build relationships and paths between these cues. Piaget defines spatial representation in a child’s cognition as the reflection of environmental behaviour as symbolic and internalized cognitive information. Piaget and Inhelder reached four main findings about spatial perception: (1) children learn space through their ongoing individual activities; (2) children’s cognitive representations are constructed through the imitation of adult behaviour; (3) spatial perception is accomplished through four stages, which are sensorimotor, preoperational, concrete operational and formal operational stages, as stated earlier; (4) the development of spatial relationships is completed in three consecutive phases, which are topological, projective and metric or Euclidian spaces (Hart and

ITU A|Z • Vol 12 No 3 • November 2015 • N. G. Çanakcıoğlu


Moore, 1973). These consecutive phases of spatial relationship are reflected in cognitive maps as follows: Topological space As children reach the age of 7, although they cannot measure and compare objects in an appropriate unit and scale, they intuit that significant relationships such as proximity, separation, order, enclosure and seriation exist between objects. They can represent these kinds of relations of objects by drawing the patterns or objects in certain manners such as within a row and sequence, receding or approaching each other (Piaget and Inhelder, 1967). Projective space Projective space parameters refer to the coordination of objects in relation to others. Objects can be observed as simple perspective expressions despite some distortions. The general coordination of perspective is set up in the drawing (Piaget and Inhelder, 1967). Metric (Euclidian) space Through gaining an awareness that objects retain their size relative to the distance in between them as they change locations, the child comprehends that the representation of objects varies depending on altered reference points. Thus, children can organize spatial representations within their cognitive maps with a better understanding of the rules of perspective, and can draw objects in appropriate sizes depending on their locations while accommodating a certain reference point of sight (Piaget and Inhelder, 1967). Along with Piaget, Lynch puts forward ‘image parameters’ to discuss how people perceive their environment, construct and recall cognitive data and remember places through some distant characteristics. 2.2. Notion of image depending on Lynch’s theory According to Canter (1977), Lynch is the researcher who brought back the notion of image to the Earth by asking the question whether some cities are more imageable than others. He addresses this issue by revealing the

link between the mind and the physical environment, and puts forward such a classification system through an extraction of his interviewees’ sketch maps. He argues in his influential book The Image of the City that five characteristics of the image of the cities can be examined: paths, edges, districts, nodes and landmarks. According to Lynch (1960), while paths refer to continuous axis, edges are defined as boundaries. Whereas districts have certain boundaries within their homogenous characteristics, nodes refer to intersection points of paths showing different modes of behaviour, and landmarks are distinct reference points of cities that give cues to the visitor. Lynch’s imaginative parameters can also be adapted to interior spaces. Nodes may represent common meeting spaces, paths may represent linking elements such as corridors, edges may represent separating elements such as walls, districts may represent spaces designed for different modes of behaviour and landmarks may represent diversely emphasized spaces or elements (Gür, 1996). So, having introduced how children perceive their environment and in what manner spatial perception is related to individual cognitive development, the next part of the paper discusses the first setting in which children meet, perceive and construct an image of home, in terms of a nest, in their cognitive schemata. 2.3. Home as the first attachment place of children The home environment is the primary setting that children recognize and get familiarized and acquainted with. Although they participate in many institutions such as daycare centres, schools and playgrounds as they grow up, home remains the primary environment and the most dominant setting that children get to know (Wohlwill and Heft, 1987). The prominent French philosopher Gaston Bachelard (1969) discusses home through poetic language about his own experiences, and evaluates home as an existence with a wide and deep meaning. Moreover, he asserts

Can cognitive maps of children be analysed by space syntax?


that the existence of a being depends on the notion of the house so that house is not only a physical space containing a recollection of time and habits but rather a place full of memories. Bachelard says that children build up their first relationship with the rest of the world through their homes because “the house shelters daydreaming, the house protects the dreamer; the house allows one to dream in peace” (1969, p.88). Additionally, the house provides children the feeling of belonging to their families and society via a network of attachment and intimacy with others. While children grow up and adapt to their home environment as an origin point where they feel a sense of belonging, they gradually become acquainted with their nearby settings and experience a feeling of freedom. They start to discover the nearby environment attached to their homes: the entrance, stairways, terrace, porch, garden and urban elements such as the sidewalks, playgrounds and urban context in which the house is located. Simultaneously, this process means that children start to construct cognitive maps of the imaginary characteristics of the nearby setting of their homes. In this manner, Hart argues that the close setting of the home environment becomes the transitional space where children begin to experience the outside world. He discusses this transition as a multilateral process between the grownup, the child and the environment which should be well balanced by families in order to both give children the freedom to explore and experience but also to protect them from the risks and dangers of the environment (Chawla, 1991). 3. Cognitive maps as a tool to analyse the interaction of children in the built environment As the vast amount of the built environment is generally designed to be appropriate for adults in terms of physical conditions, children tend to behave and become influenced in a different manner than adults. Therefore, to understand the interaction between children and the physical settings, it is necessary to develop diverse research

methods to collect data through case studies that include the participation of children. According to Ziegler and Andrews (1987), there are some alternative methods to examine how the child interacts with the environment. Children can either be requested to (1) draw pictorial sketches, (2) show some objects in a representative setting, (3) locate objects on a representative model or (4) be interviewed for verbal comments. 3.1. The scope, aim and outcomes of the previous study Since this study uses the cognitive data of a previous study, it initially gives a brief overview of previous research and outcomes for a comparison with more recent outcomes. The aim of the previous research was to investigate the factors affecting the processes of perception in children in the scope of environment and behaviour theories. The research question asks how the nearby environment with the child’s home at the centre of close surroundings and cultural setting affects the spatial perception and richness of the cognitive schemata of children. The home environment is defined as the physical setting that the child first meets in terms of a nest that brings the feeling or sense of attachment. For the comparative study, an equal number of 11-year-old children were selected from two different physical settings with contrasting urban characteristics in Istanbul: (1) a group of children were selected from an informal housing setting, or a gecekondu housing structure at the outskirts of the city; (2) a group of children were selected from a gated community setting at the periphery of Istanbul. The case study researches whether the two independent variables of gender and socioeconomic status play a role in spatial perception and cognitive maps. In the case study of the previous research, two elementary schools were selected, one from each of the above-mentioned housing communities. Each group of children were asked to draw their home environments with their nearby settings in a 40 minute session on two different days. A total

ITU A|Z • Vol 12 No 3 • November 2015 • N. G. Çanakcıoğlu


of 82 children participated in the case study, including 19 girls and 22 boys in the low socioeconomic group, and 21 girls and 20 boys in the high socioeconomic group. Consequently, 82 cognitive maps were obtained to be analysed. The evaluation phase of the cognitive maps includes: (1) Piaget’s topological space parameters (proximity, separation, order, enclosure and seriation), projective space parameters (straight lines, parallel lines and perspective), metric (Euclidian) space parameters (conservation, block expression); (2) Lynch’s imaginative space parameters: paths, edges, districts, nodes and landmarks. Cognitive maps were analysed with respect to both gender and socioeconomic status. It should also be stated that an equivalent methodology was implemented in the research of Ünlü and Çakır (2002), which comparatively investigates the cognitive maps of primary school children going to school by foot, school bus and other vehicles. The research conclusions (Çanakçıoğlu, 2011) are summarized as follows: • An analysis of the income variable shows that there is a significant difference between the two socioeconomic groups. This outcome may result because children from high-income groups demonstrate cognitive maps that are richer in cognitive data. This may be due to the tendency of higher income families introducing more toys to their children. These toys may then trigger age-appropriate senses and stimulations. • A second outcome depending on the income variable is that the children from the high-income group living in the gated community draw maps focused more on interior spaces, which are already equipped with many belongings such as pianos, computers, vanity mirrors and guitars etc. • The children living in the informal setting, draw maps accentuating outdoor spaces such as the grocery store, motorways, internet cafes and fruit trees etc. since they tend to spend their spare time playing

outside on the streets, • When the results are analysed in terms of gender, in both income groups, girls tend to draw interiors through a more comprehensive and detailed manner than the boys. This may result from the sociocultural norms of Turkish society that tends to raise girls in a comparatively more inward oriented manner than boys. Consequently, while the girls from the high-income group living in the gated community have their own rooms furnished with special possessions to keep themselves busy indoors, the girls from the low-income group living in an informal setting spend their time at home, where they do not have a separate room for themselves, and do not play outdoors as much as the boys. • The cognitive maps show that children living in the gated community are not as aware of the physical setting of their home environments as the children living in the informal setting. The reason for such an outcome could be that although a gated community is designed with recreational facilities appropriate for children, children cannot actually use them without the guidance of an adult. Since parents do not feel it is safe enough to allow a child at this age to go to the outdoor areas even though they live in a gated community, the cognitive maps of the children from the high-income group imply that they watch their immediate neighbourhood from their windows. In contrast, the children living in worse physical housing conditions reflect their urban experiences more onto cognitive maps. Although they probably cannot find appropriate facilities designed especially for them, they discover their play materials and spaces outdoors on the sidewalks and void lands. They play with mud, pebbles and bricks, and play hopscotch on the porches of their apartments, creating their own playing scenarios. As revealed in the previous study, the provided analyses are based on the spatial parameters of Jean Piaget (Piaget and Inhelder, 1967) and the image pa-

Can cognitive maps of children be analysed by space syntax?


rameters of Kevin Lynch (1960). There is an ongoing debate on the theory of space syntax whether space syntax can be used as a research tool to analyse cognitive data. The main aim of this paper is to find out whether the comparative quantitative method of analysis of space syntax may be an alternative method to examine the cognitive maps of children. Next, the paper will discuss attempting to use space syntax as a research tool to analyse cognitive maps through an interdisciplinary approach. 3.2 Space syntax as a tool to analyze spatial cognition According to the research of Zheng and Weimin (2010), there is a significant relationship between the syntactic configurations of real spaces and cognitive maps of the interviewees who participated to the study. In addition, the analyses show that despite the errors in cognitive maps drawn by the interviewees, there still is a significant consistency between the maps and the real environmental situation. So, it is possible that cognitive maps can be used as an integrated research tool in space syntax studies. Additionally, Zheng and Weimin (2010) mention that the axial maps of spaces can also be evaluated as the simulations of cognitive data of people moving around these spaces. In the scope of their study, the researchers analyse both the values and correlations between global integration, local integration, connectivity and depth values, and figure out that cognitive map representations significantly correspond with the existing settings. Thus, it is debated through a progressive manner that cognitive maps integrated with space syntax can be used as a tool to analyse the association between spatial organization and spatial cognition. Moreover, Zimring and Dalton (2003) embrace the research area of space syntax and cognition through a collaborative manner. This contributes to (1) the understanding of the relationship between the physical characteristics of environment and cognitive representations, (2) reveals how physical space is related with behaviour and (3) offers an innovative methodology

to analyse the existing spatial structures. In addition, Haq and Girotto (2003) in their study of two hospital settings contribute to this interdisciplinary research area using intelligibility as a measure to analyse cognitive maps, search for a way to analyse the idea of movement and offer a complementary grounded theory. Additionally, Kim and Penn (2004) seek to make research on space syntax interdisciplinary, and try to handle the issue through an integrated manner to reveal the relationship between the outcomes of human behaviour in the physical environment and spatial cognition. According to their survey comparing the sketch maps of residents with the existing layouts in a neighbourhood in London, they find that local integration degrees in cognitive maps are especially linked to the syntactic data of the actual environment. 3.3. Space syntax as an alternative methodology to analyse children’s cognitive maps Within the framework stated above, this study aims to: 1.search for a supplementary or alternative research tool to analyse cognitive maps of children 2.contribute to the methodology of analysing cognitive maps of individuals through the quantitative method of space syntax. To implement this aim, 72 cognitive map drawings by children obtained through a case study as part of the research of Çanakçıoğlu (2011) are analysed. These maps, which have already been analysed within the spatial parameters of Piaget (Piaget and Inhelder, 1967) and the image parameters of Lynch (1960) and have been concluded with some significant results, are now examined using justified permeability graphs within the concept of space syntax. The aim is to comprehend whether there are any unique and significant results. In this manner, it is also aimed to discuss whether space syntax can be implemented to analyse the cognitive maps of children as a substitute method. Having reviewed the previous interdisciplinary research in terms of space syntax and spacial cognition (Zheng

ITU A|Z • Vol 12 No 3 • November 2015 • N. G. Çanakcıoğlu


and Weimin, 2010; Zimring and Dalton, 2003; Haq and Girotto, 2003; Kim and Penn, 2004), this paper targets to develop a new approach to analyse cognitive maps. The scope of this paper aims to contribute to the analysis of cognitive maps drawn by 11-yearold children from two distinct social groups. 3.3.1. Steps of analysis of the study Obtaining significant results requires an equal number of drawings from both of the gender groups and income groups, meaning that 18 girls’ and 18 boys’ drawings are selected randomly from each income group so that a total of 72 cognitive maps are analysed. Before explaining the steps of analysis, the acceptance procedure used in the evaluation phase of the drawings should be mentioned. It is not possible to analyse children’s drawings as typical layouts or spatial structures through conventional methods. As each child was requested to draw his or her home environment and nearby setting through his/her eyes, there is no need to compare such subjective data with any existing plan data. Therefore, each drawing on an A4 size paper accomplished by the children with their own subjective attitude is accepted as a total cognitive space no matter whether they have tried to draw their homes within a garden, house layout plans within a façade view or houses as a façade within a street silhouette as seen in Figure 1. Indeed, children draw their cognitive maps freely as a reflection of their experience. In fact, Kaplan (1973) supports this view, defining cognitive maps as “schematic, sketchy, incomplete, distorted and otherwise simpli-

fied and idiosyncratic” and “a product of experience, not of precise measurement”. Consequently, in this study, the analysis of cognitive maps is conducted by justified permeability graphs, or justified gamma maps, which Hillier and Hanson (1984) say, “permit easy measurement of these syntactic properties. Thus, justified gamma maps are intended to allow a form of analysis that combines the visual decipherment of pattern with procedures for quantification.” In addition to Table 1, the analysis steps are as follows: (1) Following the notion of convex map described by Hillier and Hanson (1984), a simplified drawing to clearly describe the spatial structure and connections between them within the drawing is prepared for each cognitive map. (2) Each space of the cognitive map is identified by a letter in a circle, defined as a node, to create a basis for a justified graph. Following the definition of root space by Hillier et al. (1987) and Klarqvist (1993), the exterior space of the house is identified as the root space in each spatial representation. Other spaces are identified with nodes sequentially, depending on the permeability factor between the spaces. (3) To conduct a syntactic analysis, justified permeability graphs are prepared for each cognitive map in order to show the direct relation between spaces by syntactic steps, including the permeability values and depth levels of the spatial structure (Klarqvist, 1993). Both the permeability and depth values are calculated for each cognitive map. Via the justified graphs prepared, it is possible to convert the subjective data

Figure 1. Two examples of cognitive maps showing the house within a garden and within a neighbourhood. (Boy, 11; Boy, 11). Can cognitive maps of children be analysed by space syntax?


Table 1. An example to demonstrate the steps followed to reveal syntactic value of each cognitive map.

Original Cognitive Map (Girl, 11 years old)

Step 1

Simplified Drawing Showing the Spatial Structure with Nodes

Step 2

Justified Graph

Permeability Value


Depth Level


Step 3

expressed by children from a visual to mathematical verification, which can be concretely shown. (4) In the last step, tables are prepared to see whether any significant results arise and correlate with the total amounts and arithmetic mean value of permeability and depth values, dependent on gender and income level variables. Therefore, it is observed that there is the potential for a comparative discussion about boys’ and girls’ cognitive maps and different socioeconomic levels through a syntactic analysis. In the evaluation phase of the study, permeability and depth level data obtained from the cognitive maps are evaluated by Pearson chi-square analyses through the statistical software – SPSS Table 8 to find out any significant associations among the gender and income variables.

Table 2. Distribution of permeability and depth levels depending on gender.

Boys Girls

Total Permeability

Total Depth Level

163 155

76 68

Figure 2. Graph showing total amount of syntactic steps (permeability) and depth levels depending on gender.

ITU A|Z • Vol 12 No 3 • November 2015 • N. G. Çanakcıoğlu


Table 3. Distribution of total permeability of boys’ drawings dependent on income group. Total Permeability of the Boys' Drawings Low Income Group


High Income Group


Table 4. Distribution of total permeability of boys’ drawings dependent on income group. Total Depth of the Boys' Drawings Low Income Group


High Income Group


Figure 3. Scattergram of boys’ cognitive maps showing the permeability dependent on the variable of income group.

Figure 4. Scattergram of boys’ cognitive maps showing the depth level dependent on the variable of income group. Table 5. Distribution of total permeability of the girls’ drawings dependent on income group. Total Permeability of the Girls' Drawings Low Income Group High Income Group

104 51

4. Results • Firstly, all 72 maps are analysed by means of the tree-like justified permeability graphs to find the total amount of permeability and depth levels of all executed data. The distribution of these two values dependent on gender variable can be monitored in Table 2 and Figure 2. Considering total permeability, the cognitive maps of the boys reflect more syntactic steps in terms of permeability, meaning that the spaces are connected to each other through more branches, than the girls’. Considering the total depth level, the spaces drawn by boys are represented through a deeper manner than the girls’. • A second outcome is found dependent on the income variable. As seen in Tables 3-4 and Figures 3-4, the total permeability and depth levels of boys’ cognitive maps from low-income group show higher values than the ones from the high-income group. The highest permeability value is detected as 22 with a depth level of 6 in the high-income group. • However, there is also an exception that contrasts with this result. One map from the high-income group shows 0 syntactic step, meaning that the child only drew one space and nothing attached or integrated with this space within the cognitive map. The children who drew only one space in their drawings mostly drew their own rooms without any other spaces connected to the space drawn. Cognitive maps reflecting only one space are mostly observed in the high-income group. These cognitive maps are evaluated as “0” in terms of both syntactic step and depth level. • As monitored from Tables 5-6 and Figures 5-6, the total amount of permeability and depth levels of girls’ cognitive maps from the low-income group show higher values compared to the other girls’ maps from the high-income group. It is also noticed that the gap between the trend lines are wider in the girls’ scattergrams than the boys’, which means that the cognitive maps

Can cognitive maps of children be analysed by space syntax?


• • •

of the girls from the low-income group show a more interconnected and deeper structure of spaces within their drawings. Nine out of 18 girls’ drawings are evaluated as “0” since the girls from the high-income group living in the gated community only drew their own rooms without any other connected spaces to their rooms. Within these drawings, the one and only space which is highly accentuated and richly designated are their own private rooms; these children prefer to emphasize their personal equipment such as computer, TV, musical instruments etc. as seen in Figure 7. From these results, it may be inferred that these children tend to spend most of their time in their rooms and that they attach more personal significance to these because they hold their personal belongings. In the final tables and figures, the analyses of all 72 cognitive maps of children are shown. As seen in Table 7 and Figures 8-9, according to Pearson chi-square tests processed in SPSS, the changes in the total amount of permeability and depth level are significantly higher Table 8 in the drawings of the children from the low-income group than those of the high-income group: Income level; permeability: x²=12,359, df=4, p=0,015 < 0,05 (significant) Income level; depth level: x²= 16,687, df= 2, p=0,000 < 0,05 (significant) Both the permeability and depth values of the low-income group are almost twice that of the high-income group, revealing that children living in the informal setting are more aware of their nearby environments centred with their homes. Indeed,

they display the ability to represent the spaces they are living in through a more complex interrelationship and integration within their cognitive maps.

Figure 5. Scattergram of girls’ cognitive maps showing the permeability dependent on the variable of income group.

Figure 6. Scattergram of girls’ cognitive maps showing the depth level dependent on the variable of income group. Table 6. Distribution of total depth of the girls’ drawings dependent on income group. Total Depth of the Girls' Drawings Low Income Group


High Income Group


Table 7. Distribution of total permeability and total depth of all cognitive maps dependent on income group. Total Permeability

Total Depth Level

Low Income Group



High income Group



Table 8. Association between permeability and depth level with gender and income level. Item




Gender; Permeability



0,538 > 0,05


Gender; Depth level



0,480 > 0,05


Income Level; Permeability



0,015 < 0,05


Income Level; Depth Level



0,000 < 0,05


ITU A|Z • Vol 12 No 3 • November 2015 • N. G. Çanakcıoğlu


5. Conclusion The aim of the study is to find a quantitative supplementary research tool to analyse the cognitive maps of children, thus contributing to the research methodology of analysing individual cognitive maps within the theory of space syntax. Through the analysis of children’s cognitive data, it is possible to contribute the following assertions to the debate:

Figure 7. An example of a cognitive map of a girl from high income group representing only the child’s own room. (Girl, 11 years old).

Figure 8. Scattergram of all children’s cognitive maps showing the amount of permeability dependent on the variable of income group.

Figure 9. Scattergram of all children’s cognitive maps showing the depth level dependent on the variable of income group.

From the syntactic analyses of an equal number of drawings of girls and boys at the age of 11 for a total of 72 cognitive maps, it is shown that boys’ cognitive maps are more permeable and deeper in terms of spatial structure compared to girls’ cognitive maps. This difference results in both of the income groups. This outcome may be due to the sociocultural norms of Turkish society that is accustomed to raising girls in a more inward oriented manner than boys. • In addition, it is shown that half of the cognitive maps of the girls from the high-income group living in a gated community are evaluated as “0” in terms of syntactic value since they preferred to draw their own rooms in a fragmented manner, displaying their personal belongings. This finding could also enter the debate on sociocultural issues in Turkey, especially as families from both the middle class and high-income groups tend to have fewer children than those of the low-income group and have a greater chance to reserve a separate room for each of their kids. • In contrast to the above outcome, children from the low-income group, which does not have the opportunity to have a separate room, are the most likely to have more experience in the other rooms of their interiors and in the outdoor spaces attached to their homes. As low-income families tend to have more kids, they do not have the chance to reserve a separate room for each of the siblings. Also, since these children are used to going to school by foot alone or in the company of their siblings, friends or parents, they have more opportunities to observe their nearby environments within an urban context. Coming in contact with the urban layout becomes a bodily experience that a child may store in their cognition. Thus, it is possible to say that children from low-income groups living in an informal setting have more environmental experience and construct more knowledge within their cognitive schemata, thus reflecting deeper syntactic values onto their

Can cognitive maps of children be analysed by space syntax?


cognitive maps. • Another finding is that the children from the high-income group show a lower capacity, and produce shallower and low permeable cognitive maps as if they are used to watching the urban life happening around them from their windows in their own rooms. It is possible that these children may be (1) more engaged with their own technological devices such as tablets and laptops or (2) accustomed to going to school by private vehicles. Both lead to only visually experiencing the urban layout, and this is not enough for a child to store experience in cognition. Thus, it appears that children from a high-income group living in a gated community construct less environmental knowledge within their cognitive schemata and reflect shallower syntactic values onto their cognitive maps. This raises the concern about whether children from gated communities ever learn to discover the outdoors. • The three outcomes mentioned above show that the syntactic values of the cognitive maps of children depend significantly on the income level variable. The reason why the cognitive maps of children from the low-income group show a significantly higher degree of permeability and depth level may be due to the distant characteristics of the environments that the children live in and where they spend their spare time. Children who emphasize outdoor spaces in their maps may have had more opportunity to play outdoors and can therefore represent their home environments in a deeper and integrated manner. Alternately, children who merely emphasize their own rooms with their personal belongings may not have had the opportunity or preference to play outdoors as freely as the other children as they live in a gated community, an artificially designed environment with planned recreational activity areas for children. • According to these results, justified permeability graphs measuring the amount of permeability and depth values may be used to measure and

analyse the cognitive maps of children. • In conclusion, cognitive maps of children can indeed be analysed by space syntax, giving an affirmative answer to the initial research question and title of the study. References Bachelard, G. (1969). The Poetics of Space, Translated by: Maria Jolas. Boston: Beacon Press. Canter, D. (1977). The Psychology of Place. London: The Architectural Press Ltd. Chawla, L. (1991). Homes for Children in a Changing Society, In E. Zube, G. T. Moore (Eds.) Advances in Environment, Behavior and Design: Vol. 3. (pp. 187-224). New York and London: Plenum Press. Cüceloglu, D. (2009). İnsan ve Davranışı. İstanbul: Remzi Kitabevi. Çanakçıoğlu, N. G. (2011). İstanbul’da farklı sosyal grupların yerleştiği çevrelerde yaşayan çocukların algısal süreçlerinin bilişsel haritalar yöntemiyle irdelenmesi, (Yüksek Lisans Tezi), Fen Bilimleri Enstitüsü, İstanbul Teknik Üniversitesi, İstanbul. Gür, Ş. Ö. (1996).  Mekân örgütlenmesi. Trabzon: Gür Yayıncılık. Haq, S., Girotto, S. (2003). Ability and Intelligibility: Wayfinding and environmental cognition in the designed environment, 4th International Space Syntax Symposium London 2003 Proceedings. Hart, R. A., & Moore, G. T. (1973).  The development of spatial cognition: A review. Chicago: AldineTransaction. Hillier, B., & Hanson, J. (1984). The social logic of space  (Vol. 1, p. 1984). Cambridge: Cambridge University Press. Hillier, B., Hanson, J., & Graham, H. (1987). Ideas are in things - An application of the space syntax method to discovering house genotypes.  Environ Plann B, 14(4), 363-385. Kaplan, S. (1973). Cognitive maps in perception and thought. In R. M. Downs & D. Stea (Eds.), Image and environment: Cognitive mapping and spatial behavior. Chicago: Aldine. Kim, Y. O., & Penn, A. (2004). Linking the spatial syntax of cognitive

ITU A|Z • Vol 12 No 3 • November 2015 • N. G. Çanakcıoğlu


maps to the spatial syntax of the environment. Environment and Behavior, 36(4), 483-504. Klarqvist, B. (1993). A space syntax glossary.  Nordisk Arkitekturforskning, 2, 11-12. Lynch, K. (1960).  The image of the city  (Vol. 11). Massachusetts, USA: MIT press. Piaget, J., Inhelder, B. (1967). The Child’s Conception of Space. New York, USA: The Norton Library. Piaget, J. (1955). The Construction of Reality in the Child (Conclusion Part) Retrieved from University of Oregon website: http://pages.uoregon.edu/ rosem/Timeline_files/The%20Construction%20of%20Reality%20in%20 the%20Child.pdf Ünlü, A., Çakır, H. (2002). A study on perception of primary school children in home environment.  Journal of Architectural and Planning Research, 19(3), 231-246. Wilson, R. A. (1995). Nature and Young Children: A Natural Connec-

tion. Young Children, 50(6), 4-11. Wohlwill, J. F., & Heft, H. (1987). The physical environment and the development of the child, In D. Stokols, I. Altman (Eds.) Handbook of environmental psychology,  Vol.1, (pp. 281328). New York: Wiley. Zheng, L., & Weimin, G. (2010, August). A spatial cognition investigation by using the integrated methodology combined with cognitive map and space syntax. In Computer Science and Education (ICCSE), 2010 5th International Conference on (pp. 111-115). IEEE. Ziegler, S., Andrews, H.F. (1987). Children and Built Environments - A Review of Methods for Environmental Research and Design, In R. B. Bechtel, R. W. Marans, W. Michelson (Eds.) Methods in Environmental and Behavioral Research. (pp.301-336). Florida: Robert E.Krieger Publishing Company. Zimring, C., & Dalton, R. C. (2003). Linking objective measures of space to cognition and action. Environment and Behavior, 35(1), 3-16.

Çocukların bilişsel haritaları mekân dizimi yöntemiyle irdelenebilir mi? Çevre ve davranış teorileri kapsamında ele alındığında, insanlar, çocukluk çağının ilk yıllarından itibaren, duyuları aracılığıyla etkileşim içinde bulundukları ve çevresel uyaranların tetiklediği bir ortam dâhilinde, bir algılama ve öğrenme sürecine dâhil olurlar. Art arda ve zincirleme bir biçimde devam eden bu algısal süreçler sonucunda, insanlar, çevreden edindikleri bilgiyi adeta zihinlerinde yeniden inşa etmek suretiyle özgün zihinsel şemalar oluştururlar. Bilişsel olarak sağlıklı bir insanın hayatı boyunca devam eden bu süreç, bireyin öğrenme mekanizması olarak da tanımlanabilir. Bu bağlamda, çocukların çevrelerini nasıl algıladıklarını ve öğrendiklerini konstrüktivist bir yaklaşım çerçevesinde irdeleyen gelişim psikoloğu Jean Piaget (1955), çocukların çevreyle olan etkileşimini ve algısal süreçlerini, yaşa bağlı olarak değişen bilişsel gelişim teorileri bağlamında inceler. Her çocuk, çevresini birbirinden farklı algıladığı ve çeşitli çevrelere ilişkin birbirlerinden farklı bilgiler biriktirdiği ve söz konusu bu çevrelere ilişkin kendilerine özgü

zihinsel şemalar inşa ettiği için, her çocuğun ortaya koyduğu bilişsel haritanın da tekil ve özgün olarak değerlendirilmesi gerekmektedir. Çocukların çevrelerine ilişkin zihinlerinde inşa edilen ve depolanan bu bilgiyi ortaya çıkarmanın yöntemlerinden birisi, çocukların bizzat kendileri tarafından çizilen bilişsel haritaların analiz edilmesiyle gerçekleştirilebilmektedir. Zihinsel şemalar ve bilişsel haritaların analiz edilmesi konusunda çevre ve davranış teorileri kapsamında literatüre önemli katkılar sağlayan Kevin Lynch (1960), yollar, kenarlar, bölgeler, nodlar ve nirengi noktaları şeklinde sınıflandırdığı mekânsal bileşenlerin kullanıldığı güçlü ve nitelikli bir imgesellikle tasarlanmış olan kentsel çevrelerin, insanların algısal süreçlerine anlamlı bir katkı sağladığını ve bu sayede, bireylerin söz konusu bu çevrelerde oryantasyonlarını daha iyi kurabildiklerini belirtmektedir. Bununla birlikte, yukarıda belirtilen çevresel bileşenleri içeren çevrelerin, mekânların hatırlanmasında ve bununla birlikte bireylerin zengin bilişsel haritalar ortaya koymalarında önemli katkısı bulunduğunu ifade etmektedir. Bununla birlikte

Can cognitive maps of children be analysed by space syntax?


bazı araştırmalar kapsamında bilişim kavramı, mekân dizimi (space syntax) araştırmaları kapsamında irdelenmeye çalışılmaktadır (Zheng & Weimin, 2010; Zimring & Dalton, 2003; Haq & Girotto, 2003; Kim & Penn, 2004). Yukarıda belirtilen tartışma alanı çerçevesinde bu çalışma, Piaget’ nin mekânsal algı parametreleri ile birlikte Lynch’in imge parametreleri kullanılmak suretiyle, çocukların mekânsal algı düzeylerinin cinsiyet ve sosyoekonomik durum değişkenlere bağlı olarak nasıl değiştiğinin kıyaslandığı bir çalışma için elde edilmiş olan bilişsel haritaların analiz edilmesine alternatif bir yöntem sunmayı amaçlamaktadır. Bu sebeple, çalışmanın tartışma alanı, farklı iki sosyoekonomik gruptan gelen 11 yaş grubundan çocukların kendi ev ve yakın çevresine ilişkin mekânsal algılarının incelendiği “İstanbul’da Farklı Sosyal Grupların Yerleştiği Çevrelerde Yaşayan Çocukların Algısal Süreçlerinin Bilişsel Haritalar Yöntemiyle İrdelenmesi” başlıklı çalışmanın (Çanakçıoğlu, 2011), alan araştırması safhasında elde edilen bilişsel harita verilerinin mekân dizimi yöntemiyle analiz edilmesi üzerinedir. Başka bir deyişle, çalışmanın ana hedefi, bilişsel

haritaların analiz edilme aşamasında, mekân dizimi (space syntax) yönteminin alternatif bir nicel analiz metodu olarak kullanılmasına olanak sağlayıp sağlamadığını araştırmaktır. Çalışmanın içeriğinde, mekân dizimi yöntemiyle irdelenen bilişsel haritalar, “koridor, giriş holü ve merdiven gibi lineer mekânların, erkek çocukların haritalarında kız çocuklarınkine kıyasla daha fazla vurgulanıp vurgulanmadığı”; buna paralel olarak, kız çocukların haritalarında içe dönük mekânların daha fazla miktarda ortaya çıkıp çıkmadığı” gibi konular cinsiyet değişkeni bağlamında, “üst sosyoekonomik grupta yer alan çocukların çoğunluğunun, kendilerine ait odalarının bulunması sebebiyle, alt sosyoekonomik gruptaki çocuklara kıyasla daha sığ (shallow) mekânsal ilişkilerin yer aldığı haritalar ortaya koyup koymadığı” gibi hususlar ise sosyoekonomik durum değişkeni bağlamında irdelenmek istenmektedir. Anlamlı sonuçlar elde edilmesi öngörülen bu hususların irdelenmesinde, doğrulanmış geçirgenlik haritalarına (justified permeability graphs) başvurulmak suretiyle derinlik (depth) analizleri uygulanmıştır.

ITU A|Z • Vol 12 No 3 • November 2015 • N. G. Çanakcıoğlu

ITU A|Z • Vol 12 No 3 • November 2015 • 141-158

Disintegration of urban housing areas: Districts and new gated housing settlements

Suat APAK suatapak@yahoo.com • Department of Architecture, Faculty of Architecture, Istanbul Technical University, Istanbul, Turkey

Received: May 2015

Final Acceptance: October 2015

Abstract In Turkey and Istanbul, gated communities are produced rapidly and offered for use of the urban users particularly since 2000s. Urban spots created by the new gated settlement areas and the life style started appearing and spreading rapidly in outskirts of Istanbul. Starting from the emergence of these walled settlements rise, urban integrity gets disrupted and neighborhood lifestyle goes through a transition. The biggest problem posed by these islands in the city is the fact that they create non-interactive areas in the external world in contrary to the inner world. The objective of this study is to determine the interaction level between these wealth islands and existing texture and constituents of the existing texture by using the space syntax approach. First, the strength of the bonds within the current environment is determined by using the space syntax program on the current environmental plans. The methodology also followed determining the physical and social behavior patterns of the streets, mainly the paper concentrated on the intersected spaces belonging to such different life styles. These observations were made on the quality of the physical environment (order, graffiti, vandalism, etc.) and the utilization intensity. In the final stage, the data is collected by using the space syntax program and it was analysed to determine whether there is a correlation between the existence of graffiti and space syntax determinants. Results show that the interaction within the existing texture gets weaker at substantial levels around the walls of the island,contraversially the existence of graffiti increases,and sociologically existing users living around the walled islands turn to be “societies around / bottom the walls”. Keywords Districts, Gated communities, Security, Space syntax, Graffiti and vandalism.


1. Introduction By rapid production of gated housing settlements started from 1980s and in particular 2000s in Turkey and firstly we observed their emergence in Istanbul. When the world has been starting to be globalized, it brought a capital movement, but also it has effects in the society as economic dispersion around the world and income distribution disparity. Because of the capital movement effects, the economic, technologic and organizational development of the firms boosted increasing the supply and as reverse we witnessed demand and rapid expansion of the luxury gated community settlements in these era. Self-enclosed and privatization of the streets are the indicative directions of such kind of new life concept. The rich and luxury life including expensive homes, golf clubs, tennis clubs, and fitness centers have been constructed. The security and richness create the key positions, in order to overcome undesirable events such as crime, drug, vandalism, unrespectable behaviors towards public and private properties, the walls and doors that have been established as physical barriers. Protection of the settlement physically means the protection of the virtue and wealth values based on the settlement accordingly. Hence, the protection of these values is as important as physical security (İçli, 2010). On the other hand, Blakely and Snyder (1997) states that gated communities could be combined within three main categories. These are lifestyle communities, prestige communities and security zone communities. The dominant factor regarding a settlement in this categorical approaches,we may perceive that many privileges have been provided and controlled in the gated community island. This makes these settlements profitable (Görücü and Pektas, 2014) and this is the most important factor for the investment firms and entrepreneurs. Such new settlement areas and new living styles have started to appear in Istanbul and mainly outskirts in the existing neighborhoods of Istanbul. Since the walled such settlements, the urban integrity has started to be separated and street living culture has en-

tered into the transition period. The social problems have been apparently observed around these islands and we may perceive the contradiction, in spite of wealthy life style inside, the low-income families are located in the surroundings,as outside. When such islands provide the living groups as their wishes as well as living conditions and their securities, contoversially they create insecure areas in their external worlds. As it was stated by Oscar Newman (1972), “when a town is defended individually (private group)-or lived, it means that the struggle against the crime at that town is lost”. 2. The negative effects of the gated housing settlements in the urban life The Gated Housing Islands have damaged the current urban structure and have barricaded the integration and the development of the region. On the other hand, they have damaged the sustainability of the districts and traditional district culture, as contrary they have created and submitted the local artificial tissues and different life models for the community that is not peaceful to their neighbors. In some cases, these walls and fences not only prevent access to the facilities of their own but also they might make it hard for the public to access common areas such as streets, pavements, parks, beaches, rivers, paths, children playgrounds and other local resources (Blakely and Snyder, 1997). When we turn back to the current structure, it creates insecure areas in the contact lines. The treats are mostly directed to the people who are living in these islands as well as current people who are using such routes. By means of the rails, secure walls and security cameras based on twenty four hours monitoring, the precautions against all possible threats toward the inside have been taken. The undefended areas are the external belt zones as a result of such extraordinary exclusion. 2.1. The surveillance in terms of creating secure environment One of the basic components of the social control is the natural surveillance mechanism. In the books of Jane Jacobs (1961) called as “The Death and Life of ITU A|Z • Vol 12 No 3 • November 2015 • S. Apak


Great American Cities”, she signed the importance of the natural surveillance by stating the phrase of “the eyes in the street”. Jacobs mostly attributes her determinations based on the surveillance principle and she is one of them who emphasized the “defensible space” concept where such components have an important place. She stated that “in some cases to help for a stranger or to call for the police, in order to enable the person to act, he/she must own the street and he/she must believe that if he/she will need for a help, will have the urgent help”. However, she also stated that the most important think rather than acting and the condition to act are the surveillance and he stated the importance of the surveillance (Jacobs, 1961). Newman (1972) considers the surveillance function in the category of “physical design capacity that will pro-

Figure 1a. Tpical access control and surveillance concepts and classifications (Crowe, T.,D.,1991).

Figure 1b. The conceptual shift from organized and mechanical concepts has led to the natural CPTED concepts (Crowe, T.,D.,1991).

vide the surveillance possibility to the user and their relatives”. He explains this category as the mechanics towards increasing the surveillance capacity continuously and their internal and external areas of the people. He also states that they are the integral parts of the three basic mechanisms. On the other hand, Crowe (1991), in his book called as “Crime Prevention through Environmental Design” talks about the availability of three overlapped strategy in CPTED. These are natural access control, natural surveillance, and territorial reinforcement. He states that access control and surveillance are the primary design concepts of the physical design programs (Figure 1a, Figure 1b). Stollard (1991) argues about the availability of the design options to use it for deterring from the crime and he states that there are some common agreements and conciliations over some general principles. Stallord collected such principles under six groups and he placed the surveillance principle in the first instance. This principle is related to the natural or passive surveillance of the homes and public places around the homes. When this situation is considered as the indicative factor for the uninvited guests, it is accepted that deterring from the crime is the most basic and essential element. Natural surveillance can be defined as the impression that residents of the dwelling are guarded or can be guarded if necessary and that there is an eye on neighbors and on their residences. Architects and planners should take into account of current options to maximize the opportunities of the resident users (Stollard, 1991). By considering these thoughts, one of the design and settlement principles for safe residential area proposed in the “defensible space” is to direct the buildings to the streets. These directions provide opportunities that have been made by the residential building for the surveillance and also encourage them and that increase the “eyes on the street” accordingly (Figure 2). In a research that was made in Istanbul, Levent, it was understood that the house direction has been made on the streets and the streets that have high

Disintegration of urban housing areas: Districts and new gated housing settlements


natural surveillance capacity are intensively used by the pedestrians. The side of the streets that are limited by the walls and garden walls of the houses and they are less preferred by the pedestrians and they are determined that the deaf walls along the streets are mainly neglected and used as the graffiti ground (Apak, 2005). 2.2. Permeability as the integration tool The permeability is an important factor to integrate physical volumes and to give permissiveness for human behaviors. In the assessment of this factor, the scale and intensity should be primarily taken into the consideration. In the settlement where the scale is small and intensity is low, we may conclude that “the permeability is minimized”. “The settlements have been designed around the cul-de sac” exclude the foreigners and create confidential and the safe areas (Ünlü, 1986), in such areas, social control is not expected from the interaction of the in-out settlement users but it is expected from the settlement owner. Apart from the security feeling, the visibility and monitoring is controlled by the the user of the area. It is quite natural that, foreigners were regarded as suspects when they were observed in front of the tents (yurt) at nomadic times before the

permanent settlement (Küçükerman, 2007). If we observe the cul-de sacs that are surrounding square , they are inhabited by people who know each other well in traditional permanent settlement system. The existence of semi-public and private areas as shown in Newman’s chart of territorial area (Figure 3) supported with the opportunities of observation. Certainly, the subjects like intensity and urban liveliness are not desirable in these areas. In current town structures where the scale is getting larger, the blood relation, close relativeness, close neighborhood and other relevant elements are removed and where the settlement users become stranger, the subjects like the permeability and creating vitality may be considered as important factors. In 9th part of the book called as Jacobs’ “the Death and Life of Great American Cities”, the requirement for small block has been stated in detail. By considering the example of Manhattan, it is stated that stable long blocks create desperate long, monotone and dark strips. If there were streets that cut such wide blocks lies from the east to west and if there were more than shorter blocks, it is stated that there would not be any need to use such monotone way to go somewhere and the alternative routes could be chosen (Figure 4). Thus, she defenses that the distribution of the stores and

Figure 3. Schematic sketch illustrating territorial definition reinforced with surveillance opportunities (Newman,1972).

Figure 2. Houses should face each other along the street (Poyner, B., Webb, B., 1991).

Figure 4. Small block requirement in Manhattan (Jacobs,1961). ITU A|Z • Vol 12 No 3 • November 2015 • S. Apak


eligibility of the stores and commercial places will be increased as a result of the interaction. She argues that, in this way, streets will not be isolated but they will get mixed to each other which will ensure significant increase in the number of trade destinations and improvement in the availability and expansion of shops. She emphasizes that, streets do not end at successful or attractive neighborhoods. On the contrary, the number or streets increases where possible (Jacobs, 1961). Bentley & Alcock & et al. (1993) reserve the first part of the books to the permeability and deal the permeability in detail. They state that only places that let the people could provide options for them and the quality of the permeability is related to the number of the alternative ways that pass from one area. In the left diagram of Figure 5, comparing with the top layout with the sub layout, it provides large option possibility so the top layout is more than permeable. In the right diagram of the same figure, it is concentrated on the advantages of the small blocks. They deal with the permeability as physical and visual permeability, they also reiterate that a place with the small blocks provide more option than the area with the large block and they state that large block layout has only three alternative route without returning between A and B and small block version has a little short distance nine alternatives with the public route (Bentley and Alcock, et al., 1993). They draw the attention of the availability of the three design approach that works against the public location permeability and they organize them as increasing the scale of the settlement, hierarchic layout using and pedestrian/vehicle separation.

Figure 5. Different approach alternatives in the permeability sense for the same area (left) and small block requirement between A-B axles (right) (Bentley,Alcock,Murrain, McGlynn, Smith, 1993).

On the other hand, Crankshaw (2009) states that permeable street walls could be provided by the door entrances, windows and stores. A monolithic building without any openings creates insecurity and it gives you a disturbing feeling like you are all alone, walking naked along a street (Crankshaw, 2009). 2.3. Urban vitality / commerce in the sense of mixed utilization It is clear that the urban vitality in the urban structure of the commercial functions has very important place in the vitality of the life. The commercial life particularly in the town scale is related to the accommodation functions of the town, the interactions, solidarity and living together in the same area. When considering the opposite angle in terms of the solidarity of the parties, in the understanding the modern town planning (however, the functional difference is essential in this understanding), despite the fact that the sub center settlements focused on the commercial activities when they become deserted at the night time. Jacobs, who emphasized mixed-use over separated dwellings or commercial places, also prioritized the liveliness of the people in current neighborhoods in her book which is full of the most original and powerful claims of her age (Jacobs, 1961). There is no doubt that in the sense of the urban vitality and â&#x20AC;&#x153;commercial vitalityâ&#x20AC;? street understanding that is the supporting element in the mixed utilization, the interaction along the street and the supporting attitudes are more than important even if they are competitors. It is possible to see the examples in the traditional and historical structure of Istanbul. When approaching to the end of 18th century, Istiklal Street that has many opposite buildings and in which complete design was completed in the second half of 19th century (Cadde-I Kebir) is one of the most important street where the luxury buildings, European types of stores, recruitment and resting places are located after the reorganizations, nowadays, it keeps the same style indeed so that the commercial vitality has been observed. On the other hand, the spice

Disintegration of urban housing areas: Districts and new gated housing settlements


bazaar that is one of the oldest shopping centers has the covered street plan where many stores are located. The interaction between small commercial units (such as the flower bazaar, street) which stick to external, blind walls as required by the inward-oriented structure and the commercial units at the ground floors of the buildings that are located opposite the street, they are the most significant elements in the union of the bazaar with the surrounding pattern (Figure 6). When talking about the urban vitality, the availability of the stores is required. On the other hand, in our history, from the shopping stores (arasta) to the local bazaars, streets, and covered bazaars and even in our modern world, large shopping centers, shopping malls (that contrary approach to our opinion – withdrawn huge masses ) has mutual shop units. In this plan approach, face to face organization, sequence and interaction are basic elements. The pavements are social spaces. The streets, pavements need ultimate attention to continue their living powers apart from the channels where the pedestrians use. They should be encouraged to create social interaction and activity together with their adjacent external selling areas in front of the stores and pavement cafés (Crankshaw, 2009). 3. Model Ralph B. Taylor (1987) separated the scale organized from the least serious to the most serious in the figure called as the continuum of disorder. He showed the graffiti in the condition concept and he included the vandalism in the crime case based on the most serious step. The graffiti that could be defined as using the private and public properties by unpermitted letter or paints. Their existence affected the environmental image and it has created the negative effect on the sense of security. The graffiti could be performed under the direction various purposes. The “gang graffiti” that could be a sign how power the gangs and illegal organizations could be used as interaction tools to show their powers and to remove others from these areas (Taylor, 1988).

On the other hand, one of the graffiti type is the “taggers” graffiti etc. Each graffiti has not written by the member of the gang. These could be used to state how to find the drug, drug users etc. They could have some threat messages against other races, religious and genders. Some graffiti is the sample of the swearword and profanity. Some of them could reflect the politic views and some also could be classified as “gum balloon” like “ Ali loves Ayse” etc. The type of the graffiti could be changes based on the time and place. However, the graffiti is possibly coming from the “taggers” who sign their activities by their typical names. The taggers expect and hope that they would be admired because of their respective activities and personal based works. The taggers could seek excitement. Such excitements could include the danger to make graffiti in the high place or to violate the laws accordingly. The taggers could act alone or they could be gang to make graffiti as a team. In the recent researches, there are different approaches on the graffiti. Some researchers think that the graffiti is a kind of vandalism that is located on the roads, in the public transportation and stations, commercial regions and residential areas and some researchers think that these are urban artwork (Figure 7) and some of them consider that the graffiti is an activity to draw the attention of the young people to positive study fields.

Figure 6. Two- sided commercial using of the streets, İstiklal street and Çiçek Pazarı street ( Apak, S. 2015).

Figure 7. An example of graffiti as an art work and a vandalism example on the same street in Kadikoy ( Apak, S. 2015). ITU A|Z • Vol 12 No 3 • November 2015 • S. Apak


When dealing with undesirable action, it is seen that various methods have been applied to be prevented. Apart from the publications and internet possibilities that facilitate the processes what we should do especially for raising the social awareness, there are precautions towards prevention the crime factor. Such precautions have been dealt and evaluated in three categories in the researches such as “law application and implementation” where the law application precautions are provided by means of the criminal proceeding system, “perpetration on time” to minimize the damage and lost and “potential guilty prevention” to care about the young people and to draw their attention to the different places (Stafford & Pettersson, 2003). Though graffiti might give messages in some cases such as the gang graffiti, it disturbs the society as an illegal act regardless of its content or form (whether it has an artistic value or not). When it damages to the public and private properties and when it is against the authority, in the context of the environmental image, they are significant signs weakening the environment’s potential to create the feeling of safety. Apart from such messages provided to the users, it affects the environment and the population deeply. The value of the properties could be decreased and possible commercial advantages could be lost. The population could be self-enclosed, the neighborhood relations could be weaken and the sensitivity could be lost as well as these

Figure 8. The cart of the theoric model.

could be the reason for social-economic damage and disruption. While they might indicate negligence, dereliction or ignorance they might also indicate that social/environmental control is lacking or there is unwillingness to perform such control. Even if some part of the society think that these are not important, they create illegal environment and they provide clue for illegal behavior possibilities and such kind of behaviors. As a result of such negative effects of the graffiti, it has created an important fact as the indicator of the weakness related to the trust and confidence in the current urban fabric especially in the intersections of the gated islands. Without emphasizing the reasonability of the graffiti, the factors that facilitate the application and that enable the environment to be created as is defined in the theoretical framework of the activity are surveillance/ visibility, permeability, vitality, syntactic integration value. The availability or lacks of such factors are the reason of presenting, decreasing or increasing the graffiti applications. The key component of those factors is based on the permeability. Due to the fact that it has a one by one effect on other factors if a piece of urban texture is permeable or not and these could affect directly or indirectly the graffiti applications. As it is seen in the current model, permeability, within the context of visual permeability, by eliminating the possibility of identifying the route through the external user and surveillance/visibility of the external area through the internal user, the minimal occupancy, uncontrolled areas and routes have been created (Figure 8). On the other hand, when considering the physical permeability, it has minimized the neighborhood relations together with the adjacent current urban fabric. Because of the immense deathfullness created by the walls as the border element, such borders have lost the chance to transform the lively, illuminated areas such as shopping store, café, etc. It has minimized the vitality at those areas and it has prevented the possible potential development accordingly. It has also created negative

Disintegration of urban housing areas: Districts and new gated housing settlements


effects on the vitality of current tissue. In these areas, decreasing the vitality means decreasing the number of the responsible eyes that will provide the social control at those areas and decreasing the opportunity of being supported. This huge area where the permeability is not available has provided their relations with the external world by one or two controlled doors and they have turned their backs to the current urban fabric. Their relations are almost little or nothing. In this case, it has created important effects on the syntactic integration values. The syntactic integration values that are not integrated with the network of the current urban fabric is minimized and disconnected, uncontrolled, desolate environments have been created. These environments provide more comfortable opportunities for the graffiti applications that could be perceived as the indicator so that the most important crimes could be committed. These factors that have unilateral or bilateral interactions could increase or decrease the applicable physical ground and opportunity of the graffiti. 4. Method / Case studies While determining the case study areas, the attention was paid that they should be neighbor with the unplanned and undeveloped urban structure, new gated community islands. The purpose here is to catch the clues clearly to reflect the relations of such islands with their surroundings. In this framework, “Maslak-İstinye Park” that is neighbor to Pınar district that is illegally structured area (Figure 9a), “TEM1 Avrupa Residences” which are located at 4th Region slum prevention area of Gaziosmanpaşa Town which has de-

veloped badly in a mixture of planned and unplanned construction (Figure 9b) and “Zeytinburnu Kiptaş Topkapı Merkez Residence“ (Figure 9c) were selected as case study areas. Main common points of these areas are the weakness socio-economic structure along with an unplanned and badly developed urban structure and side by side closed residence islands addressing the particular income group and their ownership characters . Maslak-İstinye Park Residence among the case study area was constructed as 19 blocks was composed of 406 residential units. The entrance to the island is made from one gated access. On the other hand, TEM1 Avrupa Residences is composed of 36 blocks includes 3100 residential units and the entrance could be provided from three points. Topkapı Merkez Evleri is also composed of 803 residential units including 5+7 and totally 12 blocks. In each of two staged island, there are one entrance point. In the selected three case study areas, the method is applied to obtain data based on the observations regarding the current urban structure around the gated community areas and intermediate zones. As the assessment variables, the users living in the vicinity of islands have been taken into the account. The intensity route or points in the urban area regarding the user and their experiences have been recorded. Regarding the physical environment, the environmental quality was determined as the scale of good-medium-bad and in the adjacent environment and intermediate zones, the availability of the graffiti and vandalism is searched and the determination of the places were achieved accordingly.

Figure 9. The case study areas, a; Istinye Park Residences-Maslak, b; TEM Avrupa Residences – Gaziosmanpaşa, c;Topkapı Merkez Residences – Zeytinburnu. ITU A|Z • Vol 12 No 3 • November 2015 • S. Apak


In the parallel of these studies, updated maps regarding these three areas that have been defined based on the observation criteria. The reason for having the updated maps is to reveal the rapid development of Istanbul and to show that new settlements have been performed in last 5-6 years. The integration where the updated maps have been used as data input that is the second important step for the determination of the syntactic integration analysis. The principle spatial element is the axes of the urban structure. The organization of the building or the district is based on the organization of its axes, that is, they reflect the behaviors, the sequence of behavioral experiences. The axis is fundamental because the experience of architecture is an experience of movement (Hillier 1996). For the analysis of the syntactic integration, the Syntax 2D program by the Michigan University is used. By selecting axial line analysis and preparing axial maps, it is aimed to define the integration values of the close neighborhood around the these gated community islands. By the achieved observation data, the integration values are compared and they are searched that if the low-high integration area and graffiti is overlapped or not. According to the hypothesis, the main task that is searching if there is any correlation between the deep integration values and

increase of graffiti or not. In particular, the research concentrates on the specific places in the vicinity of gated islands, the occurrence of the graffiti applications and their meanings, their syntactic integration with the physical environment. 5. Discussion 5.1. Graffiti points observed around the sampled gated community islands According to the determinations based on the observation, around the gated community islands of the three case study areas, the availability of the graffiti is observed. The long deaf walls extending uninterruptedly and their horizontal plane form may be quite appropriate space for graffiti. In the north of Istinye Park Residences, on İğde Street in the intermediate section of “Pınar Mahallesi (District)” (Figure 10a İ1), in Sarıyer Street and its opposite and in the walls of Enka Schools (Figure 10a İ2), in the western part of Gaziosmanpaşa TEM1 Avrupa Residences, on the walls located in 1612th Street and its extension (Figure 10b A1), in western part of the Kiptaş Topkapı Merkez Residence, G-36th Street (Figure 10c M2) and on Gümüşsuyu Davut Paşa Street (Figure 10c M1), the graffiti samples are mainly observed areas. By going around the surrounding

Figure 10. The regions where the graffiti was applied on the walls of the gated community islands. Disintegration of urban housing areas: Districts and new gated housing settlements


walls of each three samples, the place and number where the graffiti has been applied are signed and determined on the plans. In the determinations that have been made, the type and size of the graffiti is ignored. The existence of graffiti has been recorded numerically. In accordance with these determinations, the graffiti has been recorded intensively applied in these case study areas. It was observed that the walls of İstinye Park Residence is on İğde Street and partly in Kılıç Street. The determined graffiti number around the wall on this street is (Figure 10-i1) (26 İğde Street+13 Kılıç Street) totally 39 points (Figure 11). There is also graffiti situated in Gaziosmanpaşa TEM1 Avrupa Residence, and at the same, an intensive graffiti has been observed on the wall throughout 1612th Street with the adjacent current urban fabric (Figure 10A1). The number of graffiti that was signed by counting on those walls are 22 indicators (Figure 12). In the Zeytinburnu Topkapı Merkez Residence part 1 and 2, the intensive graffiti applications that have been frequently used, applied and re-painted have been observed on the throughout street and stair walls that provides connection between recreation area (that located two parts of the gated residences) and Davutpaşa Gümüşsuyu Street (Figure 10-M1). The graffiti paint number that was determined at the time of observation are 12 in stair walls and 16 on the walls at Gümüşsuyu Street and totally 28 at the same region (Figure 13). It was observed and determined that the contents of these graffiti could be political, class difference and they could also be inserted as the type of “gum balloon” that is not important. 5.2. The determinations regarding the permeability specifications of the sampled gated community islands As it is known that the basic principle of the gated community is to prevent the uncontrolled entrance and to create impermeable areas where the controlled entrance is provided. In order to provide the security, to prevent the infrastructural using share, to cre-

Figure 11. Graffiti on the walls of the island of İstinye Park Residence (İ1) and İğde Street ( Apak, S. 2015).

Figure 12. Graffiti on the walls of the island of TEM1 Avrupa Konutlari (A1) and neighbourhood ( Apak, S. 2015).

Figure 13. Graffiti on the walls of the island of Topkapi Merkez Evleri (M1) and parked cars ( Apak, S. 2015). ITU A|Z • Vol 12 No 3 • November 2015 • S. Apak


there are three entrance-exit doors at Gaziosmanpaşa TEM1 Avrupa Residences due to its size of the covered area (Table 1). The heights, qualities and materials that have been used for each surrounding walls of three case studies are given in Table 2. These walls are prohibiter as the visual permeability, the transparency that provides communication,as well as the physical permeability. All these features could be completely or partly observed in each samples. The front of the walls on Abdi İpekçi Street that is the important artery was concealed by the green plants. The height of the bottom reinforced concrete part around 1612th Street where this wall is interacted with the current urban fabric was approximately increased from 40 cm. to 160 cm (Table 2). On the other hand, in order to end the visual relation, the back of the iron fences was closed by semi opaque plastic panels. Zeytinburnu, Topkapı Merkez Residences are composed of two parts. For second section, there

ate the social status living types and similar groups, firstly under the direction of providing the physical impermeability, the basic approach is to surround these areas by the walls. For the sake of minimizing the internal-external irritation and increasing the security, apart from the height and type of those walls, the visual impermeability is provided and internal-external relation has been cut completely (Table 1). The number of blocks and resident units, size of the covered areas and shortest – longest lengths of the areas belongs to these three case studies are given in table 1. Also quantities and qualities of these surrounding walls are important features for permeability. The total length of the walls that surrounds İstinye Park Residence, Gaziosmanpaşa TEM1 Avrupa Residences and Zeytinburnu Topkapı Merkez Residences are 1238 meters, 1665 meters and 1524 meters. There is only one entrance-exit door at İstinye Park Residence and Topkapı Merkez Residences 1- 2. (seperately). On the other hand

Table 1. The quantitative specifications of the sampled gated residential areas and the length of surrounding walls.

İstinye Park Residence Gaziosmanpaşa TEM1 Avrupa R. Topkapı Merkez Residence sec.1 Topkapı Merkaz Residence Sec.2

Number of block

Number of Residence

Area (m2)

Perimeter (m)









Number of Ent-Exit door 1+ (1 from mall, pedest.)

Length - short

Length - long














1+ (1 from mall, pedest.)



803 7

Table 2. The qualitative specificaitons and the heights of surrounding walls of the sampled gated residential areas.

The height of the walls

İstinye Park Residence Gaziosmanpaşa TEM1 Avrupa R. Topkapı Merkez Residence

The quality and materials of the walls Tota l (cm. )




Base (cm. )

Mid. (cm. )






reinforced concrete

natural panel wire fence






reinforced concrete

forged iron fence

double razor wires





reinforced concrete

natural panel wire fence

single razor wires


Disintegration of urban housing areas: Districts and new gated housing settlements


are also pedestrian accesses from the shopping. In the middle of both section, there is a recreation area and open shopping units for public and these two sections could be considered as two different gated islands. The quality of the walls that surround these islands could be regionally different. The recreation area between two islands is defined by the shopping units’ facades. It is also observed that 2nd part of island walls on Davutpaşa Gümüşsuyu Street and G-32th Street are composed by shopping units’ facades which is constructed under residential blocks. In Topkapı Merkez Residences, together with the physical obstacles and by the massive material such as concrete, the visual barrier was maximized. The positive side of this sample is to provide residential blocks together with the shopping units. 5.3. The integration analysis of the sampled gated community islands and surroundings In each three case studies, surrounding gated community islands, neighbor streets integration-n of Syntax 2D values have transformed into the table. For each case study, approximately 30 streets give us an idea about general characters of the close existing urban texture. For each area, the mean of integration values are calculated and when we compared the integration-n values of the cross sections around the gated community, the significant results are emerged. The mean of the integration-n values of 25 street belonged to the current settlement surrounding Istinye Park Residence is 11590,98. On the other hand, the mean average of the integration-n values belonged to the roads surrounding Istinye Park walls is 2023,35. When these values are calculated based on 30 streets surrounding TEM1 Avrupa

Residence, it was achieved to 10998,55, on the other hand integration- n values of the roads adjacent to surrounding walls are calculated as 2786,95. In Zeytinburnu Topkapı Merkez Residence, the mean value of 30 streets adjacent to the current settlement is 12039,80. Average values are calculated in the parallel streets of islands’ borders, around 1st section is about 344,71 and in the second section, it is performed as 2355,65 (Figure 14). Hillier (1996) stated that the distribution of integration in the axial map defines an “integration core” which generates not only a movement pattern but also a distribution of land uses such as shops and residences which are sensitive to movement. In the axial line analysis of Pınar Mahallesi (District) and Istinye Park Residences, when considering the integration-n diagram, the integration value of the İğde Street (3796,28) adjacent to the Istinye Residences where graffiti is applied is very low. However, Çamlıbel Street (Figure 15) that is the most populated street of the “Pınar Mahallesi (District)” has highest integration values (22555,91) in the red color scale and as is mentioned by the Hiller, it appears as the core of the integration. Naturally, on this street, due to the fact that there are many shopping units and due to the vitality of the street, it is more significant to overlap the core in the integration map. It is observed that 1612th Street and its extension where the graffiti points are highly observed on the border walls of TEM1 Avrupa Residences in Gaziosmanpaşa has lowest integration values in the dark blue color scale in the axial line diagram produced for this region. On the other hand, the integration values are rather low at G-36th Street axis hosting graffiti on deaf and castle type of walls surrounding Topkapı

Figure 14. The Axial Line Analysis and Integration-n Values of the three case studies (Istinye Park Residence – Gaziosmanpasa TEM1 Residence – Topkapi Merkez Residence). ITU A|Z • Vol 12 No 3 • November 2015 • S. Apak


Merkez Evleri and the vicinity of Gümüşsuyu Davutpaşa Street axis that are divided by the tram line. 5.4. The determined graffiti applications and syntactic integration analysis study Along with the graffiti point maps achieved by the observations, the integration-n value maps obtained by Syntax2D are also compared and interposed and this paper reinforces that there is a high correlation between the increase of graffiti and low integration-n values around the gated communities (Figure 16). The mean of integration-n values of 24 streets in Pınar District that is current urban fabric surrounding partly Istinye Park Residences is calculated as 11590,98. The axis numbered 81 that is the starting point of İğde Street con-

Figure 15. Camlibel Street (on the left), Igde Streets (on the right) ( Apak, S. 2015). Integration Values of Istinye Residence and Pinar Mahallesi district (below).

Figure 16. The overlapping the determination maps of the region where the intensive graffiti applications are available in three case studies with the value map of the Axial Line Analysis / Integration-n. (a) Istinye Park Residence – (b) Gaziosmanpasa TEM1 Residence – (c) Topkapi Merkez Residence).

nected to Camlıbel Street (where the mean value is 22555,91) is 7274,00 as integration value. The axis numbered 82 where the graffiti is intensively seen has a value of 3796,28 and it is located between İğde Street and Kılıç Street. The axis numbered 91 represents Kılıç Street. Its integration value is 2819,87 and 13 graffiti points determination is made on the walls at these streets. The integration value of the axis coded 92 is 1450,18 and it is part of the wall between Kılıç Street and site entrance door. The axes numbered 95 and 96 have the values of 787,22 and 807,81. These are connection ways between Kılıç Street entrance-exit door and Sarıyer Street connection. Due to the proximity to the controlled area with the security guard there are not any graffiti applications. On the other hand, because of their too low integration values, this area is a type of urban fabric has the weakest relations with their environments (Table 3). The mean of the integration-n values of 30 streets related to the current urban fabric around Gaziosmanpaşa TEM1 Avrupa Residences is 10998,55. The axis coded 0 has the value of 2222,43 and it is entrance point to the existing adjacent fabric from Abdi İpekçi Street. The axis coded 1 is the entrance way to the current fabric from Abdi İpekçi Street (1277,87) and it is the part that provides connection to 1612th Street and on each two axes. We may oberve here that there is not any current indicator about the graffiti points. The axes of 2 (1274,28) and 3 (3419, 37) are extensions of 1612th Street and the adjacent walls to the of TEM 1 Avrupa Residences and particularly, there are graffiti points on these walls along two axes. The number of these points is 22. The axis coded 47 (2316,06) represents the entrance door and Abdi İpekçi Street connection way and the axis coded 48 (1140,71) represents the extension of the entrance door and they show the weak relation tie in the rare points where they could integrate with the integration values less than (10998,55) of current fabric average of Gaziosmanpaşa. The arithmetic average of 30 streets was taken in the adjacent area of Zeytinburnu Topkapı Merkez Residences and

Disintegration of urban housing areas: Districts and new gated housing settlements


the integration-n values are 12039,80. The consecutive axes coded 116, 117, 118 represent the way between the 1st section walls and tramway station. Respectively, they have integration values of 960,50 – 375,17 – 237,99. We may consider that this region where the graffiti applications are intensively used. 16 graffiti points are determined. The axes numbered 148 (375,18) and 149 (228,91) represent the stairs to the public area between two sections from this region. The continuous graffiti applications are observed horizontally on the walls in both side of the stairs and this area has been restricted by painting through the administration. 12 graffiti points are determined on the day when the observation is carried out. The axis numbered 25 (175,52) in the 2nd section, the axis numbered 27 (257,14) in the 1st section that represents the pedestrian ways in front of the shopping units under the blocks that they face to the public area. Due to the fact that these areas are not eligible to apply graffiti and there are some monitoring residential units such as living rooms over the stores. These areas have very low integration-n values however it is hard to remark of traces about the graffiti activities (Table 3). The integration values obtained from the axial line analysis of the sur-

rounding streets limited by the walls are comparatively are very low, contoversially we may record that the high occurence of graffiti on the walls. The surrounding wall of the island has blue and dark blue colors that they represent the low values in the color scale. It means that the great gated islands that are not integrated with the current structure, and they make cool their environment as well. Such areas that are adjacent the gated community islands are the weakest and colorless routes of the urban/neighborhood interaction and they become the area as vacant car parking locations that may also cause blockading the pedestrian movements. 6. Result Such case study areas like the representatives of the most of the gated community islands around Istanbul have the mean of their negative specifications. The negative aspects might also be found in implementations of state based organisations like Housing Development Administration of Turkey (TOKI). The assessment of the valuable lands belonged to the public organizations within current urban structure on one public sector hand, and profitability of the different lands that are the properties of the private sector on the other hand, the both sec-

Table 3. Comparing the sampled three gated community areas surrounding and adjacent street integration values and graffiti point numbers.

Integration value (Avg.)

Number of Graffiti points

İstinye Park Residence



Gaziosmanpaşa TEM1 Avrupa K.



Topkapı Merkez Evleri



Graffiti locations (frequently seen) Cod Int. value Name of the location e 81 7274,00 Igde Street 82 3796,28 Mid part of Igde-Kilic S 91 2819,87 Kilic Street 92 1450,18 Kilic Street- Entrance 95 787,22 Entrence-Sariyer Street 0 2222,43 Abdi Ipekci St.-Existing t. 1 1277,87 Abdi Ipekci St.-1612.St. 2 1274,28 1612.St. adjacent to ex.t. 3 3419,37 1612.St. adjacent to ex.t 48 1140,71 Entrence –TEM1 Reside. 116 960,50 Davutpasa Gumussuyu S. 117 375,17 Davutpasa Gumussuyu S. 118 237,99 Davutpasa Gumussuyu S. 148 375,18 Stairs 149 228,91 Stairs

ITU A|Z • Vol 12 No 3 • November 2015 • S. Apak


tors may boost these problems around the constructed neighborhoods. The weakness in the local town-planning applications, the lack of authorization of producing local town-planning and codes and the planning and design cosiderations as the isolated islands of the Housing Development Administration of Turkey (TOKİ) create continuously the lack of the urban plan integrity, permeability and vitality problems within the existing urban life. In the location selection of the gated community islands, the economic feasibility has been become the most basic factor.However, the social-cultural-economic structures of the adjacent areas of the islands coexisted within the current urban life at surroundings have not taken into the consideration and the permeability of human movements and the structural continuity of the districts are completely ignored. The results within the scope of the research show that the interaction in the current structure is declined around the island walls. As it is seen in the syntactic integration analysis clearly, the mean of the integration values achieved by taking the current urban fabric are the strongest evidence in this comparison. It is observed that such the mean values are time by time higher than the integration values of the street surrounding the gated community areas. In these areas, when the intensive of the graffiti points increases, this indication is the proof that those areas are becoming so desolate areas. Besides, it could be seen that by the analysis that the current situation is sharply overlapping with the real life and how the vitality and permeability notions in planning issue how they are removed. By the analysis of the observed and calculated data, the zones between such wealthy islands and existing life at surroundings are isolated and ineffective formations. As supported by analyses, they come up that these available spaces for graffiti cause to disturbance and displeasure among the neighborhood situated around the endless walls. Psychologically, existing graffiti on long walls is a kind of a message and scream to the sensitive community. It is also a sign for threatening indicator

damaging safety and moreover it creates the weakening for the perception of the feeling of safe and uncontrolled environment. At this point, it could not be expected to break the walls and to bring very different social –economic and cultural communities together. Such a kind of expectation cannot been demanded by parties at this moment. At presently, the gated community users could not accept different social and physical structure so that they have been injected. They cannot communicate as much as possible in the vicinity. They attempt to make connection with the towns by their private cars without communicating with close social areas . The solution is based on defining new gated settlement locations and integrated and permeable areas. In the local applications, for the sake of the economic income, instead of producing gated community islands, it is better to develop open structures that might be integrated with the environment and to reveal the district based on the synthesis as well. In the large scale, without establishing exact lines, sharp sides, the aim must be an urban planning approach comprising more peaceful and transitive altitude (from low income to high etc.) and to process and to apply more productive urban planning understanding that will not cause any attempt to the high income groups to have a right to implement hidden and wealthy gated community solutions. References Acar, S., Kumral, B. (1997). “İş – Alışveriş Merkezleri”, YEM (YapıEndüstri Merkezi Yayınları), Tunç Matbaacılık A.Ş., İstanbul. Apak, S. (1998). “Güvenli Çevrelerin Oluşturulmasında Kullanılabilecek Kavramsal Bir Model”, (Unpublished doctoral dissertation), İstanbul Teknik Üniversitesi, Fen Bilimleri Enstitüsü, İstanbul. Apak, S. (2005). “Bir Konut Bölgesi Yaya Sirkülasyonunda Tercihlerin Güvenlik Duygusu Bağlamında Değerlendirilmesi”, Konut Değerlendirme Sempozyumu, İstanbul Teknik Üniversitesi, Mimarlık Fakültesi Baskı Atölyesi, s;83-96, İstanbul.

Disintegration of urban housing areas: Districts and new gated housing settlements


Bentley, I., Alcock , A., Murrain,P., McGlynn,S., Smith, G. (1993), “Responsive Environments – A Manual For Designers”, Butterworth – Heinemann Ltd., Printed by Hartnolls Ltd., Cornwall. Blakely, E.,J., Snyder , M.,G. (1997). “Fortress America – Gated Communities in the United States”, Brookings Institution Press., Printed by R. R. Donnelley & Sons Co., Washington D.C. Crankshaw, N. (2009). “Creating Vibrant Public Spaces – Streetscape Design in Commercial and Historic Districts”, Island Press, Washington D.C. Crowe, Timothy, D. (1991). “Crime Prevention Through Environmental Design – Applications of Architectural Design and Space Management Concepts”, The British Library Document Supply Centre, National Crime Prevention Institute, Butterworth – Heinemann Ltd., Stoneham, Massachusetts Görücü, E.,Ö., Pektaş (2014). “İstanbul – Gaziosmanpaşa- Küçükköy Mahallesi Avrupa konutları TEM-2 Projesi Değerleme Raporu”, (Emlak Konut – 12.14 – 161 nolu Rapor), Yetkin Gayrimenkul Değerleme ve Danışmanlık A.Ş., İstanbul. Hillier, B. (1996). “Space is the Machine – A Configurational Theory of Architecture”, Cambridge University Press, New York. İçli, G. (2010). “Statü Sembolü

Olarak Konut ve Konut Kullanımı Denizli Örneği”, Pamukkale Üniversitesi Yayınları, No:13, Denizli. Jacobs, J. (1961). “The Death and Life of Great American Cities”, Vintage Books, New York. Küçükerman, Ö. (2007). “Turkish House in Search of Spatial Identity”, Türkiye Turing ve Otomobil Kurumu, Euromat Entegre Matbaacılık, Istanbul. Newman, O. (1972). “Defensible Space – People and Design In The Violent City”, Architectural Press, W & J Mackay Ltd., Great Britain, Chatham. Poyner, B., Webb, B. (1991). “Crime Free Housing”, The British Library Document Supply Centre, Butterworth – Heinemann Ltd., M & A Thomson Litho Ltd., Scotland. Stollard, P. (1991). “Crime Prevention Through Housing Design”, Chapman & Hall, London. Ünlü, A. (1986). “Geleneksel Çevrelerde Tasarım Verilerinin Saptanması İçin Bir Model”, (Published doctoral dissertation), İstanbul Teknik Üniversitesi, Fen Bilimleri Enstitüsü, İstanbul. Stafford, J., Pettersson, G. (2003). http://www.dft.gov.uk/stellent/groups/ gft_mobility/documents/ page/dft_ mobility_025965-11.hcsp (http://www.gophaber.com/haber3847-arsiv.html).

Kent konut alanlarının parçalanması: Mahalleler ve yeni kapalı konut yerleşmeleri Türkiye ve İstanbul’da özellikle 2000’li yıllardan sonra kapalı konut yerleşmeleri büyük bir hızla üretilerek kent kullanıcılarının kullanımına sunulmaya başlanmıştır. Sermayesi gelişmiş özel girişimciler ve / veya özel – kamu işbirliği ile planlanan bu yeni tip yaşam anlayışlı sunumlar, bir mega kent olan İstanbul’un dahi bu oranlarda hiç karşılaşmadığı ölçek ve sayılardaki yeni konut alanlarıyla hazırlıksız bir şelikde karşı karşıya kalmalarına neden olmuştur. Bu yeni kapalı yerleşim alanları ve yaşam tarzının oluşturduğu kentsel lekeler, İstanbul’un içlerinde ve eteklerinde hızla belirmeye ve yayılmaya başlamıştır. Bu lekelerin, yapay adacıkların kent

için en büyük problemi, mevcut kent dokusuna yapışarak veya mevcut doku içlerindeki mevcut veya kentsel dönüşümle elde edilen boşlukları doldurma yöntemiyle mevcut dokuyu iteleyerek sıkıştığı alanın kendisine ait yarattığı iç dünyasının tersine dış dünyasında etkileşimsiz alanlar yaratmasıdır. Mevcut dokudan koparılan bu parçalar ile mevcut alanların birbirleri ile olabilecek olası etkileşim, birleşme şanslarının da ellerinden almasıdır. Tüm bu olumsuzluklarının yanı sıra, bazı durumlarda, duvar ve parmaklıklar ile sadece kendilerine ait donatılara erişimi değil, caddelere, kaldırımlara, parklara, plajlara, nehirlere, patikalara, çoçuk oyun alanlarına, lokal tüm vatandaşlar tarafından paylaşılacak tüm kaynaklara kamusal geçişi de engelleyebilmekte, zorlaştırabilmektedirler ITU A|Z • Vol 12 No 3 • November 2015 • S. Apak


(Blakely and Snyder, 1997). Mevcut dokunun ihmal edilip dışlanması ile karşı komşulukları ile etkileşim şansları ellerinden alınarak, bu sokak kullanıcıları yalnızlaştırılmakta, geçirgenliği sağlayacak, arttıracak, doku ve sokak süreklilikleri kesilmekte, kentsel canlılığı zayıflatılmış, ticaret potansiyeli azaltılmış çevreler yaratılmaktadır. Neticede kullanıcı yoğunluğunun endişe yaratacak boyutta azaldığı, doğal gözetim olanaklarının ortadan kaldırılıp, güvenlik hissinin zayıfladığı kuşaklar yaratılmaktadır. Makalenin teorik tabanında, güvenli ortam yaratma bağlamında doğal gözetim, kullanımı teşvik ederek sosyal kontrol perspektifinden pozitif yoğunluk yaratacak bir bütünleşme aracı olarak geçirgenlik ve fonksiyonel ayrışmışlık yerine karma kullanım unsurları ortaya konularak derinlemesine irdelenmiş ve makalenin yaklaşımı doğrultusunda kuramsal bir çerçeve oluşturulmaya çalışılmıştır. Mevcut kent dokusu içleri ve eteklerine serpiştirilen, gerektiğinde tıkıştırılan bu lekeler, bu farklı dokuların yarattığı ara kesitler bağlamında, düzen ve güven ortamının zayıflık göstergesi olarak duvar yazıları (grafiti) ve vandalizm ele alınmış, özellikle duvar yazılarının fiziksel çevrenin niteliği ve bu çevrenin kullanıcılar üzerindeki psikolojik, sosyolojik ve ekonomik etkilerini konu edinmiş düşünsel ve araştırmalara dayalı çalışmalar ortaya konulmuştur. Mevcut doku ile sırtı dönük adacıklar arasındaki temas bölgelerinin geçirgenliği, kullanım tercihliği ve yoğunluğu, canlılık ve sosyal kontrol, doğal gözetim varlığı veya yokluğu ile güvenlik hissine etkisi bağlamında duvar yazılarının varlığı, uygulanabilme ortamı bulabilmesi ve verdiği mesajlar makalenin ana omurgasını oluşturmaktadır. Bu açıdan çalışma alanlarının seçiminde mevcut kent dokusu ile yakın temas halinde olan kapalı konut adacıklarının seçimine özen gösterilmiştir ve özellikle plansız veya kötü gelişim göstermiş kentsel dokuyla komşu, iç içe geçmiş, yeni kapalı konut adalarının bu özenli seçiminde amaçlanan, lüks tüketim özellikleri gösteren bu adaların çevreleri ile olan ilişkilerinin

dışa yansımasının ip uçlarını daha belirgin olarak yakalayabilmektir. Çalışma yöntemi açısından mevcut dokunun kendi içindeki geçirgenliği, bütünleşmesi ve bu dokunun kapalı konut adacıkları ile bütünleşmesinin, aradaki bağların varlığının, bağların zayıflığı veya kuvvetliliğinin yani bütünleşme değerlerinin tespiti önemli bir bileşeni oluşturmaktadır. Diğer bileşen olarak, kapalı konut adalarının yakın çevrelerindeki mevcut kent dokusunun ve ara zonlara ilişkin gözlem ve bu gözlemlere dayalı verilerin elde edilmesi gelmektedir. Gözlem değerlendirme kriterleri olarak, ada dışı kullanıcılar ve fiziki çevre nitelikleri üzerine odaklanılmıştır. Fiziki çevreye ilişkin tespitlerde, yakın çevre ve ara zonlarda grafiti ve vandalizmin varlığı araştırılmış ve olan / yoğunlaşan yerlerin tespiti yapılmıştır. Bir binanın veya kent dokusunun organizasyonu, deneyimlenmesinin ardışıklığı nedeniyle, akslarının organizasyonudur. Mimarinin deneyimlenmesi, bir hareketin bir akımın deneyimlenmesi olması sebebiyle aks esastır, temeldir (Hillier 1996). Bu nedenle, bütünleşme değerlerinin saptanmasi için Axial line analizi seçilmiş ve çalışma alanlarına uygulanmıştır. İntegrasyon analizi için Michigan Üniversitesinin geliştirdiği Syntax2D programı kullanılmıştır. Güncel haritalardan axial line haritaları herbir bölge için ayrı ayrı Autocad programı ile hazırlanmış ve hazırlanan bu haritalar Syntax2D programına aktarılarak integrasyon değerleri hesaplanmıştır. Axial line haritalarının hazırlanmasında bölgenin sokak örüntü ağı ve bu ağı oluşturan sokaklara açılan herbir kapının eklenilmesi esas alınmıştır. Elde edilen gözlem verileri ile entegrasyon değerleri karşılaştırılması yapılarak, Düşük-Yüksek entegrasyon alanları ile grafiti Varlığı- Yokluğu örtüşmelerinin ve değişkenler arasında bir korelasyon bulunlup bulunmadığı araştırılması yapılmıştır. Özellikle sık grafiti uygulamalarının olduğu noktalara yoğunlaşılmış, bütünleşme değerleri arasındaki ilişkilerdeki anlamlılıklar sorgulanmış, fiziksel çevre niteliği ve kullanımına yansımaları irdelenmiştir. Gözleme dayalı yapılan tespitlere

Disintegration of urban housing areas: Districts and new gated housing settlements


göre, her üç çalışma alanının kapalı konut adaları çevrelerinde grafiti varlığına rastlanmıştır. Doğal olarak metrelerce kesintisiz uzanan bu sağır duvarlar grafiti için çok elverişli düşey düzlemler, grafitiye altlıklar oluşturmaktadır. Grafiti nokta haritaları ile Syntax2D ile elde edilen integration-n değer haritaları enterpoze edilip karşılaştırıldığında çok büyük bir oranda örtüşme olduğu görülmektedir. Pınar mahallesi semtinin ve İstinye Evlerinin axial line analizinde İstegrasyon-n diyagramını değerlendirdiğimizde “Pınar Mahallesi” semtinin en işlek caddesi olan Çamlıbel Caddesinin, koyu kırmızı renk sıkalasında en yüksek integrasyon değerine sahip olduğu hesaplanmakta ve Hillier’in de ortaya koyduğu gibi bölgenin bir integrasyon çekirdeği olarak belirmektedir. Buna karşın, İstinye Evlerinin duvarlarına yapışık İğde Sokağının integrasyon değerinin oldukça düşük olduğu ve gözlemlerle elde edilen verilere göre özellikle yaya kullanım yoğunluğunun çok az olduğu, ticari birimlerin yer almadığı ve duvar yazılarının yoğunlukla yer aldığı çevresiyle bütünleşemeyen bantlar olarak göze çarpmaktadır. Kapalı konut adalarında bir çok ayrıcalıkların sağlanmış ve kontrol edilebilir halde tutulması yeni yerleşim yatırımlarını ekonomik açıdan daha karlı hale getirmektedir (Görücü and Pektas 2014) ve yatırımcı firma veya girişimciler tarafından en önemli bu tercih sebebi olmaktadır. Kamu ve özel sektörün elindeki boş veya farklı fonksiyona sahip arsaların kapalı konut adaları inşaası yoluyla daha fazla kar elde edilebilir hale getirilme isteği ve ekonomik

fizibiliteler bu uygulamaların lokasyon seçiminde en önemli öncelikli unsur olmuştur. Enjekte edildikleri mevcut kent dokusunda oluşturulan adaların bitişik çevrelerinin sosyal – kültürelekonomik yapıları dikkate alınmamış, mahallelerin dokusal sürekliliğine dikkat edilmemiştir. Verilerin analizi ile ortaya konulduğu gibi, bu refah adacıkları ile mevcut doku arasındaki zonlar etkileşimsiz, yalnızlaştırılmış alanlar haline dönüştürülmüştür. Bu zonlarda doku uyuşmazlığı ortaya çıkmıştır (http://gophaber.com/haber-3847-arsiv.html). Analizlerin desteklediği gibi, uçsuz bucaksız duvarlara komşu, bitişik çevrenin rahatsızlık ve hoşnutsuzluğu grafiti olarak ortaya çıktığı gözlemlenmektedir. Gelinen bu noktadan sonra duvarların yıkılıp çok farklı sosyo-ekonomik, kültürel yapıda kesimlerin bir araya gelmesi şüphesiz ki beklenemez. Zaten böyle bir beklenti, talep taraflarda da bulunmamaktadır. Çözüm mevcut doku içindeki yeni kapalı yerleşmelerin lokasyonlarının belirlenmesinde mevcut sosyal doku faktörünün ön plana çıkarılmasıdır. Lokal uygulamalarda, ekonomik rant uğruna kapalı konut adaları üretmek yerine, mevcut bünyenin kabul edebileceği, çevre ile bütünleşebilen açık dokunun geliştirilmesi, mahalle anlayışının ön plana çıkarılmasıdır. Daha üst ölçekte ise, keskin cepheler, hatlar oluşturulmadan, daha barışkın, geçişken bir kademelenme anlayışı ile kapalı konut adacıkları uygulamalarına gerek bırakılmayacak kentsel planlama anlayışının ve uygulamasının hayata geçirilmesi olmalıdır.

ITU A|Z • Vol 12 No 3 • November 2015 • S. Apak

ITU A|Z • Vol 12 No 3 • November 2015 • 159-173

A diachronic approach on heterochronic urban space

İlgi TOPRAK1, Alper ÜNLÜ2 hacihasanoglu@itu.edu.tr • Department of Architecture, Faculty of Architecture, Istanbul Technical University, Istanbul, Turkey 2 aunlu@itu.edu.tr • Department of Architecture, Faculty of Architecture, Istanbul Technical University, Istanbul, Turkey


Received: May 2015

Final Acceptance: October 2015

Abstract As the fourth principle of heterotopias defined in Foucault’s controversial yet inspiring text named “Of Other Spaces”, heterochronies define places that accumulate time, as well as temporary spatial formations. This study interprets urban spaces with historical backgrounds, which can accommodate both the accumulation of time and the temporariness, as heterochronic urban spaces. This study aims to evaluate the reflection of socio-cultural background of historical neighbourhoods on the morphological and semantic change of their heterochronic elements throughout history. The deconstructive methodology of the study consists of a diachronic research involving three parts: deconstructing, analysing and reconstructing history. By “deconstructing history” through a multi-layered “timeline” developed with important historical thresholds and a “zoning”, morphological or socio-cultural changes and “situations” are defined. “Analysing deconstructed history” involves the syntactic analyses of these “situations” in terms of historically persistent elements and temporary formations, to grasp the morphological and socio-cultural evolution of the heterochronic urban space. “Reconstructing history”, as a synthesis, semantically interprets syntactic findings signalizing accumulations, discontinuities, shifts or losses of meaning. The case study is Kuzguncuk neighbourhood, a heterochronic urban space on the Asian side of Istanbul, an old Bosphorus village, which mainly used to consist mostly of a welcoming and peaceful residential area co-inhabited by Jews, Christians and Muslims, later by Black Sea migrants. The area is gentrified within the last thirty years. The findings of the case study show that Kuzguncuk is one of the heterochronic urban spaces, as well as a palimpsest where a majority of meanings perish, leading to fake re-valuation. Keywords Diachronic research, Heterochrony, Kuzguncuk, Space syntax.


1. Introduction Heterochronies, as in Foucault’s fourth principle of heterotopias, define multiple temporalities in a single place. Besides architectural interpretations such as libraries and museums, heterochronies can also define urban spaces in smaller or bigger scales, collecting various morphological and socio-cultural traces of time. In fact, so-called “heterotopic urban spaces” construct a perpetual time accumulation and become timeless. The first section of this paper aims to discuss how temporality is handled in Foucault’s heterotopias in “Of Other Spaces” and define what heterochrony means as a spatio-temporal notion interpreting similar and following studies mainly about architectural and urban heterochronic spaces, notably the ones of Heidegger, Sartre, Merleau-Ponty, and Rossi. This part of the work concentrates in theoretically describing “heterochronic urban space”. The second section explains the deconstructive methodology that is adopted in this study to examine heterochronic urban space in depth. The methodology consists of three parts: deconstructing, analysing and reconstructing history. This methodology aims to uncover the history by deconstructing and layering historical data of a heterochronic urban space within a “timeline”, analyse layered historical components using Space Syntax, and finally holistically reconstruct history, as today’s reality, with a synthesis of syntactic findings and their semantic interpretation. The third chapter concentrates on the case study. The case study involves the historical evolution of the Kuzguncuk neighbourhood, an old Bosphorus village on the Asian side of Istanbul using Space Syntax, diachronically analysing changing syntactic values of heterochronies: historically persistent elements as “heterochronic constants”, gatherings and “situations” as “heterochronic variables” between the years 1932 and 2014. Finally, the conclusion of the study focuses on the semantic interpretation of the current state of the neighbourhood to answer the question to what extents Kuzguncuk can be defined as

a heterochronic urban space. As the final step of the three-fold methodology, this holistic interpretation attempts to do a synthesis of the case study. 2. Heterochrony: Temporality in Foucault’s heterotopias and its translation into urban spaces Heterotopias are closely linked to concerns about time, notably time intervals, breaks, accumulations and transitions. Between the eternal and the temporary, heterotopic spaces refer to temporal formations in different contexts. Among the principles of heterotopias in Foucault’s highly controversial text “Of Other Spaces”, it is indicated that there is not a universal form of heterotopia; functions of heterotopias are variable, and that heterotopias can gather multiple incompatible spaces together. They have an opening-closing system, and are generally not freely accessible like public spaces. They always have a function relating to external spaces (Foucault, 1986). According to the fourth principle of heterotopias, which is the main concern of this paper, heterotopias linked to “slices of time” named as “heterochronies”. Heterotopias are working at full capacity in case of rupture of the traditional time for humans, as in the case of a cemetery, an intersection of loss of life and eternal rest. The first significant type of heterochrony is indefinite time accumulation as for libraries and museums. It is an endless gathering of things, accumulating and archiving in a specific place, a place that becomes itself out of time. Another type of heterotopias stands out as quite the opposite of the previous one, defining spaces as temporary structures, fugacious and finite occurrences. These heterochronies are strictly temporal and can be translated to many architectural and spatial experiences such as fairgrounds, vacation villages or Olympic villages. These two entirely opposite sides of heterochronies reflect the eternal and the temporary at the same time (Foucault, 1986). Can urban spaces really be studied as heterochronies? Mc Leod (1996) criticizes Foucault for having forsaken the “messy and in-between urban spaces” such as the residence, the

ITU A|Z • Vol 12 No 3 • November 2015 • İ. Toprak, A. Ünlü


workplace, the street and many others. The limited illustration and description of spaces depicted as heterotopias in “Of Other Spaces” seem to restrain the boundaries of the concept, referencing pre-defined spaces. However, the idea of heterochrony itself can be subject to define a double-sided reality of permanence and transience of the actual urban pattern. It can be assumed that heterochronies create a bridge between time and space. The combination of accumulation of time and fugitive experiences form them, and like in every heterotopic case, they show several inconsistencies. Heterochronic urban spaces reflect continuity and iteration, depicting history and present at the same time. Everyday experiences actualize together in urban spaces with references to the past. In that way urban space as heterochrony, can verify Foucault’s argument that urban spaces as heterochronies are sources of immediate knowledge. Like museums and libraries, they offer finite, compiled and quick information. However, the everydayness of heterochronic environments is not to be undermined. Because places such as library, a museum, a fairground or a farmers’ market welcome everyday strollers, readers, contemplators, chatters and shoppers. Two-sided experience of the urban space as heterochrony, constructs a bridge between accumulation/deletion of historical traces through time and everyday “situations” involving cultural, social changes, not in a way that a museum exposes an agglomeration of time segments, but experiential space through which different indicators of time accumulation find a place for themselves. In respect of the communities and the situations that define them, heterochronies are surrounded by discontinuities, altering and sometimes decaying meanings. The changing nature of heterochronies appears more clearly, especially with everydayness, collective experience of co-habitant communities. Otherness becomes valuable, in most of the circumstances, as each moment in history and the collective memory associated to it become a particular “situation” of its own. Many philosophers studied the spatiality of

“situation” such as Sartre and Merleau-Ponty who primarily investigated the anthropological depth and the conception of embodiment of “situation” (Ha, n.d.), according importance to personal and sociological insights of what embodies the self. However Heidegger differentiated the conception of “situation” from the general situation depicting inauthentic spatiality of the “They” (das Man) therefore, “situation” has a more socio-cultural potential (Heidegger, 1996; Ha, n.d.). Each situation has its own characteristics and cultural phenomena related to it, defining moments and slices of lived spaces. In architectural thought, “situation” might be equivalently studied with the idea of “urban artefact” (Rossi, 1982). An urban artefact can appear as a square, a building or a street that signifies a certain moment in history, in a constantly changing urban pattern (Rossi, 1982). Urban artefacts coexist and therefore form a city, which brings back the argument of the Collage City (Koetter & Rowe, 1978) suggesting a theory of urban fragmentation. Another conception, “palimpsest” that means writing, erasing and rewriting on parchment; describes a “non-contemporaneity with itself of the living present” in Derrida’s words (Derrida, 1994). Its urban interpretation “urban palimpsest” (Huyssen, 2003; Bjur&Azimzadeh, 2007) explains the multi-layered and diachronic attributes of urban environments, and especially the need to explore present pasts in order to understand past and present experiences and sensibility of time, from both historical and phenomenological viewpoints. Sometimes in urban palimpsests, urban artefacts or their cultural and sociological “situations” cannot overcome decay. At this point, it is also relevant to talk about temporal and spatial discontinuities; especially “historical discontinuities” (Teyssot, 1980) that can reflect the segmented nature of heterochronies, especially when sociological or cultural corruptions affect some “situations”, and let them become new “situations”. In historical urban areas, migration and gentrification tend to shift “situations”. In postmodern urban landscapes, historical architectural forms are amalgamated into new

A diachronic approach on heterochronic urban space


buildings, to recall a collective memory (Boyer, 1996, Mills, 2004) however in case of shifting situations, amalgams can appear fake and without identity, resulting in a decay of meaning. Time collection is the most significant issue that constructs the idea of heterochrony in urban spaces. Although Foucault’s approach seems to involve a relatively artificial time accumulation, as time accumulating heterochronies are defined as artificial collections that have been gathered from different periods in history and put in a context to which they do not belong; a majority of heterochronies in urban context has a steady and natural time accumulation. For instance, urban spaces gather different slices of time together, with buildings and street patterns that belong to various timescapes, and not at once, but in a long term, they ensure the accumulation of time in themselves. According to Rossi (1982), urban artefacts sometimes remain as they are, however sometimes they decay, and then their forms, their physical marks stay persistent. This permanence is called “locus” and it solidifies with collective memory. Locus ‘emphasizes the conditions and qualities within undifferentiated space which are necessary for understanding an urban artefact’ (Rossi, 1982). Therefore, locus witnesses many ‘situations’ and is the most interesting pathway to the exploration of an urban artefact. Rossi (1982) also points out the importance of rituals and collective nature of religious activity in the formation of historically permanent elements (monuments), as they determine the initiation of a certain religious and socio-cultural activity in an urban pattern, and provides a key to understand urban contexts. “The theory of permanences” suggests that it is incorrect to think of a persistent urban artefact as it is related to only one historical period. The dynamic nature of the city leans towards an evolutionary process that not only helps the preservation of historically permanent elements but also presents them as promoters of evolution (Rossi, 1982). This view of evolutionary process somehow contradicts with the idea of “palimpsest” which supports a more realistic viewpoint on

the change of urban pattern; given that the historically permanent elements do not always evolve but they sometimes perish as well. 3. Methodology Any urban space can accumulate time, as Doreen Massey (1995) argues, “The past of a place is as open to a multiplicity of readings as is the present”. However, historical urban spaces are more likely to host several periods of time with different social, cultural as well as morphological settings where their changes and breaks are more apparent and easier to compare than they are in newer urban environments. Historical backgrounds are suitable to uncover different fashions, interrelated slices of time and significant events, transformations, thresholds and milestones. This research aims to make a diachronic research on shifting “situations” and permanent elements of historical neighbourhoods with syntactic analysis. In order to achieve this, a three-fold deconstructive methodology is adopted: • Deconstructing history through a “timeline” • Analysing deconstructed history through “Space Syntax” • Reconstructing history through a “semantic interpretation”. The methodology is schematized in relation with the theoretical section and the case study in Figure 1, and then each step is explained in detail. 3.1. Deconstructing history through “timeline” and “zoning” Heterochronic neighbourhoods can have a chaotic and multi-layered historical background. Therefore, an ex-

Figure 1. Schematic representation of the methodology.

ITU A|Z • Vol 12 No 3 • November 2015 • İ. Toprak, A. Ünlü


tensive historical research about the history of the neighbourhood is indispensable in order to understand the temporal interrelations of important events, shifting situations, architectural and urban changes. This deconstruction will be made through a multi-layered timeline that aims to represent several aspects of time relating to the neighbourhood. Multiple temporalities are investigated with historically persistent elements, temporary formations, and socio-cultural and morphological thresholds. Therefore, the timeline is crucial in order to understand the dynamics of the current state of the neighbourhood, as many different cultural and social backgrounds are intermingled. The “timeline” concentrates on historical events, thresholds and periods (if any), maps, aerial photos, and the construction culturally significant buildings later used in the study as “heterochronic constants”. The “zoning” represents the temporary patterns of formations such as cultural gatherings or functions on a map. The zoning of residential patterns of different cultural gatherings is later used in the syntactic integration assessment of “heterochronic variables” in this study. 3.2. Analysing deconstructed history through “space syntax” A morphological-syntactical study (Griffiths, 2012) on the deconstructed history will be presented in two different methods: The first method involves the axial map analysis, in order to analyse visual interrelations and depth for comparing maps in which urban morphology changed significantly. The axial map analysis will include the comparative general intelligibility (local-global scattergram and regression analysis) and integration grid analysis of different maps referring to distinct maps throughout history, to understand the overall evolution of the syntactical pattern and intelligibility and to make inferences on the change of local/global integration levels of the area and their correlations through time. The second method will involve two different kinds of analysis: In order to understand the “situations” given in the timeline, isovist and integration analy-

ses concerning “heterochronic constants” and “heterochronic variables” will be put forward. Heterochronic constants are historically persistent elements (ex. cultural buildings, ritual buildings, monuments) and heterochronic variables are more temporary formations (ex. cultural gatherings and their housing zones) both found in urban heterochronic spaces. The first step is to detect “heterochronic constants”. According to Levy this kind of diachronic research “focuses on the role of constants, or historically persistent elements, in the fabric as the city evolves from one stage to the next. These elements play an important role in the determining the present configuration of the city” (Levy, 1999). In this study, these historically persistent elements are called “heterochronic constants”, buildings or monuments having cultural, ethnic or religious significance, associated with urban artefacts and their underlying meanings, similar to the concept of “monuments” in “The Architecture of the City” (Rossi, 1982). They signify the locus, the time accumulation, and one or several related urban artefacts throughout their entire existence. Places of rituals, as the key to understand urban context, a reference to the foundation of a city, possess a collective nature and most importantly signify an unchanging reality remaining out of time (Rossi, 1982). That is why as houses of myths and rituals, religious buildings’ time accumulation and locus would be significant to understand the alteration of their syntactic values, especially their integration on a global scale. Average depth of morphological and functional imprints on the historical area, their comparative values and mutual effects of integration/segregation degrees and sociological/cultural meanings are analysed. Isovists (area and perimeter values) are equally important to understand their impacts and visibility from surroundings, as well as their strategic location and the amount of visible area. Their syntactic significance can be measured according to the interrelated alterations of integration and isovist values. If they both rise, visually, functionally and meaningfully, those ritual spaces become more integrated in the

A diachronic approach on heterochronic urban space


area. If they both decrease, those ritual spaces become more segregated. In both circumstances, the social and cultural context and historical data will be investigated in order to understand any enhancement or decay of meaning. If integration values and isovist values alter inversely proportional, that means either the building is more integrated but less perceivable or vice versa. This shows that there may be complex issues with the building, its surroundings and its users. The historical data about the building will be explored to explain this complexity. The other isovist-based parameter that is used for measuring in order to understand the convexity of spaces is circularity. High circularity levels accentuate the shape of a circle in a space, but the centrality of the viewpoint also a determinant (Edgü et al, 2012). Therefore, in smaller urban settings, circularity is a parameter that promotes enclosure and embodiment, determining a level of perceptive inclusion. Another way to examine “situations” is to reveal more temporary formations and their evolution. These types of buildings are secondary elements constructing urban form, however their meanings remain very crucial for the society and the collective memory of citizens. In this study, they are called “heterochronic variables”. They form the everyday experiences of neighbourhood life, and according to shifting “situations”, they can change appearance, function or even disappear. Their mutual relations and patterns can alter. The first category of these less persistent formations is “gatherings”, having a rapidly changing housing pattern, as they are subject to restoration and renewal, or even abandonment, resulting from many reasons but mainly social and cultural changes in the community using them, such as migration and gentrification. This study examines cultural gatherings as changing aspects of “situations” and analyse the integration values of their settlement localizations in order to explain the social integration, daily lives and permanence in the area, comparing integration values with each other, examining every gatherings’ relation with its ritual space(s) according to integration val-

ues of respective heterochronic “variable” (gathering) and “constant” (ritual place). All syntactical analyses in this study will be made with the software “Syntax2D” developed in University of Michigan using the parameters area, perimeter, circularity to determine isovist characteristics of space, and the ease of spatial perception in heterochronic constants; and integration parameter to analyse both heterochronic constants’ and variables’ degree of adaptation in terms of integration/ segregation to the global system. 3.3. Reconstructing history through a “semantic interpretation” In the conclusion section of the study, the findings of the case study will be semantically interpreted. The aim of this interpretation is to holistically evaluate what has really happened in the heterochronic urban space. The first step is to concentrate on “situations”. Situations are discovered while creating the timeline and the zoning, or while examining heterochronic constants and variables. The idea is to retrieve useful information from syntactic findings, and to make a commentary with the aid of current and past “situations”. A more general synthesis can be retrieved from the semantic interpretation of the case study. The aim of this synthesis is not to generalise all similar cases, but to create an opening to discuss the resilience of multi-cultural heterochronic urban spaces. 4. A heterochronic case in Istanbul: history and evolution of Kuzguncuk 4.1. Kuzguncuk as a heterochronic urban space Kuzguncuk can be described as a heterochronic urban space because it involves many properties significant to the accumulation of time, as well as fleeting aspect of temporality. We can mention several heterochronic elements in the area, mainly of two different types: first, called “heterochronic constants” in this study, mainly investigate strategic places that remain intact over centuries. As ritual places witness the entire existence of a cultural gathering in an urban space, re-

ITU A|Z • Vol 12 No 3 • November 2015 • İ. Toprak, A. Ünlü


ligious buildings of Kuzguncuk (Greek and Armenian churches, a synagogue and a mosque) compose the only typology of “heterochronic constants” in this study. “Heterochronic constants” symbolize the cultural gathering spaces and ensure the existence of those gatherings, however they can lose impact with changing “situations”, and dying urban artefacts. Second, called “heterochronic variables” include the temporary characteristics and changing aspects of urban space, especially “situations” and different social and cultural gatherings deriving from those “situations”. Greeks, Armenians, Jews, Muslims, migrants and gentrifiers are some of the gatherings in Kuzguncuk throughout history, and the arrival of some of these gatherings are interrelated with new situations. For example, when gentrifiers entered the neighbourhood, “gentrification” became a new situation. 4.2. Deconstructing the history of Kuzguncuk with “timeline” and “zoning” On the Asian shore of the Bosphorus, one of the small neighbourhoods along the shoreline, Kuzguncuk was once known as a mixed community of Greeks, Jews, and Armenians (Akin, 1994). Although many sources indicate that Kuzguncuk’s name once was Chrysokeramos, many historians have not agreed this idea, as it stood for neighbourhood and church names in three adjacent Bosphorus villages Kuzguncuk, Beylerbeyi and Çengelköy (Bektaş, 1996). According to İncicyan, the name Kuzguncuk is a derivation from “Kosinitsa”, the old name of the district (İncicyan, 1976; Bektaş,1996). Kuzguncuk was a Jewish neighbourhood in the beginning, although when Jews settled is still unknown. There is a strong possibility that they settled into the area after emigrating from Spain in 1492. In 18th century, Armenians started to move into the area and started to grow their community in the 19th century (Bektaş, 1996). Starting from 18th century, mostly Jews, Greeks and few Armenians were residents of Kuzguncuk. Hagios Georgios Church was one of the churches that belonged to the Greeks in Kuzguncuk

and was constructed in 1821 on İcadiye Street (Tonguç & Yale, 2012). Another Greek Church, Hagios Pantalemion, was built in 1831 and its bell tower was added in 1890. Armenians had built the Surp Krikor Lusavoric Church in 1831, which was later rebuilt in 1861. Two synagogues were built, one in 1878 (although other sources indicate 1818) named Beth Ya’akov Synagogue, also known as the Big Synagogue on Icadiye Street, along with a smaller one in Yakup Street named Kal de Ariva Synagogue built in 1840. Finally, there are two mosques, the older one named Üryanizade Mosque with the simple building but with an interesting wooden minaret, built in 1860 on the shoreline, and the new mosque built in 1952, named Kuzguncuk Mosque. Other important landmarks are two baths: Small Bath and Mountain Bath; three fountains: İsmet Bey Fountain, İskele Fountain and Hacı Ahmed Efendi Fountain (destroyed); Kuzguncuk Pier and finally Old Police Station (destroyed). There are three cemeteries in Kuzguncuk: Jewish Cemetery, Greek Orthodox Cemetery and Nakkaş Baba Muslim Cemetery. Housing patterns in Kuzguncuk are also worth mentioning. According to Bektaş (1996) besides important mansions by the shore (yali), the housing pattern that address to people with middle income are very interesting. Greek houses commonly used timber frame and wood, while Armenians preferred masonry houses. A very special housing pattern is found in Üryanizade Street, consisting of row houses with small jetties, wooden houses with embellished façades in Simitçi Tahir Street and are some of most famous patterns in Kuzguncuk

Figure 2. Current state of Kuzguncuk neighborhood.

A diachronic approach on heterochronic urban space


(Figure 2). Non-Muslim “minorities left Istanbul in response to the frightening political climate between the 1940s and the 1960s. During this same period, rural-urban migration from Anatolian villages created a cultural shift in the old mahalle (neighbourhood)” (Mills, 2004). The main crisis was about an incident on 6th and 7th of September 1955. Greeks, Jews, and Armenians who had not left the area after this incident, moved to newer areas in Istanbul. According to Tümertekin (1997), in a research he conducted during 60’s and 70’s, Kuzguncuk’s residents mainly consist from Istanbulians with 46% of the total residents, the majority of them live closer to the shoreline, whereas migrants settle in the rear side of the area (Figure 4). Black Sea migrant community forms the majority of Kuzguncuk (İcadiye) residents today, and has since late 1930s when they migrated to larger cities. Presently, non-Muslims form a minority of the residents. (Mills, 2004). In the 1980s, with the first process of gentrification in Istanbul, Kuzguncuk become a popular place (Ergun, 2004). Many Turkish architects, poets and artists followed famous architect Cengiz Bektaş, who first bought a house in Kuzguncuk that he renovated afterwards and made a participative planning and renovation process without any charge during the following years in the area (Ergun, 2004; Uzun, 2002), making possible the first wave of gentrification in Istanbul (Ergun, 2004). The first morphological findings about Kuzguncuk show that a few buildings are present close to the shoreline and a triangular street pattern is depicted on Kauffer Map (1776), Kon-

stantin Kaminar Map (1813) as well as Moltke Map (1837). With Stolpe map (1863-1880), connections of the area with Üsküdar and İcadiye become more visible (Bektaş, 1996). Starting from the first quarter of the 1800’s, many heterochronic constants, especially religious buildings have been constructed and a majority of them survived until our days. Thresholds are related to the start of different housing developments occupied by distinct cultural groups, and important events significant to the cultural or morphological change of the area. Different maps, thresholds and heterochronic constants of Kuzguncuk and their dates can be seen on the timeline (Figure 3). Kuzguncuk has seen many changes, ruptures and historical discontinuities, that is why gatherings symbolizing different “situations” have locational importance, and their integration to the system is crucial in order to understand their subsistence. The mapping of these gatherings is made with reference to memories of a very small sample of interviewees in books and theses (Bektaş, 1996, Mills, 2004). According to the zoning, along the main street (İcadiye Street) there are mainly shops and residential area, however wealthy Turkish residents are situated along the shoreline road (Paşalimanı Avenue). There are two large residential areas, one in the centre of Kuzguncuk, where residents from all backgrounds live together; another in the outskirts of the neighbourhood towards İcadiye, mostly occupied by Black Sea migrants. Finally, Üryanizade Street is known with the start of gentrification in Kuzguncuk. The area around Üryanizade Street still welcomes many new residents as a

Figure 3. Timeline: thresholds, maps and heterochronic constants. ITU A|Z • Vol 12 No 3 • November 2015 • İ. Toprak, A. Ünlü


Figure 4. Zoning of Kuzguncuk neighborhood.

result of gentrification (Figure 4). 4.3. Analysing deconstructed history of Kuzguncuk through “space syntax” From the timeline, three maps are used for syntactic analysis. These maps and aerial photos are chosen accordingly to important dates in history and morphological and sociological changes in the area. 1932 Pervititch insurance map shows all heterochronic constants except the mosque, and all non-Muslim gatherings are present. 1966 aerial photo shows a situation after the 1955 incident when many non-Muslims left and migrants settled. Finally, 2014 map is a reflection of the present where all heterochronic constants are present and residents, migrants and gentrifiers live together. The boundaries of the analysed area are based on the Pervititch insurance map and are exactly applied to the other two sources. To understand the area globally, first axial line and grid integration analyses of the whole system are shown in Table

1. Axial line analysis shows the most integrated streets, and grid integration analysis shows the most integrated areas in the system. Intelligibility is the correlation between connectivity and integration, which permits the understanding of the global relation of space from what can be observed (Klarqvist, 1993). Looking at the intelligibility values of three maps, the r-squared value of the total system changes slightly according to the axial lines of three subsequent periods. In 1932, r-squared value is 0.96 and according to axial lines and integration values the void space towards the west is generally integrated in the system. In 1966, r-squared value slightly increases to 0.97, and the system becomes slightly more intelligible as well, and the shoreline and all heterochronic constants through it is more integrated in the system. In 2014, however, there was a decrease in the r-squared value to 0.94; the system becomes less intelligible, leaving only the intersection of the main street and the shoreline as the most integrated area in the system. According to axial line analysis, the secondary street in the east is more visually integrated than it was in the past (Table 1). Table 2 investigates the alterations of heterochronic constants and variables during the defined period. The comparative syntactic analysis of heterochronic constants is made referring to five religious buildings that are found in the spectrum of the map boundary: Hagios Panteleimon Greek Church, Hagios Georgios Greek Church, Beth Ya’akov Synagogue (Main Synagogue), Kuzguncuk Mosque and finally Surp Krikor Armenian Church. Hagios Panteleimon Greek Church has still a great imprint on the area, due to its bell tower as a special feature, although its integration values severely decrease from 851 to 405 then to 352. Its isovist area and perimeter respectively decrease in 1966, and rises again in 2014. Hagios Georgios Greek Church similarly becomes less integrated by time, however its isovist (area and perimeter) values stay steady. The reason behind this might be the decline of the main avenue’s overall integration statistics. The circularity also decreases gradually for both Greek churches, de-

A diachronic approach on heterochronic urban space


riving from the dominant linearity of the main avenue. Beth Ya’akov Synagogue entry has very close integration values to Hagios Georgios Greek Church, showing approximate fashions with it as they are neighbouring (Table 2). However Beth Ya’akov Synagogue’s isovist area and perimeter values decrease respectively from 3.674 and 620 in 1932, to 3.382 and 544 in 1966, and then increase to 3.625 and 585. Its circularity values first decreased from 104 to 87 then went up to 94. Beth Ya’akov Synagogue and Hagios Georgios Greek Church entrances seem visually hidden from the main street, and do not appear at first to everyday stroller unless they are searched for. The reason behind this fact could be that the actual religious buildings are behind the entrance walls, which separate the street from the buildings. The mosque did not exist in 1932, thus the slightly larger building in its place had an integration value of 268, which in 1966, after the construction of the mosque in 1952, rises to 293 becoming more integrated as well as the

shoreline axis, whereas in 2014 it decreases to 235, as the system’s highest integration values concentrate on the main road/shoreline intersection (Table 2). Its isovist area decreased from 2707 in 1966 to 2522 in 2014, although its perimeter increased from 499 to 608, as well as circularity rising from 91 to 146, less perceivable and less embodied as a result of the reconstruction across the road (Table 2). The most important increase in integration values are marked with Surp Krikor Armenian Church, which from 1932 to 1966 has tripled its integration values, followed by a slight decrease in 2014. Same fashions for isovist area/ perimeter/circularity values: the increase in 1966 may have been resulted from the widening of the coastline road and the demolition of the old wooden house on the coast, but both buildings (Armenian Church and Mosque) became more likely to be perceived and situated in a more embodied space (Table 2). Although according to axial line analysis (Table 1), the main (İcadiye)

Table 1. Integration, axial line analysis and intellig i ibility of Kuzguncuk.




ITU A|Z • Vol 12 No 3 • November 2015 • İ. Toprak, A. Ünlü


Table 2. Syntactic values of heterochronic constants and variables of Kuzguncuk. HETEROCHRONIC CONSTANTS 1966 2014


Integration Area Perimeter Circularity

1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

Hagios Panteleimon (GC) Hagios Georgios (GC) Synagogue Mosque Surp Krikor (AC) Hagios Panteleimon (GC) Hagios Georgios (GC) Synagogue Mosque Surp Krikor (AC) Hagios Panteleimon (GC) Hagios Georgios (GC) Synagogue Mosque Surp Krikor (AC) Hagios Panteleimon (GC) Hagios Georgios (GC) Synagogue Mosque Surp Krikor (AC)




851 604 601 268 136 5408 3612 3674 1796 1056 1014 631 620 594 361 190 110 104 196 123

405 380 386 293 442 3726 3387 3382 2709 4079 632 560 544 499 874 107 92 87 91 187

352 317 335 235 357 3996 3439 3625 2522 3839 663 524 585 608 682 108 79 94 146 116



residents +movers

Mean integration

Jews Turks Armenians Greeks Residents




191,94 490,26 216,51 240,27 284,74





Migrants Gentrifiers A diachronic approach on heterochronic urban space



street becomes highly integrated by time, isovist analysis (Table 2) shows that locally, heterochronic constants’ integration values decrease regularly, with some small exceptions showing improvement closer to 2014. The shoreline is more and more integrated over time according to axial line analysis. However strategic importance of the Armenian Church and Mosque and the alteration of the building erections around them result in the elevation of integration values around 1966, followed by a slight decrease in 2014. The increase in the integration of the axial line along the street near the Armenian Church, perpendicular to the shoreline is very important for these two buildings to become more integrated into the system. Heterochronic variables of 1932 consist of the cultural mosaic once found in Kuzguncuk that symbolized the mutual tolerance of the cultural gatherings. Jews are generally located in the rear parts of the main street, according to some witness memories (Bektaş, 1996). The mean integration of the area they lived is 191,94 a relatively low integration value, which may approve the fact that lower-income Jews settled in Kuzguncuk. The main synagogue’s integration value however was 601, one of the highest values in heterochronic constants in 1932, which may mean that Jews in Kuzguncuk attach importance to their ritual places’ location more than their homes, although the second synagogue (not included in the map) is located in a relatively segregated place. Turks living mostly along the coastline are wealthy people, their integration is highest (490,26) among others. Armenian and Greek settlements have similar integration values, 216,51 and 240,27 respectively, although Greek churches are more integrated with 851 and 604, the Armenian church is even more segregated than Armenians’ settlement mean integration, with 136. Today, Surp Krikor Armenian Church is as integrated as all other ritual places in the system, despite the reduction in the number of Armenian residents. In 1966, probably the political climate affected the integration of current residents decreasing from 284,74 to

253,77. Many non-Muslims moved and left their places to Black Sea migrants, whose locations have a mean integration of 157,09 which seems quite integrated, probably resulting from the moving decisions of non-Muslims and selling their properties to migrants below their value. In 2014 however, migrants had a lower mean integration of 90,81 and residents rise to 272,86 in consequence of the increasing economic gap between them. The integration of the mosque slightly drops in 2014 as well as the migrants’ gathering mean integration. The “gentrifiers”, more integrated than migrants but less integrated than residents with 127,70, probably prefer to rejuvenate more segregated places, but still be globally integrated and close to main street. They especially are numerous in Üryanizade Street. 5. Conclusion: Reconstructing history through a “semantic interpretation” Today, Kuzguncuk known as a socially inclusive neighbourhood can still be considered as a welcoming urban space. The axial analysis approves that the main perpendicular lines along the shoreline and the neighbourhood, create an integrated space. Through many changes and accumulations of the sociological and cultural patterns; Kuzguncuk acquires syntactical-morphological and phenomenological multi-layered structure, which can be called in this case an urban palimpsest through which several heterochronic constants are collected along with many sociological “situations” as heterochronic variables. Looking from today, Kuzguncuk’s heterotopic constants are crucial for the richness of the area, as they contribute to its positive reputation and to the idea of an inclusive neighbourhood. However, many of these heterochronic constants became less integrated in 1966, then slightly ameliorated in 2014, They show that their meanings are determined by shifting “situations”, which in this case are the desertion of non-Muslims and arrival of migrants just before 1966 resulting in the segregation; however re-appropriation and valuation of religious buildings by

ITU A|Z • Vol 12 No 3 • November 2015 • İ. Toprak, A. Ünlü


gentrifiers before 2014 resulted in the higher integration of the heterochronic constants. In 2014, the integration values of heterochronic constants become closer; objectified and clarified of underlying meanings; only their “locus” is now present. When we look at the overall picture for heterochronic constants, their syntactic values seem to make sense with changing “situations”, however heterochronic variables have internal dynamics as well, such as the segregation of migrants’ settlements relation with their identity and social belonging issues. From a wider perspective, through morphological or sociological interventions to the urban pattern, heterochrony as a notion is reinforced: heterochronic spaces have indicative traces of palimpsest environment. The definition of palimpsest is a highly sociological issue, depending on lives and lived experiences accorded with slices of time. Heterochrony has a fuzzy nature, dealing with both permanent and fleeting aspect of time and space, at this point; this study shows that even constant elements of the urban pattern result in “the loss and shift of meaning”. The loss or shift of meaning in heterochronies affect the urban fabric as a palimpsest as well, since the ageing urban palimpsest seem to accumulate a collection of memories and meanings relating to places, however in reality, a majority of these meanings perish, only their “locus” persist in heterochronic constants, giving opportunity in cases like gentrification to “fake re-valuation” of the aura of these traces amalgamated in the palimpsest, independent of lived experiences. References Akın, N. (1994). Kuzguncuk. Dünden Bugüne İstanbul Ansiklopedisi. [From Yesterday to Today: Istanbul Encyclopedia], Istanbul: Türkiye Ekonomik ve Toplumsal Tarih Vakfı, 145-146. Bektaş, C. (1996). Hoşgörünün Öteki Adı: Kuzguncuk [The Other Name for Tolerance: Kuzguncuk]. İstanbul: Tasarım Yayın Grubu. Bjur, H., Azimzadeh, M. (2007). The urban palimpsest: the interplay between the historically generated layers in urban spatial system and urban life. Paper

presented at the Sixth International Space Syntax Symposium, İstanbul. Boyer, C. (1996). The City of Collective Memory: Its Historical Imagery and Architectural Entertainments. Boston: MIT Press. Derrida, J. (1994). Specters of Marx: The State of the Debt, the Work of Mourning and the New International. Oxon: Routledge. Edgü, E., Ünlü, A., Şalgamcıoğlu, M.E., Mansouri, A. (2012). Traditional Shopping: A Syntactic Comparison of Commercial Spaces in Iran and Turkey. Paper presented at the Eighth International Space Syntax Symposium, Santiago de Chile. Ergun, N. (2004). Gentrification in Istanbul. Cities, 21(5), 391– 405. Foucault, M. (1986). Of Other Spaces. [Des Espaces Autres] transl. by Miskowiec, J., Diacritics, 16(1), 22-27. Griffiths, S. (2012). The use of space syntax in historical research: current practice and future possibilities. Paper presented at the Eighth International Space Syntax Symposium, Santiago de Chile. Ha, P. (n.d.). Heidegger’s Concept of the Spatiality of Dasein: The philosophical discourse on the localization in the global age [pdf]. Retrieved from http://www2. ipcku.kansaiu.ac. jp/~t980020/Husserl/ appliedPhenomenology/ha.pdf Heidegger, M. (1996). Being and Time: A Translation of “Sein und Zeit”. Albany: State University of New York Press. Huyssen, A. (2003). Present Pasts: Urban Palimpsests and the Politics of Memory. Stanford: Stanford California Press. İncicyan, P. G. (1976). 18. Asırda İstanbul [Istanbul in 18th Century]. İstanbul: İstanbul Enstitüsü Yayınları. Klarqvist, B. (1993). A Space Syntax Glossary. Nordisk Arkitekturforskning, 1993(2), 11-12. Koetter, F., Rowe, C. (1978). Collage City. Boston: MIT Press. Levy, A. (1999). Urban morphology and the problem of the modern urban fabric: some questions for research. Urban Morphology, 3(2), 79-85. Massey, D. (1995). Places and Their Pasts. History Workshop Journal, 39, 182-192.

A diachronic approach on heterochronic urban space


McLeod, M. (1996). Everyday and “Other” Spaces. In D.L. Coleman, E.A. Danze, C.J. Henderson (Eds.), Architecture and Feminism (pp. 1-37). New York: Princeton Architectural Press. Mills, A. (2004). Streets of Memory: The Kuzguncuk Mahalle in Cultural Practice and Imagination (Unpublished doctoral dissertation). University of Texas at Austin, Austin. Rossi, A. (1982). The Architecture of The City. Boston: MIT Press. Teyssot, G. (1980). Heterotopias and the History of Spaces. In M. Hays (Ed.),

Architecture Theory since 1968 (pp. 296-305). Cambridge: The MIT Press. Tonguç, S. E., Yale, P. (2012). İstanbul Hakkında Her Şey [All About Istanbul]. Istanbul: Boyut Publishing. Tümertekin, E. (1997). İstanbul İnsan ve Mekan [İstanbul People and Space]. İstanbul: Türkiye Ekonomik ve Toplumsal Tarih Vakfı. Uzun, C. N. (2002). Kentte Yeni bir Dönüşüm Süreci ve Yasal Düzenlemeler [A New Transformation Process in the City and Legal Organizations]. Planlama, 2002 (1), 37–44.

Heterokronik kentsel mekanların artzamanlı yaklaşımla irdelenmesi Foucault, pek çok düşünüre esin kaynağı olsa da aynı zamanda tartışma konusu olan “Başka Mekanlara Dair” adlı metninde tanımlanan heterotopya kavramının dördüncü özelliğine göre, heterokroni kavramı ile zaman biriktiren yerlerin yanısıra geçici mekanları da tanımlamaktadır. Bu çalışmada, hem zaman biriktirme hem de geçicilik özelliklerini birlikte barındıran tarihi mahalleler, heterokronik kentsel mekanlar olarak irdelenecektir. Bu çalışma sosyo-kültürel etkilerin tarihi mahalleler üzerine yansımalarını, mahallelerdeki heterokronik öğelerin morfolojik, sentaktik ve semantik değişimini inceleyerek değerlendirmeyi amaçlamaktadır. Bu amaçla, çalışmada heterokronik kentsel mekanı derinlemesine irdelemek için, yapısökümcü bir metodoloji kullanılmıştır. Metodolojik kurgu, tarihin yapısökümü, mekansal dizim ile sentaktik irdeleme ve tarihin yeniden inşası (sentez) olmak üzere üç aşamadan oluşmaktadır. Birinci evrede, heterokronik kentsel mekana ait önemli tarihi olaylar, binalar, çeşitli dönemleri yansıtan harita ve hava fotoğrafları, çok katmanlı bir zaman çizelgesi yardımıyla parçalara ayrılmıştır. Zaman çizelgesi heterokronik kentsel mekanın tarihindeki eşiklerin, morfolojik ve sosyo-kültürel değişimlerin ve bu değişimlerin arasında kalan zaman dilimlerinin belirlenmesinde rol oynamaktadır. İkinci aşamada, parçalarına ayrılmış zaman çizelgesinden yararlanarak, heterokronik bölgenin sentaktik irdelemesi gerçekleştirilmektedir. Bu sentaktik irde-

leme, heterokronik kentsel mekanın iki şekilde, yani tarihte iz bırakan kalıcı öğeler ile geçici oluşumlar üzerinden değerlendirilmesini, ve mekandaki morfolojik ve sosyo-kültürel evrimin anlaşılmasını sağlamaktadır. Tarihte iz bırakan kalıcı öğeler bu çalışma kapsamında “heterokronik sabitler” olarak değerlendirilmiştir, ve sadece farklı kültürel gruplara ait dini binalara indirgenmiştir. Morfolojik ve sosyolojik değişimleri oluşturan farklı “durumlar” (situations) ise “heterokronik değişkenler” olarak adlandırılmıştır, ve bu çalışmada farklı sosyo-kültürel toplulukların yerleşim dokuları üzerinden değerlendirilmiştir. Üçüncü aşama, “tarihin yeniden inşası”, parçalara ayrılarak sentaktik irdelemesi yapılan mekanların, zaman çizelgesinden yararlanarak, semantik bir okuma ile holistik bir şekilde yorumlanmasıdır. Sentaktik bulgular, morfolojik veya sosyolojik değişimlere, anlam kaybı, anlamsal kayma, devamsızlık ve birikmelere işaret edebilir. Bu nedenle “tarihin yeniden inşası”, çalışmada sentaktik ve semantik anlamda bir bütünlük oluşturmayı, ve heterokronik kentsel mekanların artzamanlı evriminin sentezini yapmayı hedeflemektedir. Alan araştırması, Kuzguncuk mahallesinde yapılmıştır. İstanbul’un Anadolu Yakası’nda bulunan bir boğaz köyü olan Kuzguncuk, çalışmadaki heterokronik kentsel mekan özelliklerine uyan bir mahalledir. Genel olarak, huzurlu ve yaşanılır olarak nitelendirilen mahallede, geçmişte çok sayıda Musevi, Ermeni, Rum ve Türk birlikte yaşamışlardır. Yakın geçmişte Karadeniz Bölgesi’nden çok sayıda göç

ITU A|Z • Vol 12 No 3 • November 2015 • İ. Toprak, A. Ünlü


alan mahalle, son otuz yılda ise soylulaştırma (gentrifikasyon) yolu ile sosyo-kültürel değişime uğramıştır. Alan çalışmasının sonuçlarına göre Kuzguncuk Mahallesi, bulanık bir kavram olan heterokroniyi güçlendirmektedir.

Palimpsest yapıdaki kentsel mekanda, yaşanmışlıktan ziyade, anlamsal bir birikme söz konusu olmuştur. Birçok anlam de geçicilik göstermiş, kaybolmuş veya değişmiştir.

A diachronic approach on heterochronic urban space

ITU A|Z • Vol 12 No 3 • November 2015 • 175-188

Divided shopping: A syntactic approach to consumer behaviour

Erincik EDGÜ1, Meray TALUĞ2, Nezire ÖZGECE3 erincik@gmail.com, erincikedgu@duzce.edu.tr • Department of Architecture, Faculty of Arts, Design and Architecture, Düzce University, Düzce, Turkey 2 talug.meray@gmail.com, mtalug@ciu.edu.tr • Department of Architecture, Faculty of Fine Arts, Design and Architecture, Cyprus International University, Haspolat, Lefkoşa, Northern Cyprus 3 nezireozgece@gmail.com, nozgece@ciu.edu.tr • Department of Architecture, Faculty of Fine Arts, Design and Architecture, Cyprus International University, Haspolat, Lefkoşa, Northern Cyprus


Received: May 2015 Final Acceptance: October 2015

Abstract Shopping is a socially interactive consumer activity that involves preference, selection and leisure. As historical city centres are still cores of traditional shopping and an asset improving social attraction, attractive routes and spaces for pedestrian movement provided by articulation in the setting are worth examining. Buildings on small sized plots located in a bounded environment usually encourage pedestrian flow, presenting more options of interest on a unit street scale; whereas spatial layout of the urban form, compactness of the circulation routes or visual scope of the users should also be examined. This paper focuses on the comparison of consumer shopping behaviour in such a historical city centre, Walled City of Nicosia, the capital of Cyprus Republic and TRNC. Regarding the lost centrality due to the UN buffer zone, divided city has gone through different physical and social development patterns in terms of land uses and functional changes. Assuming that physical accessibility reinforces social and economic attraction, the paper deals with the, • syntactic hints examined through line analysis underlying the physical development of the urban layout in three different periods of the city, • preferences of the pedestrians, emphasizing functional and spatial pattern that orient the consumer behaviour. The outcomes indicate that narrow long roads promote pedestrian flow in a movement based activity, while the curvy organic formed streets disperse pedestrian movement. Pedestrians tend to shop for retail based products in a linear layout, and tend to eat or drink in a dispersed organic layout. On the other hand, as an aspect of political curiosity both sides of the buffer zone also serve as attraction nodes regardless of the functions. Keywords Attraction, Historical urban core, Pedestrian behaviour, Shopping, Space syntax.


1. Introduction In terms of user satisfaction, within time, historical city centres may fail to meet the needs and expectations of the occupiers; however, Lewicka (2008) argues that historical centres create a sense of continuity with the past, embody the group traditions and facilitate place attachment, as well. That is why even though there may unpredictable population fluctuations or major factors such as wars and regime changes occur, historical city centres are still cores of attraction and they struggle to maintain their vitality. Weltevreden et al., (2005) state that historic city centres are not competitive in terms of prices but they remain attractive as a retail location only if they are able to generate new sectors offering new products. When we think of attraction in the historic city centre, shopping facility as an attractor plays a major part, but on the other hand, inner cities are also places to meet people, socialise and recreate as Weltevreden and Rietbergen (2007) discuss. They add that attractiveness of an inner city depends on four factors; these are an environment with characteristic aspects such as being historical, or having an ambiance, a concentration of a large variety of functions other than shopping such as restaurants, theatres, museums, the amount and variety of shops, and lastly the crowdedness. Among these aspects, crowdedness issue is related with the number of pedestrians. Existence of a pedestrian flow is crucial for attracting social interaction, as Hillier (1999) emphasises that attraction plays an important role in drawing people to the city centres. There are various researches which examined city centre activities through social attraction, for example, Portnov (1998) examined the social attraction in Siberian urban layout in three aspects, which were the general residential attractiveness, quality of physical environment and attractiveness for business activity. The outcomes of his research showed that perceptions of the professional specialists on spatial quality were different than of the city dwellers. While the specialists attach importance to access to city centre and recreational areas, dwellers attach

importance to ecological and functional issues such as social facilities and services. Kemperman et al. (2009) on the other hand, emphasise that attractive downtown historic centres are recognised as potential magnets for tourist shopping. Tourists prefer links/ streets that are physically attractive, having a good visibility, and are pedestrian friendly. They believe pedestrian movement is an important indicator of shopping behaviour and assume that tourist shopping behaviour is related to the motivation for shopping, the familiarity with the shopping area, and whether or not the shopping route was planned in advance. Space syntax theory introduced by Hillier and Leaman (1974) use the term syntax to refer to rules that generate different spatial arrangements, in which spaces are considered to be shaped due to certain cultural considerations and these forms in return affect social relations in one way or another. As space syntax is defined as a methodology to represent, quantify and interpret spatial configuration and visual perception of exterior or interior spaces of various scales by means of convex shapes, and axial lines, (Hillier et.al, 1987; Peponis, 2000) it is necessary to briefly explain the terms used in this sense. Integration exposes the distance to a convex shape from all points within the system. If the real or global integration value of the shape is high, it means that reaching to this certain shape from any point within the system, is relatively easy and indirect. Connectivity refers to number of cells directly connected to a shape within the system. If the shape is located somewhere close to the centre of the system, then it means that the shape has many surrounding cells, thus increasing its integration. On the other hand, if the shape is located somewhere close to the outermost parts of the system, its integration value decreases, increasing its mean depth value (Edgu, 2003). Therefore, factors such as relationship between the integration and connectivity of axial lines give ideas whether a town centre is an agreeable place with good vitality in the shopping streets (Hillier et al., 1993). Pedestrian’s perceptions on physical characteristics

ITU A|Z • Vol 12 No 3 • November 2015 • E. Edgü, M. Taluğ, N. Özgece


such as distance, building composition and urban pattern are important aspects of attraction. In the historical city centres, local’s attraction as well as the tourist attraction can be provided by a good circulation of pedestrian flow supported by facilities like recreational gathering spaces and retail functions such as eating, clothing and such shopping. Ünlü et. al. (2009) suggest that regardless of its kind, the occupancy of a space can be determined with a function attached to it, and the reason of using a space can be related with socially attractive functions such as cafeterias, small shopping units. Thus, it is important to distinguish the reason of pedestrian gathering or movement within the historic urban core. Hatz (2006) suggests that unlike retailing in US downtown areas, the historic city centres in European cities remain at the top of the retail hierarchy. By retail hierarchy he refers to the centrality of shopping streets and centres, which are defined by retail space, sales figures and catchment areas, but also by the range of goods. According to his study, middle term consumption goods such as clothing sector hold the leading position in hierarchy, and the higher the rank of a shopping district, the greater is the dominance of the clothing sector. Hatz (2006) also suggests that goods of short-term consumption indicate the opposite compared to middle term consumption goods. He emphasises that the lower ranking a shopping street or district, the higher the proportion of shops offering products of short-term consumption. On the other hand, in their research of grocery-shopping behaviour, Wang and Lo (2007) emphasised that consumption is less about economic rationality and more about cultural values and meanings, whereas location preferences are based on image and identity rather than narrow economically driven criteria. Supporting Wang and Lo, we assume that shopping is a consumer activity that involves preference, selection and leisure that allows social interaction as well. Kemperman et al. (2009) discuss that the pedestrians walk from node to node in a way that they walk forward until the next decision point which is usually on the street intersections. They

assume that the path pedestrians take, depends on the relative attractiveness, such as shopping supply and other features that may affect the attractiveness as the history of the path walked so far. Buildings on small sized plots located in a bounded environment usually encourage pedestrian flow. As Crompton and Brown (2006) indicate that small scaled places without cars may seem much larger to the walking person, than expected. Especially in complicated car-free cities with traditional architectural pattern such as Venice or Fez, tourists believe that the places felt larger than they seem on the map. With the help of entrance doors, these small sized buildings present more options of interest on a unit street scale that can be shifted to economic benefit. Kemperman et. al’s (2009) research also supports this hypothesis. They found out that tourists prefer streets/ links that are part of a long straight line offering a long view, with buildings on both sides as variations of façades. Therefore, considering the pedestrian movement, historical city centres are still cores of traditional shopping and assets which improves the social attraction of the city (Ülken & Edgü, 2005). The historical structure is an asset that improves attraction. This paper focuses on the comparison of pedestrian movement and shopping behaviour of the occupants as well as the functional differences of land use in the context of predetermined nodes of attraction located on both sides of the buffer zone in the historical Walled City of Nicosia. As it will be discussed in the following section in a more detailed manner, Walled City of Nicosia has a unique star shaped architectural layout which is unfortunately divided by UN buffer zone into two unequal halves. Therefore considering the distinction of having two different communities settled on each side, the pedestrian movement in terms of shopping also has its unique flow as well. As the research assumes that physical accessibility reinforces social and economic attraction, the paper aims to search, • the syntactic hints examined through line analysis underlying the physical development of the urban layout in three different periods

Divided shopping: A syntactic approach to consumer behaviour


of the Walled City, thus exploring the changes of accessibility through integration • spatial preferences of the pedestrians, emphasizing functional and spatial pattern that orient the consumer behaviour. 2. Case study area As an island that has been colonised by various nations throughout the history, Cyprus has encountered both the prosperity and downfall of her physical location in the midst of historic crossroads of trade and culture of the Eastern Mediterranean region. However, the conflict between the two major communities of the island, which started during the 1950’s has been dominating the political structure of the region ever since. The unresolved political situation followed by the complete division of the island and separation of Greeks and Turks after 1974, unfortunately did not help the solution of the problem. Turkish region declared a self-determinant government of Turkish Republic of Northern Cyprus (TRNC), in 1983; while Greek region was granted full membership of the EU in 2004 as Cyprus Republic (CR). Division of the island was a critical decision in terms of geography, natural resources and social structure, where the two communities remained unattached until 2003. The citizens of the Turkish region have been fatigued from the isolations and serious financial difficulties due to economic embargo, so as a manoeuvre, TRNC government opened two border gates for mutual passes. However, division of the capital city Nicosia produced even more severe outcomes. One of the harms of the buffer zone called as Green Line, occupied by the UN forces is undermining the city’s centrality.The 450 years old star shaped city walls with eleven bastions were also divided ruining the integrity of urban structure and architecture as well (Figure 1). The divided city of more than thirty years has gone through different physical and social development patterns around the divided Venetian walls, providing different functions. For example, business area has moved out of the city where the land is cheaper. As the immediate

exterior of the city walls in Turkish side are occupied by administrative and educative purposes, the Greek side is occupied by business and commercial purposes. The inner core, on the other hand remained as a retail district with small shops of clothing, food and restaurants, home supplies, and some manufacturing. The differences in the socioeconomic status of both sides affected the development and urban improvement as well. While due to lack of financial resources, Turkish side preserved the majority of historic urban layout, Greek side, renovated and transformed the historical city to some extent, such as changes in plot sizes, and vertical dimensions. On the other hand, as traffic congestion and lack of sidewalks prevented the pedestrians’ easy flow, lack of parking areas made it hard for vehicular access to the shopping zones, in both parts as well. Five years later, after the opening of initial border gates in 2008, Lokmacı Gate at the end of the Ledra Street was opened for pedestrian access therefore, though, through a weak axis, the connections beyond the Green Line was established once more. This gate working as the check point is the only civilian connection within the Walled City, and is the crucial point in pedestrian movement. As for the shopping locations within the Walled City, the division left the traditional shopping bazaar Bandabulya, few inns and arasta in Turkish side, while the expensive shopping strip remained in Greek side, excluding the plots lost to the control of UN forces. Hatz (2006) suggests that the

Figure 1. Nicosia Walled City (map from Google Earth).

ITU A|Z • Vol 12 No 3 • November 2015 • E. Edgü, M. Taluğ, N. Özgece


structure of retailing in city centres is not to be examined in terms of a shopping destination serving only local residents and customers. The city centre with its unique atmosphere becomes part of consumption as well. He also adds that, retailing in the city centre is determined by the transformation of the city into a leisure destination, in which the consumption of cultural goods and experiences become prominent. He states that transformations of downtown areas into themed shopping spaces, equipped with artistsâ&#x20AC;&#x2122; quarters, art galleries, bars and restaurants are among the aspects that help a historic city centre to maintain a consumable atmosphere. Similarly, as the shopping facilities outside of the city walls in both sides follow the market demands and trends, the atmosphere helps to preserve the social attraction within the city walls. Even though lately, promising restoration and renovation projects of significant buildings are being undertaken, emergence of workshops and depots among the residential buildings, dilapidation, and obsolescence are still crucial aspects of the urban quality in the Turkish region. 3. Methodology and syntactic analyses According to Hillier et al, (1993), even though, shops may serve as attractors for the pedestrian movement,

Figure 2. Line analysis of Nicosia Walled City before 1974.

syntactically they do not change the configuration of the urban layout. However, in order to figure out the pedestrian preferences due to shopping, we first have to present the syntactic hints examined through line analysis underlying the physical development of the urban layout in three different periods of the Walled City. As it was mentioned before, integration denotes the socio-petal, thus vital nodes and highly connected axes; therefore, exploring the changes of accessibility through the city within time shows us the shift of spatial vitality. These data acquired will then be compared with actual preferences of the pedestrians. The University of Michigan software, Syntax 2D is used in for the syntactic calculations. Syntactic properties of the case study area are analysed initially with the line analysis of the Nicosia Walled City in three phases (Figures 2, 3 and 4) before the division, during the non-contact years and after the opening of gates. These line analyses show the transformation of the city layout within a span of forty years, and also the accessibility routes that are formed through political requirements and precautions taken towards these within the recent history of a city. The syntactic properties of the nodes on the other hand, are examined through grid analyses, where each grid unit is set to be 6 x 6 metres due to the maximum public distance of interaction. Integration n levels, circularity and isovist properties are among the calculated measurements. The obtained numeric data is compared and analysed through regression analyses and Spearman correlations using SPSS. As it can be seen from the diachronic line analysis maps of Nicosia Walled City from prior to 1974, from 1974 to 2008 and after 2008, (Figures 2, 3 and 4) the mean integration n values change drastically due to the formation of buffer zone that interrupts the pedestrian or vehicular movement. It is clearly exposed in the maps that prior to 1974 the city maintains its centrality by denoting the most integrated axes located close to the centre with a mean integration n figure of 1,2320x109. In this map Ledra Street the shopping strip, seems to have a high integration level as one

Divided shopping: A syntactic approach to consumer behaviour


of the parallel streets reaching to central horizontal axis that is connecting the main roads from Paphos gate to Famagusta gate. The axis from the north Kyrenia gate on the other hand seems to have a secondary degree integration while, the central axes present a deeper structure. After the division and the settling of the UN buffer zone, the mean integration n figure drops drastically to 6,478x108. During the long thirty four years of no connection between the two regions of the city, it is seen that there were two different centres formed within the walls (Figure 4). However if we compare the situation with the previous map, we see that while the most integrated axis of north gains strength, the most integrated axis of south weakens. In both regions, this interruption was compensated by moving the vital public or commercial functions to out of the Walled City, thus abandoning the historical city centre. Finally after the reconnection of two regions with a single pedestrian passage of Ledra Street/Lokmacı Gate in 2008, we see a considerable improvement of integration n level with a figure of 8,2102x108. As this situation weakens the most integrated axis in north region, it strengthens southern axis, while reintegrating the central streets to a mild shallowness (Figure 5). We also see minor changes around the northern axis, in terms of newly built roads or opened passages that leads to mild levels of integration in northeast regions compared to the era prior to division. Crompton (2006) suggests that complexity disturbs our judgment of walking distances, Crompton and Brown (2006) discovered that the more turns, slopes, intersections, and features a walk has, the longer it appears and thus a journey will seem longer when there is more information to be observed. The Nicosia Walled City has a unique organic layout which presents exciting vistas for exploring tourists; on the other hand, it serves as a maze which prevents shortcuts. As mentioned before, comparison of consumer shopping behaviour of the occupants and the differences of land use in both sides of Walled City of Nicosia are

Figure 3. Line analysis of Nicosia Walled City between 1974 and 2008.

Figure 4. Line analysis of Nicosia Walled City after 2008.

among the main concerns of the paper. We should also keep in mind that, traditional shopping district remained in Turkish side, while the expensive and trendy shopping strip called Ledra Street remained in Greek side. In order to understand the behaviour of the occupants and the attraction activity route, observations are executed on the two determined main shopping regions, from both sides of the city.

ITU A|Z • Vol 12 No 3 • November 2015 • E. Edgü, M. Taluğ, N. Özgece


Since within the Walled City, there is currently only one connection between the divided regions, the chosen axes had to start with Ledra Street, pass the mentioned check point Lokmacı Gate, and continue to north (Turkish) side. As of coincidence, Ledra Street historically used to be the most famous shopping street in the Walled City, as well. Therefore, we initially considered examining a total of three nodes from each side; one node from each starting point that is close to check point, a middle point of some specific characteristics such as a square or an attractive function and finally an ending point before exiting the walls. The nodes were set to be virtual circles with a diameter of 30 metres, regarding the size of the largest gathering area. However, the route to out of the city walls in Turkish side disperses due to the organic layout of smaller sized plots and the pedestrian axis takes a branch like shape generating from the Lokmacı Gate, i.e. the first gathering loca-

Figure 5. Predetermined nodes of analysis.

tion from the border. Regarding this situation, and also to see the situation in the original shopping centre of the city, three nodes from the Ledra Street on Greek side, and six nodes from the Turkish side were selected for analyses and comparison (Figure 5). As it is seen in Figure 5 immediate surroundings of the selected nodes and the main axes are set to be the area of observation. The selected portion of the walled city reaches from southwest to north extending to actual central Selimiye Mosque Square in east. Within this context, Kyrenia Gate, Sarayönü (Atatürk) Square, İşbankası Node and Lokmacı Node have been chosen from northern part of the Walled City. On the other hand, original, i.e. prior to 1974, central gathering and shopping spaces of Kumarcılar Inn and Selimiye Mosque Node are also included in the analyses. Although these are surrounded by shops, nowadays, these latter nodes are frequently used as transitional spaces in order to reach main shopping axes. Analyses of nodes are executed due to pedestrian flow, functional use and syntactic properties. The pedestrian flow is analysed through observations of 15 minute video recordings. As the pedestrian flow is an important indicator of attraction, number of passers-by present implications of path selection, which in turn points out the potential shopping behaviour. Additionally shopping behaviour was also measured according to number of pedestrians with shopping bags. Thus, during the process, video recordings are counted in each node regarding the total number of pedestrians with or without shopping bags. Considering the customary preference for most of the small retail shops to be closed on Sundays and the extremely hot weather during summer months, video shooting has been done during the period between 12:00 and 14:00 on two Fridays and Saturdays in April and May. These dates are set regarding the most preferred days for shopping, the most crowded period due to lunch time, and also most pleasant weather condition for outdoor space use. Referring to Hatz’s (2006) hierarchy of middle term and short term con-

Divided shopping: A syntactic approach to consumer behaviour


sumption goods, the attractive functional uses taken into consideration are set to be the retail shops such as clothes, shoes or mobile phone like electronics and the eating-drinking spaces such as cafes, restaurants and bars. The ground floor functional uses of the buildings within the determined nodes are counted and categorized. Although counted in the total number, vacant units and public functions are disregarded in the comparison. As the actual dimensions of the shop units are excluded from the research, the number of units, thus the variations gained importance in the analyses. Visual boundaries formed by wall like dense building plots may affect ff to pedestrian perception and preference of shopping. Psychological aspect of this situation is discussed in Yönet and Yirmibeşoğlu’s (2009) gated community research, where they resemble gated communities to medieval fortress settlements, in which living behind the gates increases the fear of the unknown that is outside. Even though as we move closer to the walls of the Walled City, our visual scope widens, this theory is supported especially in the central sections of the city, where pausing and relaxing is shifted to continuous circulation. As for the syntactic aspects of the mentioned visual boundary, Ülken and Edgü (2005) point out that,

long, narrow streets possess convexity and their one-dimensional axial shape promotes movement and circulation flow, as we see this in the case of Ledra Street. On the other hand, fatter convex spaces are traditionally places that support events and occasions, such as the squares and selected nodes. Circularity analysis is an important spatial characteristic that helps to examine the compactness of the spaces or motivations that drive people to pass through certain streets. Circularity defined as the ratio of the square of the perimeter to area is one of the six geometric measures such as area, perimeter, occlusivity, variance and skewness to obtain isovist field (Benedikt, 1979; Batty, 2001). Circularity is both a measure of the shape of a space and the measure of centrality of the viewpoint within that space. In her museum analysis, Kaynar (2005) argues that deformed circularity that means lower levels of circularity motivates longer visit durations. She also adds that the movement is more distributed in areas that provide opportunities to discover new visual information. Supporting this situation, articulation in an urban setting provides attractive routes and spaces for pedestrian movement which is inevitable for historical city centre shopping. As the structural pattern gets articulated in an organic system, the envi-

Table 1. Mean syntactic values, pedestrian flow and functional values of the selected nodes.

ITU A|Z • Vol 12 No 3 • November 2015 • E. Edgü, M. Taluğ, N. Özgece


ronment tends to present surprising spaces that are appreciated especially by exploring tourists. The syntactic aspects of the research base on the comparison of spatial parameters such as integration levels, circularity and visual fields, which help us to further examine the physical structure of the predetermined nodes. The map used for these analyses is derived from the 2008 map with single connection between two regions. Therefore it is necessary to remember that after the reconnection of two regions with a single pedestrian passage, the most integrated axis in north region has weakened, while the southern one strengthened, with mild improvement on the central streets. The functional uses of streets, positions of gathering spaces, figures of pedestrian flow are also examined in the research. Table 1 indicates the syntactic values along with the environmental cues of the selected nodes from both regions. Integration n values for the Kyrenia Gate, SarayĂśnĂź and Ledra Street also comply with the line analysis results indicated in Figure 5, with the highest overall integration levels. Connectivity values are the highest Table 2. Regression analyses from the selected nodes.
















































for SarayĂśnĂź and Kyrenia Gate again parallel to the data presented in Figure 2, however, Ledra Street is lower in connectivity due to the linear nature of the path and the visual boundary of the dense buildings. While connectivity values indicate an easy accessibility through neighbouring grids, it also works reciprocally, as easily exiting to neighbouring grids. Therefore when we compare the connectivity values of these two nodes, with Ledra Street nodes in terms of the number of total pedestrian count, lower connectivity values explain the high percentage of pedestrian flow in Greek side. On the other hand, the number of total pedestrians also seems to be related with the number of shops in both sides excluding restaurants and cafes. This result also complies with Kemperman et. alâ&#x20AC;&#x2122;s (2009) findings that the touristsâ&#x20AC;&#x2122; shopping route choice behaviour is affected by the supply and accessibility of shops, however tourists do not prefer links with restaurants, lunchrooms, bars and such as these necessitate pausing and lingering for a longer period compared to retail shops. The most crowded pedestrian movement with or without shopping bags in Turkish side is observed in LokmacÄą node, followed by Selimiye node. As LokmacÄą Gate being the starting node from the check point, with relatively large number of shop units this crowdedness can be explained. In case of Selimiye however, the pedestrian flow is owed to the centrality of the location within the city along with existence of traditional covered bazaar Bandabulya. In both cases however, we see a striking fact that shopping with bags concentrates mostly on the central core of the Walled City. In both cases frequency of shopping bags decreases as we move towards the north and south exits. Comparing the percentage of shopping bags respecting the total frequency of the pedestrians, Ledra 1 node with 21.02% followed closely by Ledra 3 with 19.48% are the highest shopping nodes of Greek side. Similar numbers are also seen in Turkish side as well; IĹ&#x;bankasÄą node with 21.97% has the highest value while the second largest number is seen at KumarcÄąlar node with 20.22%.

Divided shopping: A syntactic approach to consumer behaviour


Ledra Street on the other hand, presents striking number of pedestrians compared to Turkish side. Although the number of shop units is similar in number of attraction points, the pedestrian flow is observed to be much greater. This situation complies with Kemperman et. al’s (2009) findings assuming that tourists prefer to circulate on streets/links that are part of a long straight line offering a long view, with buildings on both sides as variations of façades. Ledra 1 and Ledra 3 nodes have the highest values of isovist perimeters enabling a longer vision for pedestrians to perceive the direction of movement from a longer distance and move towards the crowd, i.e. attraction. In the regression analyses shown in Table 2, pedestrian flow and shop unit data were considered as dependent variables, while the syntactic values are considered as independent. Regression analysis is investigated with the R values with significance between -1 and +1. Strong correlations are shown in dark shaded cells while although not statistically significant, mild implications, which are worth noting are also shown in lighter shaded cells. The regression analyses indicated in Table 2 show the correlations of the syntactic and environmental figures of the selected nodes given in Table 1. Circularity of the nodes and circularity of the isovists from the centre of the nodes are significantly correlated with shopping behaviour, and pedestrian flow as a whole. This situation leads us to refer to Kaynar, (2005) once again, as she argues that lower levels of circularity motivates longer visit durations, which explains the shopping behaviour occurring on dispersed nodes of Kumarcılar and Selimiye. On the other hand, theory is supported by the higher levels of circularity promoting movement as in our cases of pedestrian flow in Ledra Street, while a distraction and dispersion is seen in branch like street axes in Turkish side. As for the location and attraction levels of commercial units, we see that while cafes and restaurants are placed randomly with a smaller number of units, they do not present significant outcomes. However as for retail shops, we see a strong correlation both with

isovist area and connectivity aspects. If the pedestrians are looking for certain types of items the size of the isovist area presents better scope of exploration. Therefore we see that the number of shop units increases in a more widely perceived layout. Connectivity correlations of the shops also support this finding as higher the level of connectivity higher the level of integration thus accessibility from the perspective of pedestrians, in their search for specific brand or good types. 4. Conclusion and discussion In Ledra Street of the Greek region, the shopping preferences of the pedestrians are observed to concentrate on retail rather than recreational spaces, thus indicating a conscious selection of brands to shop from in a linear street layout. In the case of Turkish region however, central nodes present highest number of pedestrian flow, significantly dropping at the end points of the axes, again in accordance with the number of shop units. However, in these regions, pedestrian flow is also parallel with the existence of squares with recreational attraction points, such as the cafes and restaurants. This result implies a preference of eating spaces over retail shopping spaces in the organic street layout. The research indicated that narrow long roads promote pedestrian flow in a movement based activity, while the curvy organic formed narrow streets disperse the pedestrian movement. As for the percentages of shopping bags, both sides of the buffer zone present similarities denoting an approximate rate of 20% of shopping for pedestrian count. As the retail shop units increase, the percentage of shopping as expected, also increases. If these units are located in corners of narrow street nodes, the shopping percentage also increases. However if the shop units are located on square like spaces with number of cafes and restaurants, then the shopping tendency decreases, which can be explained by losing attention or distraction of perception. The results of the research emphasise that the division of the Walled City has an immense impact on the perception of the pedestrians, which

ITU A|Z • Vol 12 No 3 • November 2015 • E. Edgü, M. Taluğ, N. Özgece


leads to the dispersion of the attraction points. It is observed that the psychological effects of the war and division on the occupants from both regions are still valid, thus this situation leads to a drawback from the buffer zone, towards more secure areas within the city. Referring back to Yönet and Yirmibeşoğlu (2009), outside the walls of the city seems to be more secure for both communities and after long years of fearing whatever is on the other side of the buffer zone has yet to be dealt with. However, we also observed that the existence of the buffer zone also serves as a type of political attractor, displaying a large amount of mobility at both sides of the check point. Kemperman et al.’s (2009) assumption of the tourist movement route ending at the starting point is also apparent in this research stressing that since the only option of turning back to the entrance is through Ledra Gate, pedestrian activity increases around the nodes closest to the check point and these nodes serve as meeting points rather than shopping preference. Pedestrian flow supports syntactic outcomes especially with isovist and circularity values. Variations of shopping spaces promotes pedestrian flow, small sized units especially serve the level of attraction in historical environments. However, actual physical interruption of any layout definitely presents a non healing setting that certainly affects the spatial preferences of the users, as seen in the case of the UN Buffer zone. References Batty, M. (2001). Exploring Isovist Fields: Space and Shape in Architectural and Urban Morphology. Environment and Planning B: Planning and Design. 28 (1), 123-150. Benedikt, M. (1979). To Take the Hold of Space: Isovists and Isovist Fields. Environment and Planning B: Planning and Design, 6(1), 47-65. Crompton, A. (2006). Perceived Distance in the City as a Function of Time, Environment and Behavior, 38(2), 173182. Crompton, A. and Brown, F. (2006). Distance Estimation in a Small-Scale Environment, Environment and Behav-

ior, 38 (5), 656-666. Edgü, E. (2003). A Syntactic Approach to Space. 1st International Symposium of Interactive Media Design, Yeditepe University, İstanbul. Hatz, G. (2006). Competition and Complementarity of Retailing in the Historic City Center of Vienna. Competition and Complementarity in Retailing, 1, 135-154. Hillier, B. (1999). Centrality as a Process: Accounting For Attraction Inequalities in Deformed Grids. Urban Design International, 4, 107-127. Hillier, B., Burdett, R., Peponis, J., Penn, A. (1987), Creating Life: or, Does Architecture Determine Anything. Architecture and Behaviour, 3(3), 233250. Hillier, B., Leaman, A. (1974), how is Design Possible?. Journal of Architectural Research and Teaching, 3, 4-11. Hillier, B., Penn, A., Hanson, J., Grajewski, T. and Xu, J. (1993). Natural Movement: or Configuration and Attraction in Urban Pedestrian Movement. Environment & Planning B: Planning & Design, 20, 29-66. Kaynar, İ. (2005). Visibility Movement Paths and Preferences in Open Plan Museums: An Observational and Descriptive Study of the Ann Arbor Hands-on Museum. 5th International Space Syntax Symposium, Delft, the Netherlands. Kemperman, A.D.A.M., Borgers, A.W.J., Timmermans, H.J.P. (2009). Tourist Shopping Behavior in a Historic Downtown Area. Tourism Management, 30(2), 208-218. Lewicka, M. (2008). Place Attachment, Place Identity, and Place Memory: Restoring the Forgotten City Past. Journal of Environmental Psychology, 28, 209–23. Peponis, J. (2000). A Syntactic Approach to Space. Rertived from http:// undertow.arch.gatech.edu/homepages/3sss/ introduction text for the 3rd International Symposium on Space Syntax/ Georgia Tech Atlanta Portnov, B.A. (1998). Social Attractiveness of the Urban Physical Environment: Cities of Siberia. Annals of Regional Science, 32 (4), 525-548. Ülken, G. and Edgü, E. (2005). Social Dynamics of Urban Transformation. 5th International Space Syntax

Divided shopping: A syntactic approach to consumer behaviour


Symposium Proceedings, 2, 669-679. Ünlü, A., Edgü, E., Cimşit, F., Şalgamcıoğlu, M E., Garip, E., and Mansouri, A., (2009). Interface of Indoor and Outdoor Spaces in Buildings: A Syntactic Comparison of Architectural Schools in Istanbul, 7th International Space Syntax Symposium Proceedings, 132. Wang, L. and Lo, L. (2007). Immigrant Grocery-Shopping Behavior: Ethnic Identity versus Accessibility. Environment and Planning A, 39 (3), 684-699. Weltevreden, J.W.J., Atzema, O. and Frenken, K. (2005). Evolution in city

centre retailing: the Case of Utrecht. International Journal of Retail and Distribution Management, 33 (11), 824841. Weltevreden, J.W.J., Rietbergen, T.V., E-shopping versus City Centre Shopping: the Role of Perceived City Centre Attractiveness. Tijdschrift voor Economische en Sociale Geografie, 98, 68–85. Yönet, N.A. and Yirmibeşoğlu, F. (2009). Gated Communities in Istanbul: Security and Fear of Crime. ENHR 2009, Prague, Changing Housing Markets; Integration and Segmantat, Prague.

Bölünmüş alışveriş: Tüketici davranışlarına dizimsel bir yaklaşım Tarihî kent merkezleri, zaman içinde kullanıcı memnuniyeti açısından sakinlerinin ihtiyaç ve beklentilerini karşılamaktan uzaklaşabilir. Lewicka (2008), tarihî merkezlerin, sürekliliği sağlama ve yere bağlılığı kolaylaştırmanın yanı sıra, var olan gelenekleri somutlaştırdığını belirtir. Bu nedenle, yaşanan beklenmedik nüfus dalgalanmaları veya savaş ve yönetim değişikliği gibi, ortaya çıkabilecek bazı temel etmenlere rağmen, tarihî kentler hâlâ çekim merkezi olmaya devam etmekte ve özellikle perakende arzı açısından ekonomik ve kültürel canlılıklarını sürdürmektedirler. Weltevreden ve Rietbergen’in (2007) belirttiği gibi tarihî kent merkezlerine olan ilgiyi arttırmada, alışveriş önemli rol oynar; kent merkezleri aynı zamanda insanların buluştukları, sosyalleştikleri ve eğlendikleri mekânlardır. Kent merkezlerinin birer cazibe merkezine dönüşmesinde dört ana faktör önemlidir; bunlar, karakteristik bir çevreye sahip olmak, alışveriş dışında, müze, tiyatro, restoran gibi farklı işlevler sunabilmek, dükkânların sayısı ve çeşitliliği ile kalabalıklık olarak sıralanmaktadır. Kalabalıklık ile sosyal etkileşimin artması, yaya akışının sağlanabilmesine bağlıdır. Kent merkezlerindeki eylemleri sosyal cazibe üzerinden değerlendiren çeşitli çalışmalar bulunmaktadır; bunlar arasında, Kemperman ve diğerlerinin (2009) çalışmasında tarihî

kent merkezlerinin turistik alışverişler için çekim merkezi olduğu vurgulanmaktadır. Bu çalışmaya göre, turistler fiziksel açıdan çekici, geniş görüş alanına sahip, yaya dostu sokakları tercih ederler; yaya hareketi ve çevreye aşinalık alışverişe özendirir. Hillier ve Leaman’a (1974) göre mekânsal dizim belirli sosyal ve kültürel varsayımlara göre şekillenen farklı mekânsal düzenler üretmek için kullanılabilecek kurallar dizisidir. Çeşitli ölçeklerdeki iç ve dış mekân biçimlenişleri ve görsel algıları dışbükey şekiller ve aks çizgileri ile yorumlanabilmektedir (Hillier ve diğ., 1987; Peponis, 2000). Bütünleşme değeri dışbükey bir şekil içinde yer alan tüm noktaların, o sistem içinde birbirlerine olan uzaklıkları ile ilgilidir. Buna göre, eğer bir noktanın gerçek bütünleşme değeri yüksek ise, diğer bütün noktalardan o noktaya erişim kolay ve dolaysızdır. Bağlaşıklık değeri hücrelere doğrudan bağlı olan komşu hücrelerin sayısıdır. Bu nedenle eğer hücre sistemin ortalarına yakınsa etrafında sayıca daha fazla hücre bulunur ve bütünleşme değeri artar. Diğer yandan hücrenin sistemin çeperlerine yakın olması bütünleşme değerini azaltır ve ortalama derinlik değerini arttırır. Bu nedenle yol olarak tanımlayabileceğimiz aksların bütünleşme ve bağlaşıklık değerleri arasındaki ilişki herhangi bir mekânın yaya hareketine uygun olup olmadığını da belirler. Kent dokusu, binaların kompozisyonu ve kent içindeki mesafeler gibi

ITU A|Z • Vol 12 No 3 • November 2015 • E. Edgü, M. Taluğ, N. Özgece


fiziksel özelliklerin yayalar tarafından algılanması, kentin cazibesini arttıran önemli etkenlerdendir. Tarihî kent merkezlerinde, eğlence mekânları ve butik, restoran gibi ticarî alanlarla desteklenen ve yaya akışını özendiren iyi bir dolaşımın bulunması yerli halk ve turistlerin ilgisini canlı tutar. Yayaların seçmiş oldukları dolaşım yolları, bu yolların göreceli olarak ne kadar cezbedici oldukları ve üzerlerinde yer alan kısa ve orta vadeli tüketim ürünlerini barındıran işlevlerle ve bunlara bağlı alışkanlıklarla da ilgilidir (Hatz, 2006; Wang ve Lo, 2007). Sınırlı bir çevrede, küçük parseller üzerine yerleşmiş olan binaların oluşturduğu tanımlı sokaklar da yaya akışını özendirmektedir. Crampton ve Brown (2006) küçük ölçekli ve araç trafiğinin olmadığı yerlerin, yaya trafiğinde artış yarattığını, yaya olarak dolaşılan bu alanın turist tarafından olduğundan daha geniş olarak algılandığını belirtirler. Dolaşımın rahat sağlandığı bu yerlerde, bir sokak üzerine dizilmiş olan küçük ölçekli binaların giriş kapıları, oluşturdukları farklı seçeneklerle ekonomik fayda da sağlayabilmektedirler. Kemperman ve diğerlerinin 2009’da yaptığı çalışmada, turistlerin iki tarafı binalarla tanımlanmış ve uzun görüş açısı sağlayan sokakları tercih ettikleri belirlenmiştir. Bu nedenle, yaya hareketleri göz önünde tutulduğunda, tarihî kent merkezleri, birer sosyal çekim noktası olarak halen geleneksel alışverişin odağıdırlar (Ülken ve Edgü, 2005). Kentin tarihî yapısının cazibeyi arttıran bir unsur olduğunun varsayıldığı bu araştırma, yaya hareketleri ve alışveriş davranışlarına odaklanmaktadır. Bu çalışma, hem Kıbrıs Cumhuriyeti’nin, hem de Kuzey Kıbrıs Türk Cumhuriyeti’nin başkenti olan Lefkoşa’da Suriçi olarak bilinen, tarihî kent merkezinde, tüketici alışveriş davranışlarına karşılaştırmalı bir yaklaşımı benimsemiştir. Birleşmiş Milletler tarafından oluşturulan ve Yeşil Hat olarak bilinen askerî tampon bölge ile eşit olmayan iki yarıya bölünen onbir burçlu yıldız biçimli tarihî kentte merkezîlik kaybolurken, buna bağlı olarak fiziksel ve sosyal gelişim örüntüleri değişmiştir. Örneğin, tarihî şehir içindeki iş alanları, arazi fiyatlarının daha ucuz olduğu şehir dışı-

na kaymıştır. Bu bağlamda, Türk tarafında sur duvarlarının hemen dışında yönetim ve eğitim amaçlı yapılaşma, Rum tarafında ise daha çok iş ve ticarî amaçlı bir yapılaşma gözlenmektedir. Buna karşılık, sur duvarlarının içindeki tarihî bölge, küçük butikler, yemek yeme alanları, imalâthaneler ve hediyelik eşya satan küçük dükkânların olduğu ticarî ağırlıklı bir bölgeye dönüşmüştür. Ekonomik tercihler Rum tarafındaki tarihî örüntünün yüksek katlı binalarla yer değiştirmesine neden olurken Türk tarafında yenilenme sınırlı kalmış ve özgünlük korunabilmiştir. Fiziksel ulaşılabilirliğin sosyal ve ekonomik cazibeyi güçlendirdiğinin varsayıldığı bu araştırma, iki konu üzerinde yoğunlaşmaktadır: • çizgi analizi yöntemi ile kentin üç farklı dönemdeki fiziksel örüntüsünün gelişimini ortaya koyan dizimsel ipuçlarının belirlenmesi, • tüketici davranışlarını yönlendiren işlevsel ve mekânsal örüntüleri vurgulayan yaya tercihlerinin ortaya konması. Alan çalışmasında Lefkoşa tarihî kent merkezinin 1974 öncesi, 1974 ile karşılıklı geçişlerin başladığı 2008’e kadar olan iletişimsiz dönem ve 2008 sonrası açılan tek yaya gümrük kapısı (Ledra Caddesi’ne bağlanan Lokmacı kapısı) ile yeniden kurulan kuzey-güney bağlantısının incelendiği haritalarda öncelikle gerçek bütünleşme değerlerine bakılmış ve kentin özgün merkezî düzeninin geçirdiği evreler yorumlanmıştır. Diğer yandan, alışveriş davranışını belirlemek üzere kentin mevcut durumunda gözlemler yapılmış, toplanma, geçiş ve alışverişe dayalı işlevsel kullanım alanları sosyal uzaklığa bağlı olarak nisan ve mayıs aylarında iki ayrı Cuma ve Cumartesi günlerinde elde edilen onbeşer dakikalık video kayıtları ile belirlenmiştir, yaya akışı ve alışveriş torbaları sayılmıştır. Bölünmeyle birlikte tarihî kentin Bandabulya olarak bilinen geleneksel pazaryeri, bazı hanları ve arastası Türk tarafında kalırken, pahalı dükkânların yer aldığı alışveriş caddesi Rum tarafında kalmıştır. Her iki bölgede de suriçi dışında kalan alanda pazar talepleri ve güncel eğilimler baskınken, suriçinin tarihî atmosferinin cazibeyi

Divided shopping: A syntactic approach to consumer behaviour


korumaya yardımcı olduğu görülmüştür. Lefkoşa Suriçi’nin 1974 öncesi, 1974-2008 arası ve 2008 sonrası artzamanlı çizgi haritalarının ortalama bütünleşme değerleri incelenmiştir. Bölünmeden önceki süreçte kentin merkezîliğinin vurgulandığı, bölünmeyle birlikte iki farklı merkez oluştuğu, güney daha ılımlı bir bütünleşme gösterirken, kuzeyde Girne kapısı aksının belirginleştiği, yeniden bağlantı sağlandıktan sonra ise kuzey aksının zayıfladığı ancak devamındaki Ledra aksının güçlendiği görülmektedir. 2008 sonrasındaki alışveriş davranışını belirlemek için Rum tarafından üç adet, dağınık organik yapısı nedeniyle Türk tarafından ise altı adet 30 m çaplı düğüm noktası analiz edilerek karşılaştırılmıştır. Zemin katlarda yer alan yiyecek, giyim, elektronik eşya türündeki perakende dükkânlar bağlamında

yaya ve alışveriş hareketi incelenmiştir. Bütünleşme değeri, döngüsellik, bağlaşıklık ve eşgörüş analizi gibi dizimsel bulgular, Rum tarafındaki uzun ve dar sokakların yaya akışını hareket odaklı eylemlere teşvik ederken, Türk tarafındaki kıvrımlı ve organik biçimli dar yolların yaya akışını dağıttığını göstermiştir. Alışveriş torbalarının sayısı merkezde yoğunlaşırken, sur dışına doğru azalmaktadır. Bunun yanında, tüketicilerin çizgisel sokak düzeninde bağlaşıklık ve bütünleşme değerlerinin de desteklediği gibi marka odaklı perakende tüketim eğilimleri oluşurken, dağınık organik bir düzende ise daha çok yeme-içme odaklı aktiviteye yöneldikleri belirlenmiştir. Bununla birlikte, tampon bölgenin politik olarak merak uyandıran iki tarafı da, işlevsel kullanımdan bağımsız olarak, öne çıkan çekim ve buluşma noktaları haline gelmektedir.

ITU A|Z • Vol 12 No 3 • November 2015 • E. Edgü, M. Taluğ, N. Özgece

ITU A|Z • Vol 12 No 3 • November 2015 • 189-207

Modeling walkability: The effects of street design, street-network configuration and land-use on pedestrian movement Ayşe ÖZBİL1, Demet YEŞİLTEPE2, Görsev ARGIN3 ayse.ozbil@ozyegin.edu.tr • Department of Interior Architecture and Environmental Design, Faculty of Architecture and Design, Ozyegin University, Istanbul, Turkey 2 demetyesiltepe@gmail.com • Department of Urban and Regional Planning, Faculty of Architecture, Istanbul Technical University, Istanbul, Turkey 3 gorsevargin@gmail.com • Department of Urban and Regional Planning, Faculty of Architecture, Istanbul Technical University, Istanbul, Turkey


Received: May 2015

Final Acceptance: October 2015

Abstract This study explores the relative association of street design –local qualities of street environment–, street network configuration –spatial structure of the urban grid–, and land use patterns with the distribution of pedestrian flows. The aim is to better understand the extent to which systematically measured street-level urban design qualities and objectively measured street network configuration are related to pedestrian movement, controlling for land use. 20 2kmx2km areas in Istanbul were studied in order to establish correlations between street design, street configuration and densities of pedestrian movement. Pedestrian data were collected on selected road segments within the areas. Same road segments were characterized through detailed field-surveys in terms of aesthetic qualities, signage, sidewalk design, pedestrian crossings/traffic lights, ground floor uses as well as GIS-based hosing plot-level (parcel-level) land use density and street-level topography. Street network configurations within the areas were evaluated using angular segment analysis (Integration and Choice) as well as two segment-based connectivity measures (Metric and Directional Reach). Linear models were developed to investigate the relationships among street design, street network configuration, land use, and walking behavior. This study contributes to the literature by offering insights into the comparative roles of urban design qualities of the street environment and street network layout on pedestrian movement. Preliminary findings imply that notwithstanding the significance of certain aspects of the street environment that relate to local urban design qualities, the overall spatial configuration of street network may prove to be a significant variable for the description and modulation of pedestrian movement. Keywords Istanbul, Land use, Pedestrian movement, Street design, Street network configuration .


1. Introduction Creating walkable urban environments have implications for public health and environmental welfare as well as urban sustainability (US Department of Health and Human Services, 1996), but intervention strategies need to be built through empirical research that identifies correlates of walking behavior (Sallis, Owen, & Fotheringham, 2000). While socio-demographic (i.e. ethnicity, income, age) correlates of walking have been widely probed in the literature (Sallis et al., 2000; Timperio et al., 2006), physical environmental variables have been studied with much less rigor. However; the limited number of studies on the link between the built environment and physical activity demonstrate that physical environmental variables are significantly associated with walking behavior controlling for socio-demographic factors (Giles-Corti & Donovan, 2002). 1.1. Street design, urban form and pedestrian movement Researches in health and urban design investigating the environmental correlates of walking have sufficiently documented associations between street-level design and pedestrian activity. The majority of emphasis is placed on the qualities of urban design, treated with reference to the immediate condition of individual streets. The local correlates of the street environment used in empirical studies range from the dimensions and design of sidewalks to the frontages of retail or the prevailing levels of environmental comfort that may encourage pedestrian movement (Badland & Schofield, 2005; R. Ewing, Brownson, & Berrigan, 2006; R. Ewing & Handy, 2009; Gehl, Kaefer, & Reigstad, 2006). Pedestrian safety, of course, is also shown to be a major factor in determining physical activity levels (Boarnet, Anderson, Day, McMillan, & Alfonzo, 2005). Safe and pleasant conditions encourage walking (Brown, Werner, Amburgey, & Szalay, 2007; C. Brown, Jones, & Braithwaite, 2007). The presence of street crossings, attractive landscaping, tree covers, and signalization (Agrawal, Schlossberg, & Irvin, 2008; Cao, Mokhtarian, & Handy,

2007), as well as aesthetic or safety features, such as cleanliness, interesting sights, and architecture (Appleyard, 1982; Gehl, 2011), have been shown to encourage walking in adults and children. In a literature review study in the health and behavioral sciences, Humpel, Owen, and Leslie (2002) concluded that accessibility to recreational facilities, opportunities for physical activity, and aesthetic attributes were consistently and significantly related to physical activity, while weather and safety attributes were less consistently associated with the behavior. Evaluating such local urban design attributes is clearly important in creating environments supportive of walking. However; walking is a context-dependent activity that requires navigating through spaces, not in spaces. Thus, it cannot be fully explained based on the local qualities of the individual street isolated from its surroundings. Any type of walking (exploratory or directed) requires pedestrians to explore perceptually available connections or exploit available connections that have been cognitively registered. Researchers in transportation and planning, on the other hand, have focused on urban form aspects of walkability, characterized in terms of proximity (distance) and connectivity (directness of traveled route) (Frank, 2000), to uncover their associations with pedestrian movement. Proximity relates to the distance between trip origins and destinations. Proximity is measured by two urban form variables. The first is density, or compactness of land uses. Density is thought to shape pedestrian activity by bringing numerous activities closer together, thus increasing their accessibility from trip origins (Cervero & Kockelman, 1997; Krizek, 2003). It is suggested that people are willing to use slower modes of travel, such as walking, for shorter distances, especially if many trips can be chained (Frank & Pivo, 1994; Marshall & Grady, 2005). The second component of proximity is land use mix, or the distance between or intermingling among different types of land uses, such as residential and commercial uses. Similarly, land use mix increases

ITU A|Z • Vol 12 No 3 • November 2015 • A. Özbil, D. Yeşiltepe, G. Argın


accessibility by increasing the number of available destinations within walking range. It is argued that commingling of offices, shops, restaurants, residences and other activities influences the decisions to walk by making it more convenient to walk to shops or to get to work (Cervero, 2002; Rodri guez & Joo, 2004) while having destinations within walking distance from origins (homes, stations, schools, etc.) increases the odds of walking (Frank & Engleke, 2000; Handy & Clifton, 2001). Whereas proximity considers airline (crowfly) distances between origins and destinations, connectivity characterizes the directness of travel between households, shops and places of employment, and the number of alternative route choices within street network (Saelens, Sallis, & Frank, 2003). The connectivity of street networks increases accessibility in two ways. First, it makes it more likely that a short or more direct route is available for any given pair of origin and destination. Second, the more the length of streets in a given area, the greater the number of frontages, and thus of destinations, that are likely to be available at walking range. Potentiality, defined as the availability of accessible streets and destinations offered by the urban fabric, is significantly related to pedestrian travel. Destinations are certainly an aspect of land use, but their number is generally proportional to the street length accessible within a walking distance. Fine-grained urban networks of densely interconnected streets improve transit and pedestrian travel by providing relatively direct routes, thus reducing the distance between origins and destinations. Prevalent measures of connectivity within the literature have been limited to average measures of street networks, such as block length (Cervero & Kockelman, 1997), block size (Hess, Paul M.; Muodon, Anne V.; Snyder, Mary C.; Stanilov, 1999; Song, 2003), intersection density (Cervero & Radisch, 1995; Reilly & Landis, 2002), percent four way intersections (M. Boarnet & Sarmiento, 1998; Cervero & Kockelman, 1997), street density (S. Handy, 1996; Matley, Goldman, & Fineman, 2001), connected intersection ratio

(Song, 2003), and link node ratio (Ewing, 1996). Apart from average measures of street density, some studies have investigated the underlying differences of street types, such as the distinctions between traditional vs. suburban and grid vs. cul-de-sac, to show a statistically significant relationship between street design with a grid-like geometry and increased frequency of walking trips (Greenwald & Boarnet, 2001; S. L. Handy, 1992; Rajamani, Handy, Knaap & Song, 2003; Shriver, 1997). However; the foregoing findings underline the multi-collinearity between such measures, hence the ambiguity of specific recommendations with regard to street network design. A number of studies have attempted to improve the explanatory power of street network design by developing composite variables that account for multiple dimensions of urban form, such as the “Pedestrian Environmental Factor” (Parsons Brinkerhoff Quade and Douglas Inc. et al. 1993) or walkability index” (Goldberg et al., 2007). 1.2. Spatial configuration and pedestrian movement While most of these studies show positive associations between measures of connectivity and walking, recent papers point out that many of these positive associations are weak, even when statistically significant (Handy, 2005; Oakes, Forsyth, & Schmitz, 2007; Rodríguez, Aytur, Forsyth, Oakes, & Clifton, 2008). One reason is the absence of measures that can systematically characterize the spatial structures of urban street networks at various scales and hierarchies. The significance of spatial structure in affecting pedestrian movement has been addressed through the framework of configurational analysis of space syntax. The methodology of space syntax involves measuring the accessibility of all parts of a network under consideration from each individual street element. The intent is to provide a generalized description of spatial structure and connectivity hierarchy without evoking information about land use or making assumptions about desirable or typical trips. In the case of space syntax, particular attention is given to the number of direc-

Modeling walkability: The effects of street design, street-network configuration and land-use on pedestrian movement


tion changes that are needed in order to move from one location to another. The claim that the ordering of connectivity, measured by direction changes, plays an important role in determining the distribution of movement is consistent with research findings in spatial cognition which suggest that direction changes, as an aspect of configuration, are related with the cognitive effort required to navigate through an area (Bailenson, Shum, & Uttal, 2000; Crowe, Averbeck, Chafee, Anderson, & Georgopoulos, 2000; B. Hillier & Iida, 2005; Jansen-Osmann, P.; Wiedenbauer, 2004; Montello, 1991; Sadalla & Magel, 1980). Earlier studies have shown that road segments that are accessible from their surroundings with fewer direction changes tend to attract higher flows (Hillier, Penn, Hanson, Grajewski, & Xu, 1993; Peponis, Ross, & Rashid, 1997). Recent research has demonstrated street network design to be significantly related to recreational (Lee & Moudon, 2006) as well as transportation walking behaviors (Ozbil & Peponis, 2012). Since walking occurs according to the fine grain of environment as well as according to its larger scale structure, appropriately discriminating measures of street connectivity are critical for designing for walkability. This study contributes to the literature by offering insights into the comparative roles of street design –local qualities of street environment–, street network configuration –spatial structure of the urban grid–, and land use patterns with the distribution of pedestrian flows. The aim is to better understand the extent to which systematically measured street-level urban design qualities and objectively measured street network configuration are related to pedestrian movement, controlling for land use. 1.3. The case of Anatolian part of Istanbul The study areas are drawn from diverse neighborhoods that vary substantially in walkability (street connectivity patterns), as well as their locations within the city (Figure 1). Kadıköy and Üsküdar are central-city districts, which include some of the

most densely walked street segments within the city. Ataşehir, which became a district in 2008, is a contemporary in-town environment with high-end residential gated-communities and office skyscrapers while Ümraniye and Kartal are peripheral districts. The underlying reason for studying the Anatolian part is due to the different urban patterns dominating each continent. The European part is mostly dominated by high-rise mass housing, service and commercial land uses, whereas the Anatolian part reflects mostly a residential character with mixed land uses prevailing the central parts. Although the selected areas represent a small cross-section of the entire city, the sum of their population equals to one-sixth of Istanbul’s total population. Table 1 presents a quantitative profile of the selected areas in terms of street patterns, population density, movement densities and land use compositions summarized based on their districts. This preliminary benchmarking demonstrates notable differences between areas. The population densities of the areas, calculated on the basis of the census blocks associated with the street segments for which pedestrian counts were taken, range from 145 to 290 per hectare with Üsküdar,

Figure 1. Locations of (a) selected districts and surveyed areas. Maps are colored based on (b) Metric Reach (800m), and (c) boundaries of the selected districts.

ITU A|Z • Vol 12 No 3 • November 2015 • A. Özbil, D. Yeşiltepe, G. Argın


Ataşehir and Kadıköy having similar densities. The mean density of moving pedestrians per 100m is 38, 29, 24, 24, and 19 for Ümraniye, Kartal, Ataşehir, Kadıköy, and Üsküdar respectively. The areas summarized in terms of their districts also differ significantly in their average street density. Average metric

reach, from high to low, is consistently in descending order from Ümraniye to Üsküdar, Ataşehir, Kadıköy and Kartal for 2km radii. However, Kadıköy has the highest two-directional reach, whereas Üsküdar and Kartal have similar lower averages. The magnitude of land use densities follows the same or-

Table 1. Urban form characteristics of selected areas summarized in terms of their districts. Ataşehir



Ümrani ye












Average population density per hectares






Average number of pedestrians per 100m






Average Metric Reach (1600 m)






Average 2-Directional Reach (20o)






Average global Integration (n)






Average global Choice (n) (in millions)






Average total residential sq mt






Average total non-residential sq mt






Average total sq mt






mixed-use entropy index






Numbers of selected areas and audited segments Number of 2kmx2km areas selected Number of segments audited Densities of residential population and pedestrians

Characteristics of street network configuration

Land use characteristics (in thousands)

Figure 2. (a) Graphic representation of observed pedestrian densities through circles of differing diameters denoting the differing densities of observed movements. (b) Location of pedestrian observations on the configurational map showing metric/directional accessibility according to Metric Reach (1600mt). Dark-to-light lines denote higher-to-lower metric accessibility within the overall Anatolian part of the city. Modeling walkability: The effects of street design, street-network configuration and land-use on pedestrian movement


der as that of street density. Ümraniye and Kartal have the highest land use intensifications while Üsküdar has the lowest total land use density. Non-residential land use density is highest in Kartal and Ümraniye, which also have relatively higher mixed-use entropy indices, and lowest in Üsküdar, which is primarily a residential district. In terms of residential building square meter, Ümraniye has the highest density while Kadıköy, Ataşehir and Üsküdar are found to have similar densities for the 2-kilometer buffer range with Kartal having the lowest density. Overall, the initial tabulation suggests a strong correspondence between the average volume of pedestrian movement and the average density of streets and land development. As higher development densities are located in areas with denser street networks, it seems plausible that the association between pedestrian density and street density is a by-product of land use. In the next section, however, the examination of the data at street segment level suggests that street connectivity has a strong role in determining the distribution of pedestrian density across and within areas.

2. Methodology This study was conducted in 5 consecutive stages. In the first stage, data on actual pedestrian volumes were recorded by conducting on-site observations in the selected areas. In the second stage, detailed field surveys were conducted to assess the degree of street-level accessibility and pedestrian quality. In the third stage, GIS-based plot-level land use compositions were measured at the street-segment scale. In the fourth stage, street network configuration of the Anatolian part of Istanbul was evaluated using various topo-geometric configurational measures. In the last stage, the associations between distribution of pedestrian movement, street design qualities, street-level land use compositions, and street network configuration were studied using linear statistical analyses. 2.1. Pedestrian observations Due to resource limitations, only 40 street segments within each 2kmx2km study area were audited. The selection of audited street segments were based on two criteria: (1) not a deadend street, (2) representative of a wide range of configurational qualities of the street network. These sampling criteria

Figure 3. (a) Land use compositions within a selected 2x2km urban area, and (b) gross densities of buildings that have their access on the individual road segment associated with each segment along the path. ITU A|Z • Vol 12 No 3 • November 2015 • A. Özbil, D. Yeşiltepe, G. Argın


ensured consistency among the sampled segments in terms of the overall pedestrian facilities. Based on these criteria, audited-segments were selected to include street-segments with differing structural levels (Figure 2). 2.2. Field surveys Street segments were characterized through detailed field surveys in terms of the pedestrian quality attributes that are shown to affect navigation in urban environments through their impacts on pedestriansâ&#x20AC;&#x2122; perceptions. These include accessibility â&#x20AC;&#x201C;sidewalk width (average width on both sides) and maintenance; street sign (presence of street name on audited segment); safety (number of pedestrian crossings and traffic-signals on audited segments relativized by street length; average width of buffer between the sidewalk and the streets on both sides as well as posted speed limit on the audited segment); aesthetics â&#x20AC;&#x201C;enclosure along sidewalks (average building setback from the sidewalk) and street trees (presence of trees on either side of the audited segment); street-front land uses (number of residential and non-residential land uses opening directly on each individual street segment relativized by street length); as well as street level topography (average degree of slope along the audited street segment). Since this study is quantitative in nature, soft-architectural-parameters, such as smell, noise and light, which are harder to quantify were not considered in the field surveys. Hence, a total number of 800 street segments were audited, and the average length of audited segments was 88.87 meters. Street segments whose total length was <%10 of average total audited segment length were excluded

from the dataset (n=10). The remaining 790 segments were included in the analysis. 2.3. Land use compositions Land use data were acquired from the 2014 GIS-based land use data provided by the Istanbul Metropolitan Municipality. Plot-based data were categorized into residential (single family and multi-family housing) and non-residential (office, retail, institution, recreation, industrial) for the purpose of distinguishing between the effects of each on the distribution of movement. Gross densities of land use density was calculated as a linear measure at the street segment scale by computing residential and non-residential building square meter associated with each individual street segment, and relativized by segment length: square meter of development per 100m of street length. Figure 3 illustrates land use compositions within a study area and demonstrates the way in which land use density was measured at the segment scale. 2.4. Configurational analyses Street network configuration of selected areas was evaluated using angular segment analysis (Integration and Choice) implemented in Depthmap10 (Turner and Friedrich, 2010-2011), as well as two parametric connectivity measures (Metric and Directional Reach) implemented in GIS. The decision to include different measures is motivated by the variety of configurational qualities (metric, geometric and topological) captured by each measure. Segment Angular Integration measures how accessible each space from all the others within the radius using the least angle measure of distance. Segment

Figure 4. Representing a study area (2kmx2km) with different configurational measures. Modeling walkability: The effects of street design, street-network configuration and land-use on pedestrian movement


Angular Choice which measures how many times a space is selected on journeys between all pairs of origins and destinations (Hillier & Iida, 2005). In other words, integration measures how easy it is to access one space (road segment) from all others in the network; whereas Choice measures how likely it is for a space to be selected moving from one space to another in the network (Hillier & Iida, 2005). These two measures represent the to and through movement potentials of the street segments (Hillier, Yang, & Turner, 2012). Choice and Integration at radii 400-, 800-, and 1600-meter were calculated. Street network configuration of the entire region was evaluated by using two parametric segment-based measures of connectivity (Peponis, Bafna, & Zhang, 2008). Metric Reach captures the density of streets and street connections accessible from each individual road segment. This is measured by the total street length accessible from each road segment moving in all possible directions up to a parametrically specified metric distance threshold. Directional Reach measures the extent to which the entire street network is accessible with few direction changes. This is measured by the street length which is accessible from each road segment without changing more than a parametrically specified number of directions. Metric Reach was computed for 1600-, 800- and 400-meter walking distance thresholds. Directional Reach was computed for two direction changes subject to a 20° angle threshold. The 20° angle threshold was selected to set the threshold low enough to make the analysis sensitive to street sinuosity. Computing directional reach for two direction changes provides an estimate of how well a street segment is embedded in its surroundings from the point of view of directional distance. In other words, it takes high values as streets become more linearly extended and as intersections to other linearly extended streets become denser. Figure 4 illustrates Integration and Choice at a radius of 800 meters, Metric Reach (800m), and 2-directional Reach (20o) respectively.

2.5. Statistical analyses Multivariate regression analyses were conducted to examine the associations between street-level urban design features, land use characteristics, and street-network configuration in explaining the distribution of pedestrian densities. The analyses were conducted in two stages. In the first stage, density of pedestrian flows was modeled for all areas considered as a single set. Street design measures were entered into the regression first to allow for the evaluation of these variables in context relative to other factors affecting pedestrian behavior. Configurational measures and land use variables were then added into the model respectively to demonstrate the effect of adding each to the model and to identify the comparative effect and significance levels of each measure. In the second stage of analyses, separate multivariate regression models were estimated for the distribution of movement densities within the individual areas summarized according to their districts. Since configurational measures computed for 800 meter radius produced higher coefficients in the analyses, these measures are reported in the following tables. Logarithmic transformation was applied to the dependent variable (pedestrian density relativized by 100m) as its distribution indicated some degree of skewness. 3. Results 3.1. Regression analyses for all areas considered as a single set Table 2 summarizes the results of regression models for 3 sets of models estimating the distribution of pedestrian densities for all areas considered as a single set. For street design measures, the most significant correlate of movement density is average sidewalk width. In fact the impact of average sidewalk width along road segments on the distribution of movement is quite consistent even when configurational and land use variables are added. The results indicate that movement densities increase with increased sidewalk width and sidewalk maintenance along the segments. Surprisingly, the signs of speed limit and the presence of

ITU A|Z • Vol 12 No 3 • November 2015 • A. Özbil, D. Yeşiltepe, G. Argın


street trees are positive and negative respectively, which is contrary to a priori expectations and earlier results. This may be due to the fact that there is not enough variability among the selected areas in terms of street speed limits (min. 10km/hour; max. 45 km/hour) and tree aligned streets. Average road segment slope is negatively and signifi-

Figure 5. Scatter plot showing the natural log of pedestrian movement densities by the multivariate regression model.

cantly associated with movement densities. Indeed the impact of street-level gradient on the distribution of movement is quite consistent across models suggesting that increased wavy topography hinders the willingness to walk. The inclusion of configurational measures, Choice and Integration at radius 800 meters as well as Metric Reach (800m) and 2-Directional Reach (20o), adds a considerable increase of 14% (p<0.001) to the explanatory power of the model. All street network measures are positive and statistically significant at the 99% level with similar standardized coefficients. However; with the addition of land use measures the overall model, there is a slight decrease in the significance levels of Integration and Choice while the standardized coefficients and significance levels of Metric and Directional Reach remain consistent. Hence; it can be concluded that the distribution of pedestrian movement is more strongly associated with metric and directional accessibility of road segments as compared to the to- and through-movement potentiality of the street. Associations between land use

Table 2. Effect tests for multivariate regressions estimating the proportion of walking for all areas considered as a single set. street design measures β

avg. slope -0.05 setback distance -0.00 avg. sidewalk width 0.00 avg. buffer width -0.00 sidewalk maintenance 0.21* street trees [no] 0.12* crosswalk existence [no] 0.15 traffic signal existence [no] -0.44 street names [no] 0.04 speed limit 0.03 Integration (800m) Choice (800m) Metric Reach (800m) 2-Directional Reach (20o) #residential land use ( ) #non-residential land use (100m) residential land use density (100m) non-residential land use density ( 790) N:


Std β

2.31* -1.67 -0.94 9.30 -1.89 2.02* 2.55* 0.40 -1.16 0.86 4.81

-0.05 -0.03 0.35 -0.06 0.07* 0.09* 0.05 -0.10 0.03 0.18

+ configurational β

-0.11 0.00 0.00 -0.00 0.25* 0.02 0.23 -0.33 0.02 0.01 0.00 0.00 0.05 0.02

t 0.12 -3.90 0.82 7.80 -1.44 2.62* 0.38 0.69 -0.96 0.58 1.94 3.13 4.05 3.90 4.40

Std β

-0.11 0.03 0.28 -0.04 0.08* 0.01 0.07 -0.08 0.02 0.07 0.16 0.16 0.16 0.14

+ land use measures β


Std β

0.00 0.00 -0.00 0.14 -0.01 0.19 -0.38 0.02 0.00 0.00* 0.00* 0.05 0.02 -0.03 0.06 0.00 0.00*

1.18 -2.99* 1.94 5.12 -0.75 1.59 -0.19 0.63 -1.23 0.58 0.05 2.59* 2.51* 3.93 3.95 -3.97 12.64 1.13 2.02*

-0.08* 0.06 0.17 -0.02 0.05 -0.01 0.06 -0.09 0.01 0.00 0.12* 0.09* 0.15 0.12 -0.11 0.38 0.03 0.05*









R adjusted

Bold: p<0.001; *: p<0.05; italic: p<0.1

Modeling walkability: The effects of street design, street-network configuration and land-use on pedestrian movement

measures and pedestrian flows are statistically significant for the objective GIS-based measures of residential and non-residential land uses (both the number of frontages and the gross densities) at the road segment scale. The most significant predictor of movement densities is the number of non-residential uses having direct access from the road segments. The results indicate that movement densities are significantly associated with both increased number of active uses and increased non-residential land use density as well as with decreased number of residential uses along road segments across areas. This suggests that increasing non-residential activities both at the ground floor level and the road segment scale and reducing residential uses would significantly increase pedestrian movement densities. Figure 5 illustrates the scatter plot showing the natural log of pedestrian densities as affected by variables in the multivariate regression model and Figure 6 shows the prediction equations for each variable in the model. Finally, street-level urban design quality attributes, configurational measures and land uses were entered together into a stepwise regression based on the forward selection method to compare each variable’s individual contribution and to identify the significant variables in explaining the distribution of pedestrian flows (Table 3). The results are similar with the previous multivariate regression models. The number of non-residential land uses associated with road segments (positive) entered the model as the most significant predictor. In fact, active frontages on the ground floor at the road segment scale alone explain 35% of the variation in movement densities. From street network configuration measures Integration within 800 meters is the most significant variable. This indicates that to-movement within urban areas is positively associated with the choice to walk. 2-Directional Reach (20o), Metric Reach (800m) and Choice at a radius of 800 meters also entered the model as significant variables along with the number of residential uses, average slope, non-residential density, setback distance, and

Figure 6. Prediction equations for the variables in the multivariate regression model.


sidewalk maintenance, but with much less contribution to the overall model. While adding Integration and av-

ITU A|Z • Vol 12 No 3 • November 2015 • A. Özbil, D. Yeşiltepe, G. Argın


Table 3. Parameter estimates for the stepwise regression model estimating the distribution of movement densities for all areas considered as a single set. R2



p value





Integration (800m)





avg. sidewalk width





#residential uses (100m)





2-Directional (20o)





#non-residential (100m)



Metric Reach (800m)





avg. slope





Choice (800m)





non-residential density (100m)





setback distance





sidewalk maintenance





residential density





traffic signal existence [no]





avg. buffer width





street names [no]





crosswalk existence





street trees [no]





speed limit







erage sidewalk width to the model results in a consequential increase in the predictive power of the model (R2 change=7-6%; p<0.001), the inclusion of the latter variables adds a modest increase of 2-1% (p<0.001) respectively. From street design variables average sidewalk width is the strongest correlate of pedestrian flows, suggesting that increasing sidewalk width along segments encourages pedestrians to walk. The signs of street design quality attributes are consistent with a priori expectations; for example, pedestrian flows increase with increased average sidewalk width, higher sidewalk maintenance, and reduced slope. 3.2. Regression analyses for individual areas summarized within their districts In order to better understand the distribution of pedestrians in each area, multivariate regression models were estimated by considering individual areas grouped according to their

districts separately. Tables 4-8 illustrate the results of regression models estimating the distribution of pedestrian densities for individual areas summarized in terms of their districts. The results suggest that the primary factors in explaining the distribution of pedestrian movement are the number of non-residential land uses at the road segment scale along with two configurational measures, Metric and 2-Directional Reach. Consistent with theory, in the analyses of individual districts movement densities are strongly associated with the number of active uses correlated with the ground floor design. The results suggest that the impact of non-residential land use on the distribution of movement is quite consistent across models. The positive coefficients of the number of non-residential uses opening directly onto the street segment suggest that movement levels increase with higher active frontages. No other significant associations are observed for other land use measures, except for Kadıköy. In Kadıköy, which is dominated by residential uses with non-residential land uses occupying the ground floors, movement densities are also sensitive to increased non-residential building square meter at the road segment scale as well as reduced number of residential uses on the ground floors. Consistent with results reporting analyses for all areas considered as a single set, results demonstrate that movement densities are strongly associated with Metric Reach (800m) and 2-Directional Reach (20o). In fact, in Ataşehir and Kartal, effect levels and significance levels indicate that Metric Reach is the main factor associated with the distribution of movement. The coefficient for the configurational variable is positive and statistically significant (at a 99 per cent level of confidence). The relationship between the distribution of pedestrian movement and directional accessibility is more pronounced in Kadıköy, which includes street segments linked more directly to their surroundings, than Kartal whose street network texture is dominated by a warped pattern. By contrast, Integration and Choice at a radius of 800 meters only appear to

Modeling walkability: The effects of street design, street-network configuration and land-use on pedestrian movement


be statistically significant for two areas (Üsküdar and Ümraniye respectively) at a 95% confidence level. This implies that pedestrian movement at the road segment scale is significantly shaped by the potentiality of a street for metric and directional accessibility rather than both for through- and to-movement. The evidence relating street design factors to walking is stronger for some of the measures. Average sidewalk width is positively and significantly associated with pedestrian flows across all models (except for Kadıköy, which has more or less a uniform standard of sidewalk width), while average road segment slope is negatively and significantly associated with movement for all areas (except for Kadıköy and Ümraniye, which have relatively smooth terrain). On the other hand, no other consistent associations are found for the rest of the street design variables. 4. Conclusions The findings of this research lend specific support for three key findings, which may have implications for urban planning and urban design decisions aimed to reduce automobile dependence and induce non-auto commuting. These will be summarized under three headings: street network configuration, street design, and land use. #1. Street network configuration is strongly associated with the distribution of pedestrian movement. The findings presented in this article confirm that the spatial structure of urban areas plays a significant role in the way movement densities of pedestrians are distributed in the city. It is shown that street network configuration, measured through syntactic measures of Integration and Choice at a radius of 800 meters as well as connectivity measures Metric Reach (800m) and 2-Directional Reach (20o), is strongly associated with movement densities when controlling for land use characteristics as well as street design attributes at the road segment scale. Linear models developed suggest that rather than the to- or through-movement potential of road segments, the density of street intersections has a greater impact on the distribution of flows. However, the results presented

Table 4. Parameter estimates for the multivariate regression model estimating the distribution of movement densities for Kadıköy. avg. slope setback distance avg. sidewalk width avg. buffer width sidewalk maintenance street trees [no] crosswalk existence [no] traffic signal existence [no] street names [no] speed limit Integration (800m) Choice (800m) Metric Reach (800m) 2-Directional Reach (20o) #residential land use (100m) #non-residential land use (100m) residential land use density (100m) non-residential land use density (100m)

β 0.04 0.00* 0.00 0.00 0.17 -0.02 -0.13* -0.11 0.02 0.01 0.00 0.00 0.03 0.02 -0.05* 0.06 0.00 0.00*

N: 238 R2 R2 adjusted

t 0.49 2.26* 1.17 0.11 0.89 -0.30 -2.01* -1.36 0.42 0.79 0.40 1.04 1.10 3.75 -3.21* 8.38 0.56 2.09*

std β 0.03 0.12* 0.09 0.01 0.05 -0.02 -0.10* -0.07 0.02 0.05 0.03 0.07 0.08 0.20 -0.19* 0.49 0.04 0.11*

0.57 0.53

Bold: p<0.001; *: p<0.05; italic: p<0.1

Table 5. Parameter estimates for the multivariate regression model estimating the distribution of movement densities for Üsküdar. avg. slope setback distance avg. sidewalk width avg. buffer width sidewalk maintenance street trees [no] crosswalk existence [no] traffic signal existence [no] street names [no] speed limit Integration (800m) Choice (800m) Metric Reach (800m) 2-Directional Reach (20o) #residential land use (100m) #non-residential land use (100m) residential land use density (100m) non-residential land use density (100m)

β -0.16* 0.00 0.00* -0.00 0.23 -0.07 -0.09 0.14 -0.09 -0.04* 0.00* -0.00 0.04 0.04* -0.02 0.12 -0.00 -0.00

N: 158 R2 R2 adjusted Bold: p<0.001; *: p<0.05; italic: p<0.1

here also underscore the significance of the spatial structure of street networks, specifically the alignment of streets and the directional distance hierarchy engendered by the street network. The fact that direction changes are as important as metric distance in describing street network configuration points to the role of cognitive factors. While Metric Reach extends uniformly along the streets surrounding a given road segment, Directional Reach may ex-

ITU A|Z • Vol 12 No 3 • November 2015 • A. Özbil, D. Yeşiltepe, G. Argın

t -2.72* 1.49 2.00* -0.15 1.05 -0.63 -0.78 0.58 -0.80 -2.08* 2.24* -0.02 1.01 2.26* -0.96 5.37 -0.62 -0.43 0.58 0.52

std β -0.17* 0.10 0.17* -0.01 0.07 -0.04 -0.05 0.06 -0.05 -0.19* 0.25* -0.00 0.11 0.25* -0.07 0.40 -0.05 -0.03


Table 6. Parameter estimates for the multivariate regression model estimating the distribution of movement densities for Ümraniye. avg. slope setback distance avg. sidewalk width avg. buffer width sidewalk maintenance street trees [no] crosswalk existence [no] traffic signal existence [no] street names [no] speed limit Integration (800m) Choice (800m) Metric Reach (800m) 2-Directional Reach (20o) #residential land use (100m) #non-residential land use (100m) residential land use density (100m) non-residential land use density (100m)

β -0.03 0.00 0.00* — -0.25 -0.02 -0.05 0.22 0.00 -0.00 0.00 0.00* 0.10 0.05* -0.02 0.05 0.00 0.00

N: 116 R2 R2 adjusted

t -0.47 0.52 2.14* — -1.10 -0.18 -0.24 0.82 0.00 -0.39 0.25 2.29* 1.78 2.47* -0.66 2.69 0.74 1.44

std β -0.03 0.04 0.18* — -0.08 -0.01 -0.02 0.07 0.00 -0.03 0.03 0.25* 0.17 0.18* -0.10 0.30 0.10 0.12

0.60 0.53

Bold: p<0.001; *: p<0.05; italic: p<0.1

Table 7. Parameter estimates for the multivariate regression model estimating the distribution of movement densities for Ataşehir. avg. slope setback distance avg. sidewalk width avg. buffer width sidewalk maintenance street trees [no] crosswalk existence [no] traffic signal existence [no] street names [no] speed limit Integration (800m) Choice (800m) Metric Reach (800m) 2-Directional Reach (20o) #residential land use (100m) #non-residential land use (100m) residential land use density (100m) non-residential land use density (N: 158) R2 R2 adjusted

β -0.11 -0.00 0.00 -0.00 -0.11 0.08 0.17 -0.16 -0.04 -0.01 -0.00 0.00 0.13 0.01 -0.02 0.06 0.00 -0.00

t -1.83 -0.10 3.49 -0.21 -0.52 0.94 0.54 -1.15 -0.59 -0.36 -0.61 1.12 3.42 0.42 -0.79 3.94 0.12 -0.06

std β -0.12 -0.01 0.28 -0.02 -0.04 0.07 0.06 -0.08 -0.04 -0.03 -0.10 0.13 0.42 0.03 -0.06 0.31 0.01 -0.00

0.50 0.43

Bold: p<0.001; *: p<0.05; italic: p<0.1

tend much less uniformly, because it is sensitive to the shape and alignment of streets, not merely to their density. This implies that street segments that constitute the primary connectivity skeleton holding together the street network is strongly associated with the way in which people navigate in urban space. Traditional models of movement patterns are based on the consideration

of distance and time, but they do not take into account the intelligibility of urban form. Integrating considerations of intelligibility can lead to enhanced models of urban form and function. The analyses presented in this paper suggest that it is possible to incorporate measures of street density and measures of cognitively significant configurational variables in the same model. However; it should be noted that the effect of spatial structure is not to determine pedestrian volume, but rather to explain how it is distributed. This is important for urban planners from the point of view of designing for urban liveliness. Space can shape land use patterns and urban densities, which are essential elements of lively cities, by affecting the distribution of pedestrian movement. To ensure urban liveliness the spatial configuration of an urban area must modulate movement densities in an economically viable manner to encourage multiple functions to occur simultaneously. #2. Higher non-residential land uses designed at the ground floor level encourage walking. The spatial structure of street network does not work independently of land use. On the contrary, based on the standardized coefficients estimated in regression models, number of active uses opening directly onto the street segment is the main driver of the distribution of flows both for all areas considered as a single set and for individual areas summarized within their districts. This supports the findings of various studies highlighting the significance of the availability of non-residential destinations nearby pedestrian-oriented nodes, such as schools and transit stations, in walking behavior (Cervero, 2002; Lee, Zhu, Yoon, & Varni, 2013). To better understand the associations between land use compositions and movement densities, street segments are categorized in terms of their ground floor designs based on the number of non-residential uses associated with each segment, as developed by Gehl et al. (2006) and Gehl (2010). Streets are classified into 4 types: active/friendly (≥10 active uses per 100m with mostly small units); mixture (6-10 active uses with a mix of large and small units); boring (2-5 ac-

Modeling walkability: The effects of street design, street-network configuration and land-use on pedestrian movement


tive uses with many blind or uninteresting units); and inactive (0-2 active uses with blind or passive units). Here active uses indicate land uses, such as retail, office and commercial establishments, dependent on passerby movement for economic viability. Figure 7 shows street scenes from four different study areas demonstrating the classification of street types based on their ground floor design. The Student’s t test (p<0.05) was used for comparing the distribution of movement densities between street types. One-way ANOVA was used for comparison of street types. As shown in Figure 8, comparisons between street types, namely active/friendly, mixture, boring, and inactive, confirm the considerable range of variation within flows at each. The coefficient of determination is 0.41 and it is significant at a 99% level of confidence. Comparison between their means also points to substantial differences in movement densities. Results of Student’s t test, presented in Table 9, indicate that there is a significant difference between each pair of types with regard to their walking shares. This analysis has consequences for at least some kinds of local urban economy: in İstanbul retail still relies on the “passing trade” because of the finer grain not only of the street network but also the size of businesses. The prevalent trait in the city, where ground floors of most residential uses function as small retail/commercial shops spread within the entire neighborhood, contributes to the higher rates of pedestrian flow encouraged by the higher street connectivity in terms of density. Here the term density refers to the amount of street which is available within a given metric range. The results also indicate that it is the number of non-residential frontages at the ground floor level rather than the gross densities of non-residential land uses at the road segment scale that matter for the distribution of movement densities within the urban fabric. It can be concluded that streets with commercial activity on the ground floor including shops, banks, cafes, and other services, are usually more stimulating to the passerby, attracting peo-

Table 8. Parameter estimates for the multivariate regression model estimating the distribution of movement densities for Kartal. avg. slope setback distance avg. sidewalk width avg. buffer width sidewalk maintenance street trees [no] crosswalk existence [no] traffic signal existence [no] street names [no] speed limit Integration (800m) Choice (800m) Metric Reach (800m) 2-Directional Reach (20o) #residential land use (100m) #non-residential land use (100m) residential land use density (100m) non-residential land use density (100m)

β 0.11 -0.00 0.00 -0.00 -0.11 0.08 0.17 -0.16 -0.04 -0.00 -0.00 0.00 0.13 0.08 -0.01 0.06 0.00 -0.00

N: 120 R2 R2 adjusted

t 1.83 -0.10 3.49 -0.21 -0.52 0.94 0.54 -1.15 -0.59 -0.36 -0.61 1.12 3.42 0.42 -0.79 3.94 0.12 -0.06

std β 0.12 -0.01 0.28 -0.02 -0.04 0.07 0.06 -0.08 -0.04 -0.03 -0.10 0.13 0.42 0.03 -0.06 0.31 0.08 -0.00

0.50 0.43

Bold: p<0.001; *: p<0.05; italic: p<0.1

Figure 7. Classification of streets based on the numbers of nonresidential uses at the ground floor level, described as (a) active/ friendly, (b) mixture, (c) boring, and (d) inactive. Developed based on Gehl et al. 2006 and Gehl 2010.

Figure 8. Variations in pedestrian densities between street types as defined by the number of non-residential ground-floor uses.

ITU A|Z • Vol 12 No 3 • November 2015 • A. Özbil, D. Yeşiltepe, G. Argın


Table 9. Results of comparisons of means of street types using Student’s t test. type



std err dif

lower CL

upper CL













































ple from immediate surroundings and further away areas. From the models it is evident that pedestrian movement is sensitive to the number of active uses opening directly onto the street both in the traditional urban areas, such as Kadıköy and Üsküdar, and contemporary in-town residential areas, such as Ataşehir, as well as peripheral districts, such as Ümraniye and Kartal. However; the higher significance and effect levels of this variable in the regression models for Kadıköy and Üsküdar indicate that since streets with commercial frontage are more evenly distributed and integrated within the surrounding urban fabric in central areas than in peripheral districts, pedestrian movement is more tuned to the ground floor-level land uses than other urban design and configurational features. Therefore, the lesson urban designers should take from these findings is that the strategic design of the ground floor at the road segment scale is critical in designing for urban vitality and sustainability; hence, is directly related to the quality of urban life. #3.The main contribution of this article, however, is the explicit consideration of street design. Findings of this study are consistent with earlier findings ar-

Figure 9. Different street sections characterizing less active space with a lower potential of generating movement (section type ‘a’) to more active spaces with a higher potential for movement densities (section types ‘b’, ‘c’, and ‘d’). Modified from Gehl (2010).

guing that sidewalk design provisions seem to be strong predictors of walking behavior (Alfonzo, Boarnet, Day, Mcmillan, & Anderson, 2008). The models developed indicate that average sidewalk width is significantly and positively associated with movement densities over and above other urban design features. In fact, the results indicate that pedestrian movement is sensitive to average sidewalk both in central city and peripheral districts. The average sidewalk width for all areas is found to be 177cm while ¼ of street segments have an average sidewalk width less than 118cm. One practical implication of this finding would be the provision of more generous sidewalks on spatially more prominent streets in the light of the association between measures of street network configuration and movement densities. Also, extensive sidewalks should be a higher priority in areas which have denser street intersections and a clearer internal hierarchy of access based on directional distances. Figure 9 illustrates various street section types with different sidewalk widths and ground floor designs to accommodate differing movement densities. As such, this study offers rules of thumbs for urban designers with regard to quality of urban life. The enhanced model of pedestrian movement demonstrated in this paper can inform specific urban design and urban master planning decisions, since it incorporates measures that can address alternative street alignments and the density of street connections simultaneously, in addition to street design attributes and land use compositions at various scales. Furthermore, the mod-

Modeling walkability: The effects of street design, street-network configuration and land-use on pedestrian movement


el can mediate between urban design and architectural design given the fact that the urban situations of buildings and land uses that can be accommodated at the ground level are sensitive to frontage and the character of the associated street. Therefore, future studies should focus on different attributes of street design, including but not limited to street widths, in conjunction with street network configuration and land use compositions both at ground floor-level and road segment-scale. Acknowledgements The authors would like to thank Dr. Fatih Terzi for his invaluable support with the analyses. This study was funded through The Scientific and Technological Research Council of Turkey (TUBITAK 113K796). References Agrawal, A. W., Schlossberg, M., & Irvin, K. (2008). How Far, by Which route and why? A Spatial Analysis of Pedestrian Preference. Journal of Urban Design, 13, 81–98. Alfonzo, M., Boarnet, M. G., Day, K., Mcmillan, T., & Anderson, C. L. (2008). The Relationship of Neighbourhood Built Environment Features and Adult Parents’ Walking. Journal of Urban Design, 13(1), 29-51. Appleyard, D. (1982). Livable Streets. University of California Press. Badland, H., & Schofield, G. (2005). Transport, Urban Design, and Physical Activity: An Evidence-Based Update. Transportation Research Part D: Transport and Environment, 10(3), 177–196. Bailenson, J. N., Shum, M. S., & Uttal, D. H. (2000). The Initial Segment Strategy: A Heuristic for Route Selection. Memory & Cognition, 28(2), 306–318. Boarnet, M. G., Anderson, C. L., Day, K., McMillan, T., & Alfonzo, M. (2005). Evaluation of the California Safe Routes to School legislation: Urban Form Changes and Children’s Active Transportation to School. American Journal of Preventive Medicine, 28, 134–140. Boarnet, M., & Sarmiento, S. (1998). Can Land-use Policy Really Affect Travel Behaviour? A Study of the Link between Non-work Travel and Land-

use Characteristics. Urban Studies, 35(7), 1155–1169. Brown, B. B., Werner, C. M., Amburgey, J. W., & Szalay, C. (2007). Walkable Route Perceptions and Physical Features: Converging Evidence for En Route Walking Experiences. Environment and Behavior, 39(1), 34–61. Cao, X., Mokhtarian, P. L., & Handy, S. L. (2007). Do Changes In Neighborhood Characteristics Lead to Changes in Travel Behavior? A Structural Equations Modeling Approach. Transportation, 34(5), 535–556. Cervero, R. (2002). Built Environments and Mode Choice: Toward a Normative Framework. Transportation Research Part D: Transport and Environment, 7(4), 265–284. Cervero, R., & Kockelman, K. (1997). Travel demand and the 3Ds: Density, diversity, and design. Transportation Research Part D: Transport and Environment, 2(3), 199–219. Cervero, R., & Radisch, C. (1995). Travel Choices in Pedestrian Versus Automobile Oriented Neighborhoods. Transport Policy, 3(3), 127–141. Crowe, D. A., Averbeck, B. B., Chafee, M. V., Anderson, J. H., & Georgopoulos, A. P. (2000). Mental Maze Solving. Journal of Cognitive Neuroscience, 12(5), 813–827. Ewing, R., Brownson, R. C., & Berrigan, D. (2006). Relationship Between Urban Sprawl and Weight of United States Youth. American Journal of Preventive Medicine, 31(6), 464–74. Ewing, R. H. (1996). Best Development Practices : Doing the Right Thing and Making Money at the Same Time. Chicago : American Planning Association. Ewing, R., & Handy, S. (2009). Measuring the Unmeasurable: Urban Design Qualities Related to Walkability. Journal of Urban Design, 14(1), 65–84. Frank, L. D. (2000). Land Use and Transportation Interaction: Implications on Public Health and Quality of Life. Journal of Planning Education and Research, 20(1), 6–22. Frank, L. D., & Engleke, P. (2000). How Land Use and Transportation Systems Impact Public Health : A Literature Review of the Relationship Between Physical Activity and Built Form. Environments. Atlanta, GA.

ITU A|Z • Vol 12 No 3 • November 2015 • A. Özbil, D. Yeşiltepe, G. Argın


Frank, L. D., & Pivo, G. (1994). Impacts of Mixed Use and Density on Utilization of Three Modes of Travel: Single Occupant Vehi- Cle, Transit, and Walking. Transportation Research Record: Journal of the Transportation Research Board, 1466, 44–52. Gehl, J. (2010). Cities for People. London: Island Press Gehl, J. (2011). Life Between Buildings: Using Public Space. London: Island Press Gehl, J., Kaefer, L. J., & Reigstad, S. (2006). Close Encounters with Buildings. Urban Design International, 11(1), 29–47. Giles-Corti, B., & Donovan, R. J. (2002). The Relative Influence of Individual, Social and Physical Environment Determinants of Physical Activity. Social Science & Medicine, 54(12), 1793–812. Goldberg, D., Frank, L., Chapman, J., Kavage, S., & Mccann B. (2007). New Data for a New Era: A Summary of SMARTRAQ Findings. San Francisco: Smartraq Technical Report. Greenwald, M. J., & Boarnet, M. G. (2001). The Built Environment as a Determinant of Walking Behavior: Analyzing Non-Work Pedestrian Travel in Portland, Oregon. Institute of Transportation Studies: Irvine Handy, S. (1996). Methodologies for Exploring the Link Between Urban Form and Travel Behavior. Transportation Research Part D: Transport and Environment, 1(2), 151–165. Handy, S. (2005). Critical Assessment of the Literature on the Relationships Among Transportation, Land Use, and Physical Activity. Washington, DC: Transportation Research Board and the Institute of Medicine Committee on Physical Activity, Health, Transportation, and Land Use. Handy, S., & Clifton, K. (2001). Evaluating Neighborhood Accessibility: Possibilities and Practicalities. Journal of Transportation and Statistics, 4, 67–78. Handy, S. L. (1992). Regional Versus Local Accessibility: Neo-Traditional Development and Its Implications for Non-Work Travel. Built Environment, 18(4), 253–267. Hess, P. M., Muodon, A. V., Snyder, M. C., & Stanilov, K. (1999). Site De-

sign and Pedestrian Travel. Transportation Research Record, 1674, 9–19. Hillier, B., & Iida, S. (2005). Network effects and psychological effects: a theory of urban movement, 5th International Symposium on Space Syntax, Delft. Hillier, B., Penn, A., Hanson, J., Grajewski, T., & Xu, J. (1993). Natural Movement - Or, Configuration And Attraction In Urban Pedestrian Movement. Environ Plann B., 20 (1), 29 - 66. Hillier, W., Yang, T., & Turner, A. (2012). Normalising Least Angle Choice in Depthmap - And How It Opens up New Perspectives on The Global and Local Analysis of City Space. Journal of Space Syntax, 3(2), 155-193. Humpel, N., Owen, N., & Leslie, E. (2002). Environmental Factors Associated with Adults’ Participation in Physical Activity: A Review. American Journal of Preventive Medicine, 22(3), 188–99. Jansen-Osmann, P., Wiedenbauer, G. (2004). The Influence of Turns on Distance Cognition: New Experimental Approaches to Clarify the Route-Angularity Effect. Environment and Behavior, 36(6), 790–813. Krizek, K. J. (2003). Residential Relocation and Changes in Urban Travel: Does Neighborhood-Scale Urban Form Matter? Journal of the American Planning Association, 69(3), 265–281. Lee, C., & Moudon, A. V. (2006). The 3Ds+R: Quantifying Land Use and Urban Form Correlates of Walking. Transportation Research Part D: Transport and Environment, 11(3), 204–215. Lee, C., Zhu, X., Yoon, J., & Varni, J. W. (2013). Beyond Distance: Children’s School Travel Mode Choice. Annals of Behavioral Medicine, 45(1), 55–67. Marshall, N., & Grady, B. (2005). Travel Demand Modeling for Regional Visioning and Scenario Analysis. Transportation Research Record: Journal of the Transportation Research Board, 1921, 44–52. Matley, T. M., Goldman, L. M., & Fineman, B. J. (2001). Pedestrian Travel Potential in Northern New Jersey: A Metropolitan Planning Organization’s Approach to Identifying Investment Priorities. Transportation Research Record, 1705. Montello, D. R. (1991). The Mea-

Modeling walkability: The effects of street design, street-network configuration and land-use on pedestrian movement


surement of Cognitive Distance: Methods and Construct Validity. Journal of Environmental Psychology, 11(2), 101– 122. Oakes, J. M., Forsyth, A., & Schmitz, K. H. (2007). The Effects of Neighborhood Density and Street Connectivity on Walking Behavior: The Twin Cities Walking Study. Epidemiologic Perspectives & Innovations : EP+I, 4(1), 16. Ozbil, A., & Peponis, J. (2012). The Effects of Urban Form on Walking to Transit. Eigth International Space Syntax Symposium, Chile. Parsons Brinkerhoff Quade & Douglas Inc., C. S. & C. A. (1993). Making the Land Use, Transportation, Air Quality Connection (LUTRAQ), National Transportation Library, Oregon. Peponis, J., Bafna, S., & Zhang, Z. (2008). The Connectivity of Streets: Reach and Directional Distance. Environment and Planning B: Planning and Design, 35(5), 881–901. Peponis, J., Ross, C., & Rashid, M. (1997). The Structure of Urban Space, Movement and Co-Presence: The Case of Atlanta. Geoforum, 28(3-4), 341– 358. Rajamani, J., Handy, S., Bhat, C., Knaap, G., Song, Y. (2003). Assessing The Impact of Urban Form Measures on Nonwork Trip Mode Choice After Controlling for Demographic and Level-Of-Service Effects. Paper presented at TRB Annual Meeting. Reilly, M. , Landis, J.(2002). Influence of Urban Form and Land Use on Mode Choice: Evidence From the 1996 Bay Area Travel Survey. Institute of Urban and Regional Development University of California, Berkeley. Rodríguez, D. A., Aytur, S., Forsyth, A., Oakes, J. M., & Clifton, K. J. (2008). Relation of Modifiable Neighborhood Attributes to Walking. Preventive Medicine, 47(3), 260–4. Rodriguez, D. A., & Joo, J. (2004). The Relationship between Non-Motorized Mode Choice and the Local Physical Environment. Transportation Research Part D: Transport and Environment, 9(2), 151–173. Sadalla, E. K., & Magel, S. G. (1980). The Perception of Traversed Distance. Environment and Behavior, 12(1), 65– 79. Saelens, B. E., Sallis, J. F., & Frank, L.

D. (2003). Environmental Correlates of Walking and Cycling: Findings From the Transportation, Urban Design, and Planning Literatures. Annals of Behavioral Medicine : A Publication of the Society of Behavioral Medicine, 25(2), 80–91. Sallis, J. F., Johnson, M. F., Calfas, K. J., Caparosa, S., & Nichols, J. F. (1997). Assessing Perceived Physical Environmental Variables that May Influence Physical Activity. Research Quarterly for Exercise and Sport, 68(4), 345–351. Sallis, J. F., Owen, N., & Fotheringham, M. J. (2000). Behavioral Epidemiology: A Systematic Framework to Classify Phases of Research on Health Promotion and Disease Prevention. Annals of Behavioral Medicine : A Publication of the Society of Behavioral Medicine, 22(4), 294–8. Sallis, J. F., Prochaska, J. J., & Taylor, W. C. (2000). A Review of Correlates of Physical Activity of Children and Adolescents. Medicine and Science in Sports and Exercise, 32(5), 963–75. U.S. Department of Health and Human Services (1996). Physical Activity and Health: A Report of the Surgen General. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion. Shriver, K. (1997). Influence of Environmental Design on Pedestrian Travel Behavior in Four Austin Neighborhoods. Transportation Research Record: Journal of the Transportation Research Board, 1578, 64–75. Song, Y. (2003). Impacts of urban growth management on urban form: a comparative study of Portland, Oregon, Orange County, Florida and Montgomery County, Maryland. National Center for Smart Growth Research and Education, University of Maryland. Timperio, A., Ball, K., Salmon, J., Roberts, R., Giles-Corti, B., Simmons, D., Baur, L. A., & Crawford, D. (2006). Personal, family, social, and environmental correlates of active commuting to school. American Journal of Preventive Medicine, 30(1), 45–51. Turner, A. & Friedrich, E. (20102011). Depthmap Software, Version 10.14.00b. University College London: London.

ITU A|Z • Vol 12 No 3 • November 2015 • A. Özbil, D. Yeşiltepe, G. Argın


Yürünebilirliğin modellenmesi: sokak tasarımı, yol-ağı örgütlenmesi ve arazi kullanımının yaya hareketine etkisi Bu çalışma sokak tasarımı –sokak çevresinin yerel nitelikleri-, yol-ağı örgütlenmesi –kentsel dokunun mekânsal yapısı-, ve arazi kullanım örüntüsünün yaya hareket dağılımıyla olan ilişkisini incelemektedir. Çalışmanın amacı, sistematik olarak ölçülen yol ölçekli kentsel tasarım nitelikleri ile nesnel olarak ölçülen yol-ağı örgütlenmesinin yaya hareketi ve arazi kullanımı ile arasında ne derecede bir bağıntı olduğunu ortaya koymaktır. Sokak tasarımı, yol örgütlenmesi ve yaya hareket yoğunlukları arasındaki bağıntıyı ortaya koymak için İstanbul’un Anadolu yakasında seçilen 20 adet 2kmx2km’lik alan çalışılmıştır. Yaya yoğunlukları ile ilgili veriler alan içinden seçilen yol-parçalarında gerçekleştirilen yaya sayımları aracılığıyla elde edilmiştir. Aynı yol-parçaları; estetik nitelikler, sokak işaretleri, kaldırım

tasarımı, yaya geçitleri/trafik ışıkları, giriş katı kullanımları ve CBS kaynaklı parsel ölçekli arazi kullanım yoğunlukları ve sokak ölçekli topografya verileri bağlamında detaylı saha analizlerine tabi tutulmuştur. Alanlardaki yol-ağı örgütlenmesi açısal yol-parçası analizi (Bütünleşme ve Tercih) ile parametrik bağıntılılık ölçütleri (Metrik ve Açısal Erişim) ile değerlendirilmiştir. Sokak tasarımı, yol-ağı örgütlenmesi, arazi kullanımı ve yürüme davranışı arasındaki ilişkiyi incelemek için doğrusal modelleme yöntemi kullanılmıştır. Bu çalışma, sokak çevresinin kentsel tasarım niteliklerinin ve sokak ağı örüntüsünün yaya hareketi üzerindeki karşılaştırmalı rollerine dair ortaya koyduğu bulgularla önceki çalışmalara katkıda bulunmaktadır. Ön bulgular; -sokak çevresinin yerel kentsel tasarım nitelikleriyle ilintili bazı özelliklerinin önemi göz ardı edilmeksizin- yol ağının genel mekânsal örgütlenmesinin, yaya hareketinin tanımlanması ve düzenlemesinde önemli bir değişken olduğunu göstermektedir.

Modeling walkability: The effects of street design, street-network configuration and land-use on pedestrian movement

ITU A|Z • Vol 12 No 3 • November 2015 • 209-225

Does favorite design lead to good design?: Taxi design competitions in Istanbul and New York City

Harun EKİNOĞLU1, Gülname TURAN2 harun.ekinoglu@gmail.com • Department of City and Regional Planning, Graduate School of Science, Engineering and Technology, Istanbul Technical University, Istanbul, Turkey 2 gulname.turan@gmail.com • Department of Industrial ProductDesign, Faculty of Architecture, Istanbul Technical University, Istanbul, Turkey


Received: May 2014

Final Acceptance: June 2015

Abstract The changes in both urban and national bureaucracy over the last few decades have been described as a shift from “government” to “governance” or as a move from the “old government” to “new governance” (Kjaer, 2009, p. 138). A shift in the public affairs from old public management to the new public management has reasoned the emergence of pluralism and open-discussion platforms into the public tradition, which has also triggered a new paradox: The challenge of pluralism. Although pluralism is an essential ingredient of participation, it accommodates difficulties. The society consists of a huge diversity of different social, cultural, anthropological and emotional attachments. We have different moral codes. When our design ideas compete, so do our values and societal commitments. We recognise those tensions in the rival claims of ideology, ethnicity, gender, religion and locality (Bellamy, 1999, p. 1). Within this article the struggle between the searches for good design vs. popular phenomena is being questioned out of Istanbul and New York City’s (NYC) taxi design competitions’ methodologies and results. Experiences revealed from both case studies prove that the challenge between the popular taste and search for good design may not always promote either design itself or the promoter.

Keywords Design, Society, Good design, Popularity, Competition, Governance.


1. Introduction 1.1. Theoretical approach According to Kjaer (2009), since the 1980s, public administration has been reformed towards a new focus and hierarchy between politics and public administration through the term governance. This was also a shift to an organisational set up emphasising the functionalities of the networks and juxtaposing roles of politicians, public administrators and civic actors (Kjaer, 2009, p. 138). This is no less the case in the local state where the change has been described as a move from “local government” to “community governance”. Urban Governance theory thus highlights changes in urban bureaucracy such as the move towards a blurring of public private boundaries, the rise of an increasing number of governance networks and a greater inclusion of actors other than the local state in the pursuit of community goals (Kjaer, 2009, p. 138). In late 1960s, rationalist architects such as Christopher Jones, Christopher Alexander, Tom Markus and Ray Studer claimed that considering some mathematical models could objectify a better programmed decision-making process (Broadbent, 1980). However, by the 1970s these highly rationalist design methods had been rejected since they were also setting the goals in a top-down approach defining the problems away from a world made up of a great diversity of values and priorities (Comeiro, 1990). Scholars underlined the difficulties faced when applying mechanical-rationalistic methods to design problems. Contemporary society has complicated issues related to everyday urban life that straight methods of planning can no more deal with. Conventional approaches to already defined problems could no longer be adapted to the wide array of today’s wicked problems which are in fact symptoms of some other everyday problems with no formulated solutions. Different layers of the society with different value judgements would no longer face the problems triggered by some other problems with the same set of criteria. There is no common “true” or “false” where of incomplete

and contradictory knowledge with a vast amount of people and opinions involved. In such ambiguity, the search of a solution that works for everyone becomes a useless struggle with conventional top-down decision making traditions (Comeiro, 1990). Thus urban life-related design problems would be best solved in a process that paves the way for direct involvement and deliberative dialogue among the all concerned ones. Hence, design comes up as a public discourse and paves the way for public debate in order to expose and spread the current state of art knowledge about the problem, since no one in fact has expertise to come with a solution. Fischer (2000) in “Symmetry of Ignorance” draws a core understanding and states that individuals are not equal in what they know, but in what they do not know. The knowledge or the opinion that each of us has, may greatly alter and evolve the content and profile of the solution. This view, especially in metropolitan cities, encourages rejecting the conventional top-down methods to find out solutions to design related urban problems. Thus, it surely encourages democratic and open-discussions where all the stake holders or interest groups should have their interests and opinions articulated. By the late 1980s a clear change occurred in local administrations’ approach to commissioning design. Accordingly, they started to plan the entire design process in a participatory manner, opening up to all stake holders in collaboration with actors of private sector and academia. This new pluralist, open and transparent “Second Generation Design Method” (Comeiro, 1990) resulted not only in a new role of the public authorities but also for designers. Governance in design triggered a trend from user-centered to user-led design (Sanders, 2002). Blurry lines between the designer and user ushered in a rise of highly programmed interactive platforms for understanding the user’s experiences, needs and reactions. In a transition from government to governance, the notion of design has surely had its share from this transition. The focus of the design activity shifted from the design object itself towards participative and inclu-

ITU A|Z • Vol 12 No 3 • November 2015 • H. Ekinoğlu, G. Turan


1 Istanbulites experienced this type of a collective decision making before the new taxi competition, for selecting the new city boat in 2006. The Department of Istanbul Maritime Lines asked people to vote for their favorite boat design out of 8 different alternatives. The majority (41% of the voters), voted for the design resembling the former the most (Hurriyet, 2006). Today, the old and new boats float among giant oil tankers on the Bosphorus, surrounded by armies of seagulls (ZeroIstanbul, 2012, p. 110).

sionary character of the process of the design. The core contribution of participatory design methods is, to the extent possible, that the user’s knowledge and intervention is collected and utilized in design process. Theory of participatory design methods also underlines that in addition the democratizing the design it also promotes the identification of the point-in-question design matter. Participation is also considered as a method for legitimizing the majority’s views and unifying the opposing views in a collective way. Thus, participation as a tool acts like an educative and socialization catalyser (Comeiro, 1990). The reflection of “new governance” in design finds itself as “participatory design”. Participatory design in architecture and planning, together with its theory and techniques have been on the agenda of city planners and architects since the 1970s, as a considerable movement towards the direct involvement of the public in the definition of their physical environment. Participatory industrial design later followed the way. Participation is a matter of control over decisions by the participants. Having explained the shift from local government to a new governance and its synergy of win-win interactions between public, private and community; and the relationship between the new governance and participatory design, two participatory design intend from two different cities -Istanbul and New York- that are comparable in certain aspects will be discussed. In the two design competitions discussed here, the activity of “participation” is enabled through competitions where professional/independent designers or manufacturers were invited to make new taxi designs and public to select its favourite by the local governments of two metropolitan cities1. According to Wulz (1986/1990, p. 39), a design competition is a solution to alienation between designers and users through allowing users to employ their influence in the design or decision making process. Participation is a general concept with different forms of decision making methods by the involved parties. Participation can be active or passive as Wulz defines in seven different forms and stages: Representation,

Questionnaire, Regionalism, Dialogue, Alternative, Co-decision and Self-decision (Wulz, 1986/1990, p. 41). According to Wulz (1986/1990, p. 41), representation is a passive form of participation where designer, with a clear social sensitivity, considers himself/herself as the user or client in design process and designs accordingly. Questionnaire is another passive form of participation of the anonymous user. The objective is clear; more observable and statistically comprehensible data can better take the designer to the user needs and experiences. Regionalism appears as a solid concern for the place based values and references of a particular territory where the design is made for. Dialogue is based on the concept of using people’s knowledge as a source and asking them to comment on the designer’s proposal while the design process is in progress. This sort of dialogue may either happen face to face or through a dynamic interface. Alternative is a form of participation where the users are encouraged to make selection out of a number of different design alternatives. The critical aspect of this type of participation is the possible restrictive effects of the presentation. The way that the alternatives are presented is vital in terms of the impression created. Co-decision is a participation method where the citizen/user has the biggest role starting from the design process to decision making out of alternatives. The promoting agency manages the entire process without imposing or dominating the design. The citizens are both active designers and the decision makers throughout the entire process. Competitions from ideas, emerged from the public, to the selection of the final design among the alternatives have a lot to offer in sense of co-decision way of participation (Volker, 2010) (Nasar & Kang, 1989). Self-decision is a participation model where not only the majority but everybody has equal right to influence the design. Wulz (1986/1990, p. 46) states that this is a form of self-build or selfhelp method where the designer or architect is engaged as a consultant. However, self-decision design method can only be applied in small groups for productive design processes. If not,

Does favorite design lead to good design?: Taxi design competitions in Istanbul and New York City


representatives of the groups need to work with the consultant and this may undermine the synergy of this method. Below the steps taken in “taxi design” attempt of both cities are discussed with reference to aforementioned participatory design methods where there are similarities. 2. A brief review of both design competitions: commonalities and differences The necessity of new design concepts led to design competitions for taxi cars in NYC and Istanbul during approximately the same time period. Istanbul’s competition launched in March 2011 while NYC’s occurred in May 2011. In NYC, pre-competition process and R&D phase started in 2007 with a vast amount of public consensus studies. The city administration issued a request for proposals for the manufacturers and designers to submit their designs for a purpose-built taxi design to serve as NYC’s taxi of tomorrow (DTPS, 2007) (nyc.gov, 2011). The shortlisted entrants to NYC competition were announced in February 2011 for the online voting and the winning entry was announced in April 2011. Istanbul’s early pre-competition studies started in 2009, through meetings with Chamber of Istanbul Taxi Drivers, Istanbul Metropolitan Municipality’s (IMM) Mass Transportation Services, independent designers and manufacturers, lasting from 2009 to 2010 (IBB, 2010). The competition design brief was prepared in 2010 and the competition has been launched in March 2011 (www.taksitasarim.ibb.gov.tr, 2011). This paper draws a critical perspective on the participatory design attempts for designing the next generation of taxis by two world class cities, NYC and Istanbul, with diversified rankings of business activity, human capital, information exchange, cultural experience and political engagement (Hales, Peterson, Pena, & Gott, 2014). The paper builds a critical review on both cities’ participatory design activisms in their differences and similarities. Apparently, two cases have methodological differences as well as some similarities. Differences of both cases are considered to be factors contrib-

uting to the argument for questioning the tension between the search of good design and challenge of popular phenomena. Two important points need to be stated in the beginning of the paper, one of which is about the term good design, and the second is the involvement of the authors in the mentioned competitions. It is obvious that there are different sets of characteristics attributed to good design when gazed into design history and theory, and the paper is held within this acception. On the other hand, there are common characteristics such as usefulness, understandability, unobtrusiveness, aesthetics, honesty, sustainability and environmental friendliness, most of which can be traced in the ten commandments of good design put by the German designer/architect Dieter Rams in the 1970s. The first author of the paper, free from evaluation of the entries, was in charge of providing independent design consultancy to the City of Istanbul throughout the entire competition. Both cities witnessed a thorough R&D process as the private sector and NGOs focused on the quality of taxi services. Thus both cities perceived the issue as a societal matter as well as design and they planned an open process for designers and for the citizens who would be able to vote for their own favourite taxi concept. 3. Istanbul and New York City cases: Istanbul case Taxi service in Istanbul is a complex issue. Considering Istanbul’s heavy traffic congestions make it clear that taxi cabs are effective for the citizens, thus they bear an importance not only for their service but also for the aesthetic effects created by the taxis within the city. It is not only a vital part of daily transportation but also the centre of controversies with its ever growing chronic problems on city’s agenda. IMM’s Department of Mass Transportation Services reports that there are 18,000 registered cabs, and between 30,000 to 60,000 illegal taxis giving service around Istanbul, a city of 14 million (IBB, 2010). According to Aydinonat general public perception

ITU A|Z • Vol 12 No 3 • November 2015 • H. Ekinoğlu, G. Turan


about the service quality of the taxis in Istanbul is remarkably low (Aydinonat, 2013). Principally IMM restricts the number of registered taxi plates, which can be leased for taxi service. Thus an ever increasing demand at the city for taxi service is making the value of a single taxi plate a very profitable investment tool, in free market. Today the value for a taxi plate is listed at almost 1,2 million Turkish Lira (approx. $500.000) (milliyet, 2013). Ownership of constantly increasing taxi plates is perceived as an alternative investment instrument. Therefore providing a high quality taxi service has become less important. Problems with Istanbul taxis’ low service quality are more or less a result of this fact together with the issue that there exists no rivalry on the quality of service provided by taxis. There are also illegal taxis around the city because of growing demand for taxi services. These illegal ones are fulfilling this need. It is estimated that government’s annual tax loss due to illegal taxis is approximately 100 million Turkish Liras (cnbce, 2013). Under these circumstances, the rationale behind the restrictions on the number of registered plates becomes dubious. One answer could be the vested interests of current holders of these plates. In Table 1, Aydinonat (2013) sheds some light on facts and figures about number of the taxis around the world metropolitan areas as in the following (TUBITAK & Cetin&Oguz, 2007, Table 1. Number of the taxis around some of the world’s important metropolitan areas (TUBITAK & Cetin&Oguz, 2007, 2010, 2013, 2008).

2010, 2013, 2008); The above numbers show that on average a single legal taxi is giving service to 800 people in Istanbul, versus 662 people in NYC. Although the numbers for Istanbul point out an insufficient amount of legal taxis in the city, authorities still prefer not to increase the number of the registered plates. Apparently as long as the numbers of the registered plates remain constant, the service quality of the legal taxis will have no reason to increase in Istanbul. In other words, currently there is no reason for emergence of competition for a higher quality taxi service among the taxi service providers in Istanbul apart from IMM’s concerns. Nevertheless under these circumstances in the period between 2009 and 2010, IMM’s Department of Urban Design began to work organising a taxi design competition to increase the standards and service quality of Istanbul taxis (IBB, 2010). City of Istanbul’s main concern was to make a process of design involving open dialogue, communication and trust as Sanoff ff cites as the integral parts of participatory design (Sanoff, ff 1990, pp. 5-21). Istanbul’s attempt, almost in the same period with NYC, to launch a taxi design competition for the city, was a national and two-phased “Taxi System Design Competition”. In March 2011, IMM has launched Istanbul’s new Taxi System Design Competition as a design idea competition to seek and promote creative ideas, approaches and new design concepts under 4 sub-design themes: a taxi for the general purposes, eco-taxi, a taxi communication system and a taxi stop/service point, and with 3 different ff designer profiles; professional designers, university teams and high school students. The city’s participatory design proposition as Sanoff ff (1990, pp. 6-7) also describes was based on a belief that people affected ff by design decisions should be involved in the process of making those decisions. In other words, IMM paved the way for the emergence of the local creativity. Having understood the reasoning behind the attempt of a design competition for the city’s new taxi concept, searching for a consensus at second

Does favorite design lead to good design?: Taxi design competitions in Istanbul and New York City


stage for the shortlisted design proposals presented another challenge: finding balance between the creativity of emergence and stability of design (Hamdi, 2004, p. 18). The finalists in each design case were announced on the competition’s web site and their proposals were put to a popular vote. All around the world, nearly 1.5 million people visited Istanbul’s taxi design competition web pages and more than 340.000 people voted for the shortlisted design proposals over 15 days (www.taksitasarim. ibb.gov.tr, 2011). Finally an interdisciplinary jury made its decisions for the award winning designs in line with people’s votes through considering the amount of the online votes for each design proposals. The whole process was a step-by-step interactive design activity from the beginning till the end and an “open to everybody’s ideas” process intending to make the new taxis “everybody’s taxi”. Sanoff states his experience in user participation in design as “the main source of user satisfaction is not so much the degree to which his or her needs have been met, but the feeling of having influenced the decision” (Sanoff, 2006, p. 140) 3.1. The general structure of Istanbul’s design competition Istanbul’s competition granted a feeling of control over the new design ideas was also a social contract between the city and citizens as Sanoff (1990, pp. 5-21) depicts, implying that their needs, values and ideas would be taken into consideration. The competition’s design brief explained that the city authority invited all the professionals, college and high schooled designers into the competition no matter what company or manufacturer they were working for. Istanbul city authority and the jury both accepted this principle to provide and strengthen the equality and avoid any conflict of interest between the contestants, the jury and the city authority (IBB_DesignBrief, 2011). As a step by step interactive design activity, Istanbul’s taxi design competition was a collaboration between the contestants, executing authority of the city and the members of a multidisciplinary jury. By definition of par-

ticipatory design Wulz defines this as co-decision in his approach of seven forms and stages of participatory design (Wulz, 1986/1990, p. 41). Prior issue of co-decision is developing a balance between design process and decision-making. According to Wulz (1986/1990) co-decision involves the stake holders from the beginning of a design process and aims at user’s direct and active participation. The jury of Istanbul’s competition has been chaired by a keynote personality, Önder Küçükerman, who pioneered both the establishment of industrial design education and corporate design activities in Turkey. By the

Figure 1. Shortlisted final 4 design proposals presented at the online poll (www.taksitasarim.ibb.gov.tr, 2011).

ITU A|Z • Vol 12 No 3 • November 2015 • H. Ekinoğlu, G. Turan


Figure 2. General view of Istanbul’s winning taxi design proposal (www.taksitasarim.ibb. gov.tr, 2012).

Figure 3. General view of Istanbul’s winning taxi design proposal (www.taksitasarim.ibb. gov.tr, 2012).

Figure 4. General inner view of Istanbul’s winning taxi design proposal, (www. taksitasarim.ibb.gov.tr, 2012).

end of the first step evaluations of the design proposals, high school students’ category award winning designs were determined by the jury (IBB-Rec1, 2011). The shortlisted design proposals on each of the four sub-themes belong to professionals’ category and university students/teams’ category have been chosen and announced, as entrant codes, on the competition’s website (IBB-Rec2, 2011). Another two months of project developing process was given with a new set of criteria developed by the multi-disciplinary jury for the shortlisted projects in each design cases (IBB-Rec2, 2011). By the end of the

second phase, the jury again chose the properly and satisfyingly developed four design proposals in Figure 1. to present at online polls at competition web site during fifteen days (IBB-Rec3, 2011). According to the city authorities, people’s votes were meant to distinguish the good designs and to end up the all process without ambiguity and discussion (IBB-Rec5, 2011). Online poll results clearly pointed people’s favourite designs on each theme for both contestant profiles. Sanoff defines it as “the issue of individual influence in decision making and its proportional impact that can best be resolved by the participants themselves” (Sanoff, 1990, p. 1). Taxi for the general purpose of the Professionals Category online poll results were shaped as in the following (IBB-Rec4, 2011) (www.taksitasarim. ibb.gov.tr, 2011): According to the jury evaluations, the project with P4 code was one of the most promising one in terms of its satisfying and advanced project presentation shown as in Figures 2, 3 and 4 (IBB-Rec4, 2011). This fact was appreciated by the jury at the last meeting before the online voting process (IBB-Rec4, 2011). However, the project also conveyed some unqualified design decisions and incorrect technical resolutions (IBB-Rec4, 2011). Even though the jury was impressed with its professional presentation, the odds with specific design and technical proposal were also openly criticised (IBB-Rec4, 2011) (IBB-Rec5, 2011). The jury articulated the general problems as exaggerated proportions, bad seat positions and inappropriate placement of the sliding doors (IBB-Rec4, 2011) (IBB-Rec5, 2011). According to the members of the jury, the majority of the voters would have never seen the projects through the eye of a technical expert but his/her own personal taste of aesthetic or beauty (IBB-Rec4, 2011) (IBB-Rec5, 2011). The modest and appropriate design features and the consistent inner and outer technical resolutions in appropriate proportions of the taxi design project with the P21 code were also noted and appreciated by the jury (IBB-Rec4, 2011) (IBB-Rec5, 2011). 129,575 votes had

Does favorite design lead to good design?: Taxi design competitions in Istanbul and New York City


Figure 5. Jury and technical committee studies ( IMM-Department of Urban Design Archive, 2012).

been clicked for the finalist four design projects in general taxi design category (IBB-Rec5, 2011). The design proposal P21 was voted by 41,2%, while project P4 was preferred by 28,5% of the voters (IBB-Rec5, 2011). The jury has chosen P21 as the winning design of the general taxi design category and it was awarded with 100,000 Turkish Liras (IBB-Rec5, 2011). In spite of the written objections raised by the designers of P4, the jury and the technical committee in Figure 5, noted that “We have been looking for a taxicab with clear taxi proportions not a taxicab with minivan proportions” (IBB-Rec5, 2011). Jury also states that “Online poll results are also clearly pointing out to this vital detail. This is why the winner of this theme is the project P21” (IBB-Rec5, 2011). According to the jury, citizens’ favourite design was obviously the fruitful result of the search of “good design” in Istanbul’s taxi design competition (IBBRec5, 2011). 3.2. New York City case “A taxi is not a car; it is a moveable public space. It may have four wheels and carry passengers, but the circumstances are completely different” (DTPS, 2007). A Ford’s Crown Victoria Taxi cab, shown in Figure 6, in NYC are both loved and hated by the New Yorkers. They are assumed as practical and alternative modes of transportation over Manhattan throughout the routes not sufficiently fulfilled by the subways (www.ny.com). One can say that the traffic that results alongside the Manhattan’s streets is the flood of the yellow cabs. With over 10,000 yellow cabs giving service to NYC is the prior reason to the mid-town traffic

Figure 6. In yellow-cab form, Ford’s Crown Victoria has been an integral part of the New York cityscape for years (http://news. drive.com.au/drive/motor-news/revealednew-yorks-new-yellow-cab-201105041e79l.html).

(DTPS, 2007). Goldberger, the Dean of the Parsons The New School for Design, says that “What is troubling about the NYC taxi is not that it is ubiquitous, but that it is so ill-suited to its job. There is something brightening to the cityscape in the constant flow of deep yellow vehicles along the city streets but then you get into one of them, and you are reminded that it is hard to enter, hard to leave, uncomfortable to sit in, and awkward to carry luggage in. It is as likely as not to be dirty, and it may or may not have a functioning air conditioner. It is hard to communicate with the driver. And, although you are unlikely to realize this is as a passenger, the NYC taxi is no friend to the environment (Goldberger).” There is no doubt, hailing a cab in NYC -with its spirit of freedom, power, and anonymity - is more and more a “must to do” thing in the city. A number of people visiting the city and experiencing the taxi cabs are making the “NYC cab notion” among top tourist attractions (DTPS, 2007). Although taxis still accommodate a crucial NYC experience, few would oppose to the idea that they should be more comfortable, better designed, and accessible for all. Hence at the hundredth anniversary of the gasoline powered taxi in 2007, the NYC’s Design Trust for Public Space has been studying how this iconic wealth of transportation could be revised, with the eventual goal of making a new taxi design for the century turn over. In total, over 50 designers and taxi stakeholders took part in a spirited discussion of all things about taxi where the participants discussed trends in taxi design, the taxi’s role as

ITU A|Z • Vol 12 No 3 • November 2015 • H. Ekinoğlu, G. Turan


a public space, and the ideal taxi of the future, from the perspective of a taxi passenger. On the basis of the effective ff and operative efforts ff developed by the NYC’s Design Trust for Public Space, Table 2. NYC’s poll results for the “Taxi of tomorrow” (Taxi of Tomorrow Survey Results, NYC, TLC, 2011).

Fiigure 7. Design F i proposals for the New York’s “Taxi of Tomorrow”(www.taxioft o omorrow. com, 2011).

the studies and workshops have been done for capacity building processes in the search for NYC’s taxi of tomorrow (DTPS, 2007). Later on, in 2007, city officials convened a group of stakeholders, including representatives of taxi drivers, owner and passengers, to create a set of goals for the next NYC taxi cab, a project called the “Taxi of Tomorrow”. In December 2009, the authorities initiated a “request for proposals,” inviting auto manufacturers and designers to submit their best ideas for a purpose-built vehicle to serve as a NYC taxicab. On May 3, 2011 it was announced that the NV200, designed by Nissan had been chosen as the winner of the competition (nyc.gov, 2011) (DTPS, 2007) where the other two finalist producers were Ford, and a Turkish company Karsan. Participatory point of NYC’s process of seeking the “Taxi of Tomorrow” is its online poll process for the three finalist automotive companies’ design proposals. According to the poll results, as appears in Table 2, Karsan was voted as the “most loved” design, with 38,9% of the “Love it” votes, by the New Yorkers. The poll raised a unique question; “What do you think of the new designs?” and requested five diff ferent feelings for each design from the citizens: The results, for the design proposals shown in Figure 7,were shaped as in the Table 2 (nyc.gov, 2011); For each of the following design proposals; Eventually, NYC authority chose Nissan as the “Taxi of Tomorrow” by putting the poll results aside assuming that the New Yorkers “loved” the Nissan’s design too. However, this decision has led to some controversy. Journalist and blogger James Wagner wrote in

Fiigure 8. Prototype of selected design F i concept of Ni N ssan NV 200 model. (www. taxioft o omorrow.com, 2011).

Does favorite design lead to good design?: Taxi design competitions in Istanbul and New York City


his e-article “NYC gets yesterday’s taxi, not ‘The Taxi of Tomorrow’ “ stating “Although the very modern, beautifully-designed, extraordinarily-roomy and fully-accessible Karsan V1 was hailed by New Yorkers as their favourite, the City ended up choosing the least popular entry, the hideous Nissan NV 200” (Wagner, 2012). He continues by saying “While the Nissan was certainly the most conservative response to an important challenge, in the end it will prove to have been the most impractical choice, and therefore the most radical, given the parameters of the search: Of the three finalists it responds the least well to current taxi needs, and its environmental and accessibility inadequacies, among others, will look even more grotesque as time goes by. According to Wagner “in picking the barely-adequate, ungainly and unlovely Nissan “they” struck out once more, embarrassing New Yorkers who actually care about the city’s ability to get things right (both better than and before others do, if possible)” (Wagner, 2012). As a response to the critics and reactions, city authority informed that a French designer François Farion would work on the Nissan’s proposal to make it better in design as the “Taxi of Tomorrow” for the New Yorkers. Immediately after TLC has announced their final choice of the competition, City Authority of NYC signed an agreement worth $1 billion with Nissan to supply city’s new unique taxi cabs each for $29,000 market price for the next ten years. Queens based Taxi Safety Committee came out with a claim that Nissan NV200, shown in Figure 8, models were not up to date in terms of taxi cab safety issues and they litigated. Recently in early October of 2013, the Supreme Court decided for a suspension of execution on the account to the fact that the city has no privilege to designate one single company to provide city’s new taxi cabs (NYTimes, 2013). Apparently this has brought a new and unexpected twist to the city’s agenda on the process of implementation of the new taxi cabs. Wulz in his article of “The Concept of Participation” claims that there is always a controversy in participatory

design when majority’s choice dominates minority’s expectations (Wulz, 1986/1990, pp. 41-44). In NYC case, city’s final choice that paying no attention to citizens’ decision is beyond disturbing for both parties. Although there may be many different reasons on the selection of the Nissan’s design, NYC’s decisions reminds the challenge of “what happens when the most favourite design is not approved as the good design?” 4. Potential outcomes and critical comparison of both cases In NYC the designer companies have been technically briefed prior to making their bids. However in Istanbul, IMM purposely let the designers free for encouraging original ideas in the first phase and briefed the shortlisted entries on technical details in the second phase of the competition (IBBRec3, 2011). People in NYC expressed their feelings for each design entry in 5 different ways from “like it” to “take it or leave it” while in Istanbul, people have only been asked to choose their favourite design out of four finalist entries. This paper considers the ratio of “love it” votes in NYC as an equivalent parameter to the ratio of “favourite” votes in Istanbul’s poll. Other four feelings for each of the three finalists’ design proposals in NYC are valuable and yet might be considered as data for another case study. Post-competition agendas of Istanbul and NYC were slightly different. NYC authorities openly expressed that it was a purpose built design competition while City of Istanbul stated that the winner design proposal could partially or fully be implemented or considered as a capacity building activity for better taxi service standards for the future (nyc.gov, 2011) (IBB_DesignBrief, 2011). In addition to that, City authority of Istanbul also required the entrants to contest, free from their corporate attachments, as independent designers. These methodological differences affected the profile of the entrants in both competitions. Due to differently articulated purposes of the competitions, corporates and manufacturers such as Nissan, Ford and Karsan entered the competition

ITU A|Z • Vol 12 No 3 • November 2015 • H. Ekinoğlu, G. Turan


in NYC while designers participated in the competition in Istanbul as independent contenders. From this point of view Istanbul’s design competition aims at involving the citizens in this design experience as independent designers while in NYC companies are in charge of design and citizens are only getting involved in the process via online voting. The backbone of the participative process of NYC design case is mainly structured on getting people’s feedback on the features of each design proposal. Istanbul in this respect is programming the entire competition with a concern for promoting the citizen involvement as much as possible for different designer profiles and different design themes (IBB_DesignBrief, 2011, p. 5). Both processes had different methodological characteristics. Even though NYC had three design proposals with complete R&D backgrounds developed by the manufacturers, the entire process was finalized by the “least loved” design (NYCTLC, 2011). On the other hand, Istanbul’s design competition for taxi and its system pieces have promised to be multidisciplinary design opportunities for the independent designers and design teams. Even though Istanbul’s method seems more risky and challenging especially for the post-competition progress and the industrialization of the designs, the high calibration between people’s favourite designs and jury selections have proved the success of the participation. Istanbul’s design competition in Wulz’s (1986/1990, pp. 44-45) terms is a co-decision participation model that mostly occurred as a balanced decision making process. Likewise influencing population had a direct influence on the final decisions in Istanbul’s competition. Jury showed highest effort to get in line with people’s favourite designs in all themes (IBB-Rec5, 2011). As an exception, although one of the finalist design proposals on “taxi stop/service point” category was the most favourite one by the poll, due to a considerable amount of objection emails from the public against it, the jury approved the second most favourite design as the winner (IBB-Rec5, 2011). Putting the manufacturers or de-

signers at the very centre in the process and encouraging them for new design concepts for the city’s new taxi design is somehow a serious task and a heavy responsibility. Considering the vital claims of “taxi is a design object” and “taxi is not a car but a moving public space,” its comfort, security, and urban identity issues are respectfully vital. Thus design of a taxicab may become an even more complex issue. When this sort of responsibility is programmed as a step by step interactive design process between city authority, designers and citizens genuinely, then the new design ideas inevitably may have the chance to emerge from the inside of the city. That is to say, the collective taste of design may appear on the stage. The good potential of the pluralism in the society may activate a fruitful discussion for the different design ideas. However, academia emphasizes that good design does not always have to mirror the users’ wishes and tastes. Users’ tastes, desires and interests on a particular design object do not always imply the success of it (nyc.gov, 2011). One can say that the politicians behind both Istanbul’s and NYC’s processes did not only disregard this claim but also the question of “is publicly favourite design enough?” (Maile, 2012). Numerous similar experiences exist between the cases of NYC and Istanbul. One of the design proposals in both cases had clearly been highlighted by significant public support. The city authorities would have never felt better about the results had they not consulted with the broader society. Apparently none of the political figures in both cities would disagree on these clear outcomes. Thanks to the well-organized propaganda which is one of the greatest weapons of pluralist western democracies, missions in both cities created successful PR processes out of taxi and design issues. Yet the results evolved differently. Leading political authorities’ final decisions in both cities had vital role on the results of both processes. There existed a respectful amount of a public support for one particular design idea, and the rejection of that design for concerns of avoiding failures, and continuing with a “least loved” de-

Does favorite design lead to good design?: Taxi design competitions in Istanbul and New York City


sign proposal, triggered enormous controversy in NYC. Avoiding the institutional failures but faced with unpleasant critics may be defined as the unforeseen risk of the game. According to McGrath “most organizations are profoundly biased against failure and make no systematic effort to study it. Executives hide mistakes or pretend they were always part of the master plan. Failures become undiscussable, and people grow so afraid of hurting their career prospects that they eventually stop taking risks” (McGrath, 2011). This is, more or less, what happened in NYC where the process was barely supported by the civic society until the politicians decided to continue with the “least loved” design concept of the competition (nyc.gov, 2011). This situation had increased a public tension especially at intellectuals’ side of the city. They all raised the same question: Is the taxi of tomorrow the least loved design idea? Where is the vision in it? The fact is that the visions may be revised. Lootsma states “let us define our terms” and continues: “revisionism is a tendency where in general politics take position (Lootsma, 2011). This is a tendency to conserve what once existed and still aims at radical change in the present system. Revisionism is a path between revolution and conservatism. Needless to say that this is the route and the tension of post-modernism (Lootsma, 2011). Apparently this tension was the major struggle that politicians in both cities have faced. Relatively Istanbul’s experience is no more different in this respect. The Istanbulites’ favourite design and the common taste of design of the public majority was awarded as the winner design which had a support of 41,2% amongst all (IBB-Rec5, 2011).To this end, Istanbul’s competition did not only award the designers but also to the public opinion and its common taste. Due to no prior commitment for the production of the winning design, designers had no concrete expectation for the production. Following the city’s responsible authorities decided not to produce the winner design but considering and implementing its high service and design standards for the different type of taxicabs that will pro-

vide service within the city in the near future (hurriyet, 2011). Being different from NYC, Istanbul’s design competition and its following process, with 4 major design cases, 3 major designer profiles and online voting to search for a public consensus, was a city wide capacity building process. Istanbul has never committed the production of the winning design concepts as a whole as New York did. To this end public discomfort against to all process in public opinion in Istanbul is much harmless when comparing to the opposition voices in New York. When considering the controversial topic on the post-competition progress of NYC’s “taxi of tomorrow” project, one can say that NYC does not have a publicly-supported new taxi design concept as Istanbul achieved at the end of the whole competition. Differently from Istanbul, NYC wanted to develop the new design concept for the city’s new taxicabs through a process that invited and included manufacturers. Similarly, both NYC and Istanbul made the citizens to vote online for the shortlisted design ideas. Istanbul and NYC political authorities’ enthusiasm to consider public reactions for taxi and its system design made both processes into a popular design activity. 4.1. Comparative and theoretical discussions: The challenge of plurality in design The Reagan and Thatcher era of the 1980s was characterized by an array of new public management reforms that were, if not global, then at least very widespread in geographical scope. Although there is no agreed upon definition of new public management, most observers seem to agree that it entails at least seven aspects transferred from private sector management principles to the public sector, such as handsoff, professional management, explicit standards and measures of performance, managing by result and value for money, privatization, agentification, competition, decentralization and citizen empowerment (Kjaer, 2009, p. 138). However, a new paradox appeared: challenge of pluralism. Needless to stay, discussions in open and pluralist platforms are far more chal-

ITU A|Z • Vol 12 No 3 • November 2015 • H. Ekinoğlu, G. Turan


lenging than those happening behind the doors. According to Bellamy, such pluralism gets into modern societies, the mixed blessing of their differentiation and openness (Bellamy, 1999, p. 1). When the associated diversity enhances one’s everyday life’s efficiency, it may also reason many of its troubles being obliged to choose between conflicting constraints, purposes, rules and considerations. The theory and practice of politics are no less stable in a world of rapidly growing social and value pluralism. Due to a trend of becoming more and more heterogeneous, citizens hold ever more atypical and often contrary identities, ideologies and interests in western developed societies. Not only private but also public life may encounter problematic and occasionally even tragic choices. However these circumstances place contemporary orders in a dilemma: can they respect plurality yet produce collective contracts that govern an unobligated loyalty? When reconsidering the concepts of the popularity of design and the power of plurality in the same pot, Istanbul’s taxi design competition is inspiringly convincing to develop inquisitorial analogies about the tension between popularity of design objects and the search of good design. Both cases suggest that crediting the majority’s “good” as “good” without experiencing it, emerges to be a paradoxical issue. 5. Discussion and conclusion Populist Politicians react with a postmodern, but confused, way (Lootsma, 2011). Confusion may appear in any way that the politician would not prefer to be a part of, especially after all his attempts to reach to the pluralist creative opinion of the society via a well-organized process. A dilemma between public opinion and some technical or legal concerns may put the politician in an opposite position in comparison to overall mission of the process. Beyond a doubt this kind of change in expression and in mind setting might result by a decrease in his or her public support. Looking for innovation to upgrade the urban life may be energized through populism from a political point of view. However, possible

conflicts between populism and real world constraints especially for design ideas obtained via open public processes may naturally bring up some unexpected concerns on public stage. One can say that developing creative ideas for design objects for wide public use throughout an open and participatory process has its own rules. This may imply a shift in literal and technical definition of governance in design issues. Evidence gained from both metropolitan cities’ participatory design processes suggests that when the deal is design, populism does not always take us to good design and the political promoter may somehow be harmed through the process. Instead of making generalizations on the issue of good design versus the popular one and its relation to participatory approach, two concrete competitions with similarities depicted a fruitful platform to engage critical review. Is the most popular design always the good design? or “Is publicly favourite design enough?” Design thinkers and academia suggest that the answer might just be “no” (Maile, 2012) (McGrath, 2011). Experience revealed from both of the different approaches of NYC’s and Istanbul’s design competitions and post-competition discussion and decisions inform us that the design preferred by the public on paper may not always be the good design unless it is produced, used and proved. On the other hand a good design as defined and agreed by the experts may not always be the most popular design for people who may have different criteria and taste from those who determine what the good design is. References Aydinonat, S. (2013). “Korsanla Müdacele ve Devletin İstanbul Taksi Piyasası ile İmtihanı” Türkiye Ekonomi Politikaları Araştırma Vakfı, TEPAV. Retrieved April 23, 2013, from http://www.tepav. org.tr/upload/files/1366701330-1.Korsanla_Mucadele_ve_Devletin_Istanbul_Taksi_Piyasasi_ile_Imtihani.pdf. Bellamy, R. (1999). Bellamy, R. “Liberalism and Pluralism: Towards a Politics of Compromise” . London: Routledge. Broadbent, G. (1980). The Morality

Does favorite design lead to good design?: Taxi design competitions in Istanbul and New York City


of Design. Design Science Method: Proceedings of the 1980 Design Research Society Conference (pp. 309-311). Westbury House Publishers. cnbce. (2013). cnbce. Retrieved June 1, 2013, from www.cnbce.com: www. cnbce.com/haberler/ekonomi/taksi-plakasi-altin-degerinde Comeiro, M. C. (1990). Community Design: Idealism and Entrepreneurship. In H. Sanoff, Participatory Design: Theory & Technics (pp. 21-37). Bookmasters. DTPS, D. T. (2007, November 5). Designing the Taxi, Rethinking New York City’s Moveable Public Space. Retrieved April 1, 2010, from www.designtrust. org/pubs/05_Designing_the_Taxi.pdf Fischer, G. (2000). Social Creativity, Symmetry of Ignorance and Meta-Design. In Special Issue on “Creativity & Cognition 1999” of the International Journal of “Knowledge-Based Systems” (pp. 527-537). Oxford, UK,: Elsevier Science B.V.,. Goldberger, P. (n.d.). Taxi as Icon: News, Clips and Selected Essays. Retrieved May 1, 2011, from www.nyc. gov/html/media/totweb/taxioftomorrow_history_links_designtrust1.html Hales, M., Peterson, E., Pena, A. M., & Gott, J. (2014). ATKearney 2014 Global Cities Index and emerging Cities Outlook. ATKearney. Hamdi, N. (2004). Small change: About the Art of Practice and The Limits of Planning in Cities. Introduction: Design, emergence and Somewhere In-Between. London, UK: Earthscan. Hurriyet. (2006). İstanbullular Bu Vapuru seçti. Retrieved February 2015, from hurriyet.com.tr: http://www.hurriyet.com.tr/gundem/9490965.asp hurriyet. (2011). “Taksilerde Devrim Olacak”. Retrieved 2011, from hurriyet.com.tr: http://www.hurriyet.com. tr/gundem/17339926.asp IBB. (2010). Istanbul Icin Taksi sistemi Tasarimi Yarismasi Onayi. Istanbul. IBB_DesignBrief. (2011). Istanbul Taxi Design Competition Design Brief. Istanbul. IBB-Rec1. (2011). (TUT1) Istanbul Taksi Tasarim Yarismasi Juri Degerlendirme Tutanagi-1(Istanbul Taxi Design Competition Jury Record1). Retrieved 2011, from Istanbul Taksi

Tasarim Yarismasi: http://taksitasarim. ibb.gov.tr/Sayfalar/Duyurular.aspx IBB-Rec2. (2011). (TUT2) Istanbul Taksi Tasarim Yarismasi Juri Degerlendirme Tutanagi-2(Istanbul Taxi Design Competition Jury Report2). Retrieved 2011, from Istanbul Taksi Tasarim Yarismasi Web Sitesi: http:// taksitasarim.ibb.gov.tr/Sayfalar/ Duyurular.aspx IBB-Rec3. (2011). (TUT3) Istanbul Taksi Tasarim Yarismasi Juri Degerlendirme Tutanagi-3 (Istanbul Taxi Design Competition Jury Records3). Retrieved 2011, from Istanbul Taksi Tasarim Yarismasi Web Sitesi: http:// taksitasarim.ibb.gov.tr/Sayfalar/ Duyurular.aspx IBB-Rec4. (2011). (TUT4) Istanbul Taksi Tasarim Yarismasi Juri Degerlendirme Tutanagi-4(Istanbul Taxi Design Competition Jury Record4). Retrieved 2011, from Istanbul Taksi Tasarim Yarismasi Web Sitesi: http:// taksitasarim.ibb.gov.tr/Sayfalar/ Duyurular.aspx IBB-Rec5. (2011). (TUT5) Istanbul Taksi Tasarim Yarismasi Juri Degerlendirme Tutanagi-5(Istanbul Taxi Design Competition Jury Record5). Retrieved 2011, from Istanbul Taksi Tasarim Yarismasi Web Sitesi: http:// taksitasarim.ibb.gov.tr/Sayfalar/ Duyurular.aspx Kjaer, M. A. (2009). Globalization and Urbanism in the Non-Western World: Governance and Urban Bureaucracy. Sage. Lootsma, B. (2011). From Pluralism to Populism: Architectural Criticism in Times of Internet. Serbian Architectural Journal 3, 254-269. Maile, M. (2012). Design Criticism For the 21st Century. Retrieved September 25, 2012, from www.core77. com: www.core77.com/blog/education/design_criticism_for_the_21st_ century_11236.asp McGrath, G. R. (2011). Failing By Design. Retrieved April 5, 2013, from Harvard Business Review: https://hbr. org/2011/04/failing-by-design McGrath, G. R. (2011). Failing By Design. Harvard Business Review. milliyet. (2013). http://ekonomi.milliyet.com.tr. Retrieved August 5, 2013, from milliyet: http://ekonomi.milliyet.com.tr/istanbul-da-taksi-plakasi-

ITU A|Z • Vol 12 No 3 • November 2015 • H. Ekinoğlu, G. Turan


primde-yuzde-40-i buldu/ekonomi/ ekonomidetay/31.12.2011/1482692/ default.htm Nasar, J. L., & Kang, J. (1989). A Post-Jury evaluation: The Ohio State University Design Competition for a Center for the Visual Arts. . Environment and Behaviour, 21, 464-484. nyc.gov. (2011, April). Taxi of Tomorrow. Retrieved May 20, 2011, from http://www.nyc.gov/html/media/totweb/taxioftomorrow_home.html NYCTLC, N. Y. (2011). Taxi of Tomorrow Survey Results. NYC: NYCTLC. NYTimes. (2013). Court Blocks City’s Plan for ‘Taxis of Tomorrow’. Retrieved October 15, 2013, from nytimes.com: http://www.nytimes.com/2013/10/09/ nyregion/judge-blocks-new-york-cityplan-for-taxi-of tomorrow.html?adxnnl=1&adxnnlx=1381862412-wukwgX0OPZUTEsOa7ZMuFw Sanders, E. (2002). From User-Centered to Participatory Design Approaches. In J. (Ed.), Design and the Social Sciences. Taylor & Francis Books Limited. Sanoff, H. (1990). Participatory Design, theory and Technics. North Caroline State University, USA. Sanoff, H. (2006). Multiple Views of Participatory Design. METU Journal of the Faculty of Architecture.

TUBITAK, & Cetin&Oguz. (2007, 2010, 2013, 2008). Gathered from Various Sources. Volker, L. (2010). Designing a Design Cpmpetition: The Client Perspective. Retrieved December 2014, from www.designresearchsociety.org (The proceedings of the DRS 2010 conference in Montreal): http://www. designresearchsociety.org/docs-procs/ DRS2010/PDF/125.pdf Wagner, J. (2012). NYC gets yesterday’s taxi, not ‘The Taxi of Tomorrow. Retrieved May 12, 2012, from jameswagner.com: jameswagner. com/2012/04/post_28.html Wulz, F. (1986/1990). “The Concept of Participation” edited by Henry Sanoff in his book of “Participatory Design, Theory & Techniques” . North Caroline State University, USA. www.ny.com. (n.d.). ny.com. Retrieved June 2013, from http://www. ny.com/transportation/taxis/ www.taksitasarim.ibb.gov.tr. (2011, March). Retrieved from Istanbul Taksi Design Competition Web site: www. taksitasarim.ibb.gov.tr ZeroIstanbul. (2012). Zero Istanbul Design Guide. Retrieved February 2015, from http://www.zero.eu/: http://www. zero.eu/download/istanbuldesignguide/istanbuldesignguide2012.pdf

İyi tasarımı “favori tasarım” yoluyla aramak: İstanbul ve New York taksi tasarım yarışmaları

toplum arasındaki kesin çizgilerin bulanıklaşması olarak da tanımlarken söz konusu ağ yapılanmasının başarısını; ilgili tüm aktörleri dahil edebilmesi ve hedeflenen amacı gerçekleştirmeye dönük kurduğu ilişkileri ile de orantılar (Kjaer, 2009, p. 138). 1960’larda Christopher Jones, Christopher Alexander, Tom Markus ve Ray Studer gibi rasyonalist düşünce temsilcisi mimarlar bir takım matematiksel modellerin tasarımcıları daha iyi karar alma süreçlerine götürebileceğini savunmaktaydı (Broadbent, 1980). Ancak 1970’lerde oldukça rasyonalist bu önerilerin aslında muazzam bir değerler ve öncelikler çeşitliliğini içinde barındıran bir dünyaya ait problemleri doğru tanımlamada yetersiz kaldığı görüşü önem kazandı (Comeiro, 1990). Günümüz çok sesli ve çok katmanlı modern toplumları için kent hayatına dair karmaşık

Son bir kaç on yıldır ulusal ve kentsel bürokraside yaşanan değişimler “Yönetimden” “Yönetişime” veya “Eski Yönetim” biçiminden “Yeni Yönetişim” biçimine geçiş olarak tanımlanmaktadır (Kjaer, 2009). Bu aynı zamanda çeşitli ortak özel amaçlar için bir araya gelen yerel yöneticiler, kamu idarecileri, akademisyenler veya kar amacı gütmeyen sivil aktörlerden oluşan bir ağ yapısının hedeflenen belli bir amaca dönük tanımlanan bir organizasyon yapısı olarak kurulmasıdır. Literatürde bu şekilde özel bir amaca dönük yatay ve dikey hiyerarşileri olmayan esnek ve çoğulcu örgütlenme biçimi “Yerel Yönetim” den, “Toplumsal Yönetişim”e geçiş olarak da tariflenmektedir. Yönetişim teorisi bunu kamu, özel ve sivil

Does favorite design lead to good design?: Taxi design competitions in Istanbul and New York City


sorunlara çözüm üretmede geleneksel yukarıdan-aşağı tek yönlü karar alma süreçleri yetersiz kalmaktadır. Bu yetersizlik kendini en çok geniş halk kitlelerini ve kent hayatını doğrudan ilgilendiren tasarım problemlerinin doğru tanımlanması ve iyi tasarım önerileri geliştirilmesi ihtiyacında göstermektedir. Bu ihtiyacın beraberinde getirdiği sorunsallardan en önemlisi ise “çesitlilik” olarak kendini göstermektedir. Toplumsal yapı farklı sosyal, kültürel, antropolojik ve duygusal bağları olan yüksek bir çeşitliliği bünyesinde barındırmaktadır. Hepimizin farklı estetik ve ahlak/etik kodları bulunmaktadır. Bu nedenledir ki; kamu çalışmalarında eski kamu yönetiminden yeni kamu yönetimi anlayışına kayış; çoğulcu ve tartışmalara açık platformların kamu geleneği içinde kendini göstermesiyle birlikte yeni bir paradoksu tetikledi: Çoğulculuk Meselesi. Çoğulculuk katılım için olmazsa olmaz bir girdi olmasına rağmen bir takım zorlukları da beraberinde getirmektedir. Kendi içinde yüksek bir çeşitlilik ve çok seslilik barındıran bir kent toplumunun günlük yaşam standartlarını ilgilendiren tasarım problemlerine çözüm ararken genel-geçer doğru ve yanlışlardan söz etmek ve herkes için işleyen bir çözüm geliştirmek güç bir durum halini almaktadır. Çünkü toplumsal çeşitlilik ve farklılıkların tasarımdaki karşılığı yaklaşımların, fikirlerin ve önerilerin çeşitliliği olarak ortaya çıkmaktadır. Bizler tasarım fikirlerimizi yarıştırırken diğer yandan değerlerimizi ve toplumsal, sosyal ve mekansal yorumlarımızı da yarıştırmaktayız. Tüm bu gerilimi, eğitimin, ideolojinin, etnik yapının, toplumsal cinsiyetin, dinin ve yerelliğin birbiriyle örtüşmeyen, rekabet eden veya çelişen kabulleri içinde teşhis etmekteyiz (Bellamy, 1999, p. 1). Bu makalede “iyi tasarım arayışı” ile “popülerlik / favori olma durumu” ya da daha anlaşılır bir ifadeyle “halk tarafından belirlenme ve benimsenme” olgusu arasındaki gerilim, İstanbul ve New York kentlerinin taksi tasarım yarışmalarının yöntemleri ve sonuçları üzerinden sorgulanmaktadır. Tasarım yarışmaları bu çesitliliği tasarıma girdi olarak değerlendirirken halk oylamaları herkesi kapsamayı hedefleyen çoğulcu bir karar alma yön-

temi olarak karşımıza çıkmaktadır. Bu açıdan değerlendirildiğinde katılımcı tasarım kuramlarına göre tasarım yarışmaları ve halk oylamaları “birlikte karar alma/co-decision” modeli olarak tanımlanmaktadır (Wulz, 1986/1990, p. 41). Bu makalede anlamlı benzerlikleri olan iki örnek olay olarak, “İstanbul için Taksi Tasarım Sistemi Yarışması” ve “New York için Geleceğin Taksisi / Taxi of Tomorrow for New York”, tasarım yarışmaları ve halk oylamaları yukarıda tanımlanan “birlikte karar verme” katılımcı tasarım uygulamaları örnekleri olarak incelenmiştir. Her iki tasarım yarışması kendine özgü farklılıklara sahip olmakla birlikte özellikle katılımcı tasarım uygulamaları açısından benzerlikler arz etmektedir. İki yarışmanın birbirinden ayrılan en belirgin özelliği, yarışmaya katılabilecek kişilerin nasıl tanımlandığıdır. İstanbul’un yarışması özgün fikirleriyle katkı sunabilecek her kesimden insanı -orta öğretim öğrencileri, üniversite ekipleri, profesyonel tasarımcılar kategorileriyle - bağımsız tasarımcılar olarak tasarım sürecine dahil ederken; New York yarışma komitesi, tasarım sürecini davet ettiği otomotiv sektörünün temsilcileri, Nissan, Ford, Karsan gibi profesyonel tasarım oluşumlarıyla sürdürmektedir. Her iki yarışmada finalist tasarım önerilerinin yarışma web siteleriyle halk oyuna sunulması izlenen temel ortak katılım yöntemidir. Makalenin argümanı İstanbul’un halk oylamasında profesyonel kategoride genel amaçlı taksi tasarımı için en yüksek oyu alan, diğer bir deyişle “en favori” tasarım, New York’un halk oylamasında ise en beğenilen, yani en yüksek oranda “love it” oyunu alan, tasarım önerileri ve bu tasarımlara dair tartışmalar üzerine kurulmuştur. Her iki örnek olayda ortaya çıkan durum şöyle özetlenebilir: Popüler beğeni ile iyi tasarım arayışı arasındaki mücadelenin, temelde aynı amaca dönük arayışlar gibi gözükseler de, farklı kaygıları bulunabilmektedir. Daha konforlu, daha “modern” ya da “güncel”, daha güvenli ve kent kimliğine katkıda bulunacak başarılı taksi tasarımlarının kent yaşamının standartlarını yükseltmeye dönük, katılım boyutu ön planda olan bir “iyi tasarım arayışı” sürecinin veya yarışmasının toplum tarafındaki

ITU A|Z • Vol 12 No 3 • November 2015 • H. Ekinoğlu, G. Turan


algılanışı ile hedeflenen amaca dönük ağ yapısını kurgulayan ve yöneten yerel yönetim tarafındaki algılar, kaygılar ve anlayışlar farklılık arz edebilmektedir. Bu makalede incelenen örnek olaylar şu ilişkiyi açıklamaktadır; kullanıcının tasarım sürecine katılım şansı ve süreci etkileyebilme yetisi, onun süreçten beklentisini doğru orantılı olarak etkilemektedir. Bir diğeri de kentsel yaşam kalitesini ilgilendiren tasarım problemlerinde tasarım sürecine, tasarımcı veya en beğendiği tasarım için oy kullanan kullanıcı olarak katılım sağlayan aktör ile süreci kurgulayan ve yöneten aktörün süreçten beklentisi arasındaki farklılık ne denli fazla ise de, elde edilen sonuçların toplam başarıya etkisinin o denli az olduğu görülmektedir. Bir diğer deyişle; iyi tasarım arayışını “en favori veya en popüler olan tasarım” üzerinden kurgulayan süreçlerin yukarıda sayılan nedenlerden dolayı bazen ne hedeflenen tasarımın kendisini ne de süreci başlatan, ağ yapısını oluşturan ve organize eden aktörü yüceltmeyebileceğini göstermektedir. Sonuç olarak, bu çalışma ile irdelenen argüman ve örnek olay incelemeleri, hedeflenen şey geniş kitleleri

ve kent yaşamını ilgilendiren tasarım konuları olduğunda, yönetişim olgusunun teorik ve teknik tanımının yetersiz kaldığına işaret etmektedir. Her iki örnek olaydan elde edilen çıkarımlar, ulaşılmak istenen amaç iyi tasarım olduğunda, “popüler olan” veya “en beğenilen” tasarımların bizi her zaman iyi tasarıma götürmeyebileceği gibi, süreci kurgulayan ve yöneten aktörün/ ağ yapısının süreçten olumsuz etkilenmesine de yol açabildiğidir. “En beğenilen” veya “herkesçe en popüler olan tasarım” her zaman iyi tasarım mıdır? sorusuna cevap ise “hayır” olmaktadır (Maile, 2012) (McGrath, 2011). Makalede karşılaştırmalı olarak incelenen iki örnek olay da buna işaret etmektedir. Sorgulanan argüman özelinde incelenen örnek olaylar ile şu sonucun olumlandığı görülmektedir: Çoğunluğun beğendiği ancak henüz üretilmemiş ve deneyimlenmemiş tasarım her zaman iyi tasarım olmayabilir. Öte yandan uzmanlarca iyi tasarım olarak tanımlanmış ve üzerinde uzlaşılmış tasarımlar ise farklı kriter, beğeni ve deneyimlere sahip kullanıcılarca en beğenilen veya en favori olan tasarım olarak kabul görmeyebilir.

Does favorite design lead to good design?: Taxi design competitions in Istanbul and New York City

ITU A|Z • Vol 12 No 3 • November 2015 • 227-247

Continuity of architectural traditions in the megaroid buildings of rural Anatolia: The case of Highlands of Phrygia Alev ERARSLAN aleverarslan@gmail.com • Department of Architecture, Faculty of Architecture, Istanbul Aydın University, Istanbul, Turkey

Received: May 2015

Final Acceptance: October 2015

Abstract Rural architecture has grown over time, exhibiting continuities as well as adaptations to the different social and economic conditions of each period. Continuity in rural architecture is related to time, tradition and materiality, involving structural, typological, functional and social issues that are subject to multiple interpretations. This fieldwork was conducted in an area encompassing the villages of the districts of today’s Eskişehir Seyitgazi and Afyon İhsaniye districts, the part of the landscape known as the Highlands of Phrygia. The purpose of the fieldwork was to explore the traces of the tradition of “megaron type” buildings in the villages of this part of the Phrygian Valley with an eye to pointing out the “architectural continuity” that can be identified in the rural architecture of the region. The methodology employed was to document the structures found in the villages using architectural measuring techniques and photography. The buildings were examined in terms of plan type, spatial organization, construction technique, materials and records evidencing the age of the structure. The study will attempt to produce evidence of our postulation of architectural continuity in the historical megara of the region in an effort to shed some light on the region’s rural architecture. The study results revealed megaroid structures that bear similarity to the plan archetypes, construction systems and building materials of historical megarons in the region of the Phrygian Highlands. These structures were classified in a typology that evidenced the existence of an architectural continuity of megaroid building tradition, which this study seeks to present. Keywords Megaron, Megariod buildings, Architectural continuity, Rural architecture.


1. Introduction Rural architecture is a type of architecture that is based on local needs and construction materials, and it reflects local traditions at the same time. The general characteristics of rural architecture are traditionality, functionality, adaptation to environmental conditions and local materials. Rural houses are important elements of heritage that have historical and cultural continuity. They are less liable to be affected by rapid cultural changes and has grown over time, exhibiting continuities as well as adaptations to the different social and economic conditions of each period. Some part of Turkish communities preserved their nomadic lifestyle until the end of the 19th century when they began to abandon their yörük tents to build permanent houses during their transition from a nomadic existence into a settled lifestyle. It is believed that it was in this period that they must have adopted the housing plans used by the local populations in the places in which they settled. 2. The aim and the methodology of the research The process of exploring Anatolia’s rural settlements is still in its beginning stages. There is still controversy over which parameters were influential in the choice of the house plans, materials and building systems used in the transition of the Turks into permanent settlements. The general belief is that the nomads adapted to the housing culture of indigenous societies in that period. It is for this reason that research on Anatolian rural architecture is of great importance. This fieldwork was conducted in the area encompassing all villages of the districts of Eskişehir Seyitgazi and Afyon İhsaniye, referred to in history as the “Highlands of Phrygia” (Figures. 1, 2). The main goal of this research is to reveal the similarities between the megaroid structures in the Highlands of Phrygia and the plan, spatial organization, building systems and materials used in the historical megara, and to uncover any architectural continuity as is believed to exist in the area. This will shed some light on the question

Figure 1. Eskişehir Seyitgazi. Researched villages (drawn by Uğur Süleymanoğlu).

Figure 2. Afyon Ihsaniye. Researched villages (drawn by Uğur Süleymanoğlu).

of whether there is in fact cultural and architectural continuity in Anatolian rural architecture. To reach our goals, we investigated all of the villages in the region. All the houses were examined in terms of the plan type, spatial organization, materials, construction systems and records evidencing the age of the buildings. The samples chosen were documented

ITU A|Z • Vol 12 No 3 • November 2015 • A. Erarslan


using architectural measuring methods and photography. In addition, interviews were held with the members of the households living in the houses and with village elders to learn about the age of the buildings, the history of the villages, the daily lifestyle and the use of space. 3. The historical background of the megaron Megaron (plural megara) refers to an elongated rectangular building with an entrance on one of the short sides, provided with a porch. In Homer, the megaron refers to the great halls of the Mycenean palaces (Knox, 1973). Homer regards the function of the megaron as “the hall of the men”. Heredotus holds it equal to the sacred room of the adyton of the temple dedicated to Helen (Işık, 1998). According to Deroy, “megaron” is a word in Sanskrit that means “a room with a hearth” (Deroy, 1948). Dörpfeld, Schliemann and Blegen were the first to apply the term “megaron” to prehistoric remains, in the palace at Tiryns in 1885 and later for the large buildings of Troy II (Dörpfeld, 1902); Schliemann, 1885). They used the term “megaron” in the Homeric sense of a large hall or a main hall in a palace (Ivanova, 2013). In his 1953 excavation report, Blegen described the megaron as “a room of great size, the principal apartment of the palace”. It is referred to as a megaron of the classic mainland type, consisting of a great hall, a vestibule, and a two-columned portico fronting a court, in most respects similar to corresponding suites at Mycenae and Tiryns (Blegen, 1953). The term has subsequently been used to refer to other buildings in Greece and elsewhere that contain a long hall fronted by a porch, as well as freestanding buildings with this alignment of rooms (Warner, 1994). There are different views about the roots of the megaron. The simplest type, namely an isolated rectangle, is attested for Thrace, Macedonia and Thessaly in Neolithic (Müller, 1944). Buildings of this type consist of a single room without porch or anteroom. They have pitched roofs, flat roof or the barrel roof.

Other early examples can be seen in Sesklo and Dhimini in Eastern Thessaly in the late Neolithic (Bintliff, 2012). Sesklo has in fact been named the earliest “megaroid-style building” in the middle of the acropolis. The structure is made up of a porch, a main chamber and a back room. In the same way, the acropolis at Dhimini boasts a megaron larger than others that stands in the middle of the circular walls. These are centrally-located structures and protected by fortification walls. They have been interpreted as the rulers’s residences or the temples (Bintliff, 2012). Poliochni and Thermi had row-houses of megaroid character during EBI. These long and narrow buildings are composed of a closed antechamber and of a main chamber. These structures have been placed on the street perpendicularly. They were constructed as row houses with common side walls (Warner, 1994). Mellink presumes that these long houses are the ancestors of the megaron (Mellink, 1986). Lerna IV (Early Helladic II) is a small one-room megaron of classical form facing east onto a large courtyard (Warner, 1994). In the Middle and Late Helladic, the megaron plan continued to develop and a number of new types emerged, particularly in houses. In the Peloponesus, the Mycenaean palaces of ruler forts of Late Helladic III, Mycenae, Tiryns and Pylos exhibit structures of the megaron type. The most well-known is Nestor’s palace at Pylos, of which we hear much in Homer’s Odyssey. It consists of a hall, a forehall, and a porch with two columns in antis to support the roof. The main hall contains a large circular hearth at center, surrounded by four columns (Blegen, Rawson, 1966). Looking at Anatolian examples of megaron-type, Hacilar IIA from the Early Chalcolithic presents buildings of megaroid character. These contiguous buildings of one or two stories have features similar to megaron type because they display a forecourt of a sort. The houses are arranged with their backs to the defensive wall. Each consists of a main room with a hearth set in the middle of the floor and an anteroom (Mellart, 1970). Another precurser of the mega-

Continuity of architectural traditions in the megaroid buildings of rural Anatolia: The case of Highlands of Phrygia


ron-like structure in Anatolia can be seen in Yümüktepe XVI. Here stand a series of houses of megaroid character adjoining the city walls, all with a closed porch and a main room in back (Garstang, 1953). At the start of 4000 BC in the Late Chalcolithic, in Beycesultan XXIV, appeared a structure that may be said to be a precursor to the megaron type in Anatolia. Here, at the west end of a building, projecting walls created an open porch; from this, one entered through a doorway with a raised step into the main chamber, which had a circular hearth in the center (Lloyd, Mellaart, 1962). An increase in megaron-type buildings appeared in Anatolia by the EB. Troia I consisted of parallel rows of long buildings of megaroid character. House 102 has one room and a porch and without rear antae. It had a hearth at the center (Blegen, 1937, Ivanova, 2013). At Troia II, large houses of the megaron-type have been brought to light on the citadel. Megaron IIA with a central hearth probably served as an assembly room or an audience-hall, and maybe in its last phase (Ilh), it was a place of cultic activity (Mellart, 1959). The long-room units arranged in rows were very common in the western part of Anatolia in the EB. The coastal communities along the Anatolian littoral at Beşiktepe, Bakla Tepe and Liman Tepe VI, and on the eastern Aegean islands at Yeni Bademli, built mainly long-room dwellings arranged in a row (Ivanova, 2013). Some of these long houses may be defined as of megaroid-style because of their side wall extensions (Erkanal, 1996). The most numerous examples of the megaron plan were uncovered in an EB village at Karataş-Semayük. The basic characteristics of these freestanding rectangular structures are two structural long walls with cross-walls inserted to form a main room and a front porch which are entered axially on the short side. The long walls end in antae at the front; the rear cross-wall is often set back from the ends of the long walls, which thus project as rear antae (Warner, 1979, 1994). Each is entered through a door centrally located in the front cross wall between the porch and

the main room. The roofing sytem is the gabled roof. In all of the occupation levels dated to the EB at Beycesultan, megaroid houses and shrines are quite prevalent. They have main rooms with hearths, sometimes with small rooms in the back and with a porch. This continued to be a feature of megaron-type buildings right up to the end of the LB at site (Lloyd, Mellaart, 1962). The Antalya Bademağacı, Eskişehir Küllüoba, Demircihöyük and Keçiçayırı settlements of the EB had two-roomed structures in the megaron style. They contained a central open area around which there were examples on a radial plan adjoining the city walls. In Bademağacı, the EBA II town were megaron-like houses with openporched and with rear rooms (Duru, 2003; Korfmann, 1983). Küllüoba in Early Bronze II consists of an upper city, at the center of which there are two megara complexes that have public functions. Surrounding these structures on three sides are long houses and two- or three-roomed houses in megaroid shape, their rear rooms abutting against the fortification wall (Efe, Ay-Efe, 2001; Efe et al. 2011). The megaron-type in Marmara however can be seen at the end of the EB. The acropolis of Kanlıgeçit has three adjacent large megara arranged linearly. These buildings are observed to be of the type that has a single main room with an open porch in front with rear antae (Özdoğan, 2002). Phrygian architecture represents the period in which the megaron plan prevails as the most characteristic plan type. The buildings in the citadel in Gordion are freestanding and each consisted of a large hall with central hearth and a porch and anteroom, and with a flat or double pitched roof (Young, 1960). The megaron-type can also be established in central Anatolia. In the period of Phrygian expansion to the east, the Iron Age of Büyükkale, small megaron houses are typically of Phrygian construction and plan. The most frequent type displayed an open entrance hall/ porch in antis. There are also other variations with semi- or completely closed porches (Neve, 1996).

ITU A|Z • Vol 12 No 3 • November 2015 • A. Erarslan


At Kültepe in the EBIII, a megaron-shaped architectural complex is a temple. In the middle of a large room lies a round hearth, surrounded by four posts. Smaller rooms are grouped around this central hall (Özgüç, 1963). A megaron-shaped architectural complex was also found on Stratum IIa in Kaman Kalehöyük. The structure was surrounded by corridors (Omura, 1999). At Kerkenes Dağ, some structures that can be identified as megaron have been uncovered. They are freestanding buildings that have one main room with a central hearth and an open porch (Summers et al., 2004). They have double-pitched roofs and served some special public function or the residences for the ruling elite (Summers et al., 2004). Tell Tayi’nat Building XVI was a long room divided into a portico, a main hall and a shrine. It is a temple complex and its plan is in antis style (Harrison, Osbourne, 2012). Besides these examples, some researchers assert that there are buildings of the megaroid character in the Near East as well. Some have named the long-axis corridor house/pierhouse type of house found to be widespread in the southern Levant in Middle PPNB “megaron” (Wright, 1985; Garstang, Garstang, 1940). A central hearth is a common feature of the plan. Wright states that the PPNB temple at Jericho E is a megaron since the flank walls project as antae (Wright, 1985). Wright also claims that the Langbau type of temples in the Near East in the LB and Iron Ages are of the megaron style (Wright, 1985; Davey, 1980). They have a single long room and a shallow entrance porch designed in antis (Hundley, 2013). 4. A history of the settlement of the region The Highlands Phrygia cover the whole of the districts of Afyonkarahisar and the districts of Ihsaniye, Işcehisar and Bayat as well as a part of Bolvadin and Seyitgazi and Han in Eskişehir, and a part of central Kütahya and Tavşanlı (Aşılıoğlu, Memlük, 2010; Haspels, 1971; Kortanoğlu, 2011) (Figure 3). The region has been named

Figure 3. The Highlands Phrygia (https:// www.academia.edu/1670748/Highlands_ of_Phrygia-Map).

after the Phrygians. Ancient Phrygia was a neighbor to Cappadocia to the east and later to Galatia, the regions of Lykaonia, Pisidia Kabalis, Milyas Kibyratis in the south, and Mysia, Lydia, Karia to the west and the regions of Bithynia and Paphlagonia to the north (Sevin, 2007). The capital Gordion, Pessinous, Mideon, Dorylaeum, Laodikeia and Kolossai may be mentioned as the most important cities of ancient Phrygia (Sevin, 2007). The oldest name for Afyon was Akronio. In the period of the Hittite Empire, Afyon gained importance because of the campaigns of Murshilish II against the kingdom of Arzava. It was after the fall of the Hittite Empire and following the ensuing Dark Ages that the Phyrgians entered the scene. The region was known as Phyrgia thereafter up until the end of the Byzantine Era (Ilaslı, 2004). As a result of the Cimmerian attacks, dominion over Central Anatolia passed from the Phyrgians to the Lydians (Akurgal, 2000). From the middle of the sixth century BC, the Persians captured sovereignty over the Afyon region. From 30 BC onward, Anatolia was under the leadership of Rome. During the Roman Era, new towns and cities were established in the region. The city of Amorium gained importance in the Byzantine Era and Phyrgia was divided in two, one part becoming

Continuity of architectural traditions in the megaroid buildings of rural Anatolia: The case of Highlands of Phrygia


annexed to Galatia, the other remaining as Phrygia Salutaris. During 1275-1343, Sahipataoğulları was established in Afyonkarahisar (Karazeybek, 2004). Yıldırım Beyazıt annexed the region to the Ottoman lands in 1390. After WWI, Afyon too was invaded by the Greeks. The city was liberated on August 27, 1922 in the Great Offensive. Seyitgazi is a small district of Eskişehir, 4 km. to its south. The oldest settlements in the area dates back to the EB (Altınsapan, 1999). The region was conquered by the Hittites in the fifteenth century BC and the Phyrgians entered the area around 1200 BC, establishing a strong kingdom in the environs of Eskişehir. In the eighth century BC however, with pressures from the Lydians, and because of the steady loss of strength as a result of the the Assyrian invasions, the Phrygian kingdom was destroyed by the Cimmerians, remaining under Lydian rule until the Persian invasion in 546 BC. In the Hellenistic Era, Rome began to reign in Anatolia and Seyitgazi, named “Nacolea”, became an important guard post of Rome (Aşılıoğlu, Memlük, 2010). After 395 AD, many Byzantine cities were established in the region. In the period of the Seljuks, Nacolea was conquered by the Danişments and the Seljuk tribes. Seyitgazi was annexed to Ottoman lands during the reign of Murat I. Seyitgazi participated in the War of Liberation with a special battalion; the troops were partially incapacitated in the Greek invasion and on 1922, with the coming of the Turkish armies, it became a part of the Turkish Republic (Altınsapan, 1999). 5. Plan typologies and general characteristics of the structures in the region In this work, the term “megaron” has been used to signify structures with 2 long walls ending in antae at the front, a front porch which is entered axially on the short side, and a rectangular single main room. However, to be precise, these structures have also been described as megaroid buildings, megaroid-style buildings, megaroid-shaped buildings, megaron-like buildings and the like.

The region’s megaroid buildings have a main room and a front porch, comprised thus of two structural archetypes. They constitute a longitudinal mass. The buildings vary in size. The main rooms are almost square or in the form of a longitudinal rectangle and form the core of the building. They are used for sleeping, eating and for most domestic activities. The main room sizes vary. They are usually of the size 3x3 m. None of the examples exhibit a central hearth. The hearth would occupy the wall across from the door. There are niches in the walls of the main room. The buildings can be directly accessed from the main room. In none of the examples is the door of the main room on the central axis, but either to the right or the left. The term “porch” is used to refer to the roofed area between the extensions of the long walls of the main room. The two long walls of the main room extended beyond the short front and constituted a porch at the front. The roof extended over the porch and provided additional working space. The porches are shallow in general. On the average, the depth of the porches is 1 m. In only a few examples, the porches are deep. The people of the region refer to the ground level porch as the “ev önü” (front of the house), and to the porch on top of the barns as the “hayat”. In no example are the porches paved with a special floor covering. None of the houses had a rear anta/rear porch. While most structures are one-storied, some buildings rise above a barn. Examples rising above a barn are few in the area. Some of the single-storied structures are built on ground level while some have been built on a stone basement. In this case, the structure has a raised porch with stairs leading to a porch. The stairs can be at the center of the house or on one side of the porch. The stairways open to the porch. The people in the region call the houses on ground level as “Yer Ev” (Ground House). The region has six types of megaroid structure. This typology was attained from a total of 33 buildings among the houses scanned for the research throughout the region (Figure 4, 5). Some of them are in ruins. Of those

ITU A|Z • Vol 12 No 3 • November 2015 • A. Erarslan


Figure 4. Graph showing the dispersion of the plan types according to villages in EskiĹ&#x;ehir Seyitgazi.

Figure 5. Graph showing the dispersion of theplan types according to villages in Afyon Ihsaniye.

Figure 6. Plan type 1 (Drawn by Salih Ceylan).

Figure 7. The megaroid building built in type 1A.

surviving, some have either been abandoned or converted into storage space. Rebuilding has occurred in some houses. The full age range of the houses is not known exactly. However, the average age of all of the examples, according to the testimony of the villagers, is 120-140 years. 1. Type with 2 antae: This is the most widespread plan in the region (Figures 4, 5). This type consists of a main room and a porch at the front (Figures 6-8). The side walls of the main room extend to form a porch at the front. They are antae walls. The antae form a solid wall. A building is built on a stone basement. In this case, the structure has a raised porch with stairs leading to a porch (Figures 6: B, 8). In general, these types of structures are freestanding on a road. 2. Type in antis: This is the second widespread plan (Figures 4, 5). In this type, the structures are single-roomed with an open porch and one, two or

Continuity of architectural traditions in the megaroid buildings of rural Anatolia: The case of Highlands of Phrygia


Figure 12. The megaroid building built in type 2B.

Figure 8. The megaroid building built in type 1B.

Figure 13. The megaroid building built in type 2C. Figure 9. Plan type 2 (Drawn by Salih Ceylan).

Figure 10. The megaroid building built in type 2A.

Figure 14. Plan type 3 (Drawn by Salih Ceylan).

Figure 11. Inside of the same house. The hearth.

three wooden posts placed between antae (Figure 9). In this form, the structures appear to be monostyle in antis, distyle in antis and tristyle in antis (Figures. 9-13). The monostyle in antis is the simplest form. The porches are very shallow and there are

Figure 15. The megaroid building built in type 3A.

ITU A|Z • Vol 12 No 3 • November 2015 • A. Erarslan


Figure 16. The megaroid building built in type 3B.

Figure 21. Plan type 4 (Drawn by Salih Ceylan).

Figure 17. The megaroid building built in type 3C.

Figure 22. The megaroid building built in type 4A.

Figure 18. The megaroid building built in type 3D.

Figure 23. The megaroid building built in type 4B.

Figure 19. The megaroid building built in type 3E.

Figure 20. The megaroid building built in type 3F.

no rear antae. In only one example of these buildings, which are generally on ground level, is the structure accessed with stairs consisting of a few steps (Figure 9: C, 13). While the structures are generally freestanding on a road, one of the buildings is attached parallel to the courtyard wall (Figure 10). 3. Type with a single anta: This is the third plan type in the region (Figures 4, 5). In this type, there is a single anta wall of the building. One of the antae forms the solid long side of the main room and one or more wooden post/ columns are found at the other end (Figure 14). Generally this type appears as a single wooden post/column

Continuity of architectural traditions in the megaroid buildings of rural Anatolia: The case of Highlands of Phrygia


Figure 24. The megaroid building built in type 4C.

Figure 25. Inside of the same house. The hearth.

Figure 26. The megaroid building built in type 4D.

on the other end of a solid long side wall which is one of the antae walls (Figures 14: A-D, 15-18). Sometimes 2 or 4 wooden posts may be found in front of the porch (Figures 14: E-F, 1920). In applications with a single post/ column, sometimes a wooden door opened to the courtyard is built on the post/column side of the porch (Figures 14: B-C, 16-17). Both freestanding and attached structures parallel to the courtyard wall structures can be seen at the location. In the structure on the location attached parallel to the courtyard wall, the courtyard wall forms the single anta (Figure 18). 4. Prostyle type: This is the other type in the region (Figures 4, 5). Prostyle is a term defining freestanding columns across the front of the building and refers to a building having

posts only along the front side. This type of building does not possess antae. The prostyle porch has been used in the region in the distyle and tristyle prostyle (Figure 21). In the distyle prostyle, a wooden post/column is found at each end of the porch in front of the structure (Figures 21: A-B, 22-23). In the tristyle prostyle, the structure has 3 wooden posts/columns in front (Figures 21: C-D, 24-26). In general, the prostyle type has been implemented in examples built on a high sub-basement or barn and these structures are accessed by stairs. The structures are freestanding on a road. 5. Closed porch type: This type of plan, which is only encountered in a single example in the region, comprises a main room with a double entrance and a closed porch/anteroom in front (Figures 4, 27: A, 28). It is without rear antae. The building is freestanding on a road. 6. Type with a rear room: In this plan, which is only uncovered in a single example, the main room is subdivided by a wall to create a back room (Figure 4). A partition wall divides the interior into two rooms. Thus, a small rear room is formed in back of the main room of the structure (Figures 27: B, 29), the long main room being flanked by the smaller room at the back. There is a deep porch with a single anta in front of the structure. The building is attached parallel to the courtyard wall. While the courtyard wall forms the single anta of the structure, in the other direction, 3 wooden posts stand in front of the porch (Figure 29). These megaroid-style buildings are characterized in three locations; freestanding on a road or on a courtyard, adjacent or abutting another structure and attached to parallel to the courtyard wall. In the first type of location, the structures are freestanding and self-contained structures on a road or on a courtyard. They are single unit structures. Their porticos have been built to face the south. Their orientation is looking out toward open spaces and streets (Figures 8, 17, 23, 24, 26). In the second type of location, the structures are buildings where a few families live together and which also

ITU A|Z â&#x20AC;˘ Vol 12 No 3 â&#x20AC;˘ November 2015 â&#x20AC;˘ A. Erarslan


have a barn, a hayloft and stand in a courtyard along with other houses. These buildings are either freestanding inside the courtyard, looking out toward the road, or they have been built attached to parallel to the courtyard wall. In this case, one wall of the house is also the wall of the courtyard (Figures 10, 18, 29). In other words, the courtyard wall forms one of the antae of the building. In the third type of location, the structure is leaning on another structure at the back. The building usually leans on another structure (an annex), most likely a barn or hayloft. Only in very few examples, this annex is a house (Figure 17). The annex is not entered from the main building. It is separately roofed. There is only one example of an annex added to the long side of the building, but the construction of this indicates that the annex was built later (Figure 20). The organization of the façades of the buildings is made up of a door and one adjacent window opening out into the main room. With this arrangement, the buildings have a two-element structure on their façades. Only in some examples were the windows on the façades closed off afterwards.

Figure 27. Plan type 5 and 6 (Drawn by Salih Ceylan).

Figure 28. The megaroid building built in type 5.

The side walls of the buildings too have windows but some structures remain windowless. The rear façades of the buildings however are without windows. All of the megaroid structures in the region have a roof that is one of 3 types—flat roof, low-pitched roof or gabled—but all of the types are wooden. In all the types of roof, the roof extends over the porch. In the case of the flat roof, some of these are sometimes tiled with bricks but sometimes covered with earth. The low-pitched roof and the gabled roof however are tiled with brick. Because the rooms are small, no internal wooden posts or a central post carrying the roof in the main room have been encountered. Two types of wall technique have been used on the walls of the structures- stone wall construction and timber-reinforced stone construction. In the timber-frame supported stone wall, walls were supported by a wooden framework of horizontal, transverse and vertial timbers. While rubble stone was generally used as material for the walls, it can also be seen that both rubble stone and finely cut stones were used together. In some buildings, mudbrick was used along with the stone. In addition to this mixed material, mixed wall construction can also be observed. In very few examples, timber-frame stone and mud-brick construction can be seen together in a mixed wall construction system. In one example, bağdadi (lath and plaster) was used on one anta wall of the structure (Figure 7). Sometimes the walls were covered with clay and straw plaster. None of the houses have toilets. All of the toilets are outside. In one corner of the main room stands a wooden platform that serves as a bath.

Figure 29. The megaroid building built in type 6.

Continuity of architectural traditions in the megaroid buildings of rural Anatolia: The case of Highlands of Phrygia


6. Comparison of the historical megara When the megaroid buildings in the region are compared with historical megarons of ancient Anatolia and Near East, it can be noticed that the plan types, construction systems, building materials and roofing sytems display some common elements. The first plan type in the region is the “type with 2 antae”. This type consists of an almost square or longitudinally rectangular main room and a porch at the front (Figure 6). There are no rear antae. Similar types of this plan can be seen in ancient Anatolia in Troia I (House 102), Troia II, Karataş-Semayük, Bademağacı, Büyükkale, Beycesultan X-IX, Keçiçayırı, Gözlükule Tarsus and Kerkenes. All have megaroid structures that are single-roomed, with an open porch and without rear antae. The side walls of the main room extend to form a porch at the front and they constitute the antae walls. So, the antae walls of the buildings are the solid side walls of the main room. In general, these types of buildings are freestanding. Some of the buildings are attached to each other because of the “Anatolian settlement siedlungschema” (Blegen, 1937; Neve, 1996; Summers et al., 2004; Efe et al, 2011; Lloyd, Mellaart, 1962; Warner, 1979; Naumann, 1998; Duru, 2003). Another plan in the region is the “type in antis”. In this type, one, two or three wooden posts are placed between antae (Figure 9). The porch is very shallow and there are no rear antae. These structures display a porch arrangment of monostyle, distyle and tristyle in antis type. This type porch arragement resembles the façades of Phrygian rock-cut shrines and Phrygian, Lycian, Hellenistic and Roman periods rock-cut tombs (Kortanoğlu, 2011). It was Ch. Fellows and later Benndorf and Niemann who asserted that these tombs might have been influenced by wooden houses in ancient Lycia (Işık, Yılmaz, 1996). The façades of the tombs resembled house façades because the tombs were considered the residences of the dead (Ambrossini, 2011). Thus houses began to be seen as the precursers of the Lycian rock tombs and Phrygian rock-cut shrines

and tombs (Kjelden, Zahle, 1975). In the book he wrote in 1853, “Ein Ausflug nach Kleinasien und Entdeckungen in Lycien”, Ch. Fellows drew pictures of the Turkish houses and storehouses he saw in the Xanthos plains and he called attention to the resemblance between these houses and the wooden house architecture of ancient Lycia and the Lycian rock tombs (Figure 30) (Işık, Yılmaz, 1996). These house tombs imitated the appearance of wooden Lycian houses, with their stone reproductions of wooden architectural features (Metzger , Coupel, 1963). These tombs have the same shape, all small temples with pediments supported by columns. Their façades have a pediment and columns between the projecting side walls (antae). They generally have façades with an arrangement of 1, 2, 3 or 4 columns between antae (in antis). Besides having gabled roofs, there are also examples of flat roofs. The Phrygian rock-cut façades provide a clue about the megaron façade arrangements at Gordion. Some of them consist of a façade, varying in size, generally depicting the front of a house. The most prominent feature of the Phrygian rock-cut façades is a focal niche with a surrounding façade decorated with geometrical motifs. They are thought to have imitated the front of a building of public importance. The appearance of Phrygian houses may be gauged from the carved rocks representing the façades of buildings, probably temples, illustrated in stone in Arslankaya, Bahşayiş, Demirkale or Midas City (Barnett, 1967). Berndt-Ersöz assert that the rock-cut façades are not true copies of Phrygian houses, but may be imitations (Berndt-Ersöz, 1998; 2006). The porch in antis is also reminis-

Figure 30. 19th century houses and storehouses in the plain of Xanthos from Ch. Fellows (Işık, Yılmaz 1996: 178).

ITU A|Z • Vol 12 No 3 • November 2015 • A. Erarslan


cent of Greek temples. Greek temples generally made use of the “distyle in antis” plan. The houses in the region with the in antis plan also resembled the columned prostas house of the classical Greek period. Here, in front of the oikos was a columned porch (prostas) which opened out onto the courtyard. House plans with prostas were derived from megaron. Other similar examples of porch in antis can be seen in Ain Dara and Tell Tayinat Building II . These temples are buildings where the lateral walls of the main hall of worship are continued on the façade with a vestibule/porch on either side. Both have two columns between antae. Another plan encountered in the region is the “prostyle type”. This type does not possess antae. There is a wooden columned porch in front of the main room. This type has been used in the region in distyle and tristyle prostyle (Figure 21). In distyle prostyle porch, a wooden post/column is found at each end of the porch in front of the building. The distyle prostyle is reminiscent of the rock-cut tombs façade of the Hellenistic and Roman in Highlands of Phrygia (Kortanoğlu, 2011). The façade of Ayazini and Yapıldak, there is a distyle prostyle porch arragement (Figure 31). Here, a column is placed in each end of the porch. The prostyle porch is also reminiscent of prostyle Greek temples. A Greek prostyle temple has a colonnaded porch in front of the cella. There are however generally 4 columned porches in front of the cella. Besides in the Greek temples, the prostyle porch is encountered in the Neolithic wooden houses of the Cucuteni-Tripolye-Ariuşd cultural groups. In the houses of these cultures, a porch made up of wooden posts stands in front of the main room (Figure 32: A) (Laszlo, 2000). A shrine in Jericho, dated to the PPNP, is the other resemble of this type. Here on level XI, in front of a building that Garstang believes to be a shrine is a veranda-like vestibule supported by six wooden pillars (Figure 32: B). Garstang describes it as a prostyle porch (Garstang, Garstang, 1940; Banning, Byrd, 1988). However, both Cucuteni-Tripolye-Ariuşd and the shrine in Jericho have semi-antae,

a feature that is different from the prostyle types in our field of study. The other plan in the region is the closed porch type. This plan is only encountered in a single example. The building is without rear antae. It consists of a main room with a double entrance and a closed porch/anteroom in front (Figure 27: A). The closed porch is very widely used in the megaron style of buildings in Anatolia. Examples of a single-entrance main room and a closed porch/anteroom can be seen in Gordion, Küllüoba, Demircihöyük, Mersin, Hacılar IIA and Büyükkale. The main room with a double entrance is seen in Anatolia at Karataş-Semayük V-VI and in Bademağacı. At Bademağacı, in the EBII settlement, there are megaron-style buildings with a main room having double entrances with open porches (Figure 32: C, D). There is also a building with a double entrance and a closed porch/anteroom in front (Figure 32: E). At Semayük Karataş V-VI, although the main room in the megaron structures have a double entrance, these have open porches (Figure 32: F) (Warner, 1979). These buildings have no rear antae, but a rear room. Another type of megaroid structure in the region is the plan type that has a rear room. In this plan, the main room is subdivided by a wall to create a back room. Thus, a small rear room is formed in back of the main room (Figure 27: B). In front of the main room is an open porch. At Bademağacı, House 30 is a building with a rear room (Figure 32: G). At Karataş-Semayük V-VI, rear rooms are found in several houses. There are megaroid structures at Küllüoba with a closed porch and rear

Figure 31. Afyon Ayazini ( ht t p : / / w w w. w eb re h b e r i . n e t / y e re l / afyonkarahisar/#!prettyPhoto[gallery2]/1/).

Continuity of architectural traditions in the megaroid buildings of rural Anatolia: The case of Highlands of Phrygia


room. These rear/back rooms are used as storage facilities. The other type of plan in the region was the “type with a single anta”. In this type, one of the antae forms the solid long side of the main room and one or more wooden posts/columns are found at the other end (Figure 14). This type with a single anta wall is not encountered in the historical megara and appears to be a local characteristic completely unique to the region. The porches in the megaroid structures in the region are shallow in general. This type porch resembles the Anatolian megarons of Küllüoba, Semayük-Karataş and Bademağacı. These structures all have porches in front that are not very deep. However, there is some example of a considerably deep porch in the region; this example brings to mind a pronaos (Figure 14: D-E; 27: B). Megarons with deep porches in Anatolia are seen at Troia I (House 102), Troia II, Bademağacı, and Küllüoba. None of the megaroid structures in the region have rear antae. They have front antae only. Examples of megarons without rear antae are found in Anatolia at Karataş-Semayük, Bademağacı, Gordion, Küllüoba, Troia I (House 102) and the Troia II, VI. Three types of roof have been used in the region—a wooden flat roof, a low pitched roof and a gabled roof. In all of the roof types, the roof extended over the porch. Of the historical megarons, Gordion exhibits the use of 3 distinct groups of roof systems—the gabled roof, the pitched roof and the flat roof (Berndt-Ersöz, 2006). The megara at Gordion probably had gabled roofs, as indicated by a completely preserved poros akroterion whose lower parts follow the outline of a pitched roof, and three double-pitched poros blocks found at Gordion. Gabled buildings are also seen in the drawings incised on the exterior walls of Megaron 2, inscribed on potsherds from Midas City, and represented by three building models (Roller, 2009). In Troia, the roofs are flat. At Küllüoba too, because of the “Anatolian settlement siedlungschema” and since the houses are laid out in a row, it is hard to use a gabled roof. For this reason, the megarons in the

Figure 32. The comparision of the historical megaroid structures. A: A house of Cucuteni-Tripolye-Ariuşd culture. B: A shrine in Jericho. C, D, E: Bademağacı. F: Semayük Karataş. G: Bademağacı, House 30 (drawn by Salih Ceylan).

settlement are covered with flat roofs. The long houses built as independent structures, however, may have used the gabled roof form (Fidan, 2012). In Bademağacı for the same reason, the megaroid buildings are flat roof-covered. In freestanding buildings at Karataş Semayük too there is evidence that the gabled roof was used (Warner, 1994). In all of these historical examples, the roof extended over the porch. In the structures of the region, two types of wall technique are observed— stone wall construction and timber-reinforced stone construction. The wall materials were generally rubble stone but there are also examples of rubble stone and finely cut stone used together. In some buildings, a smooth mudbrick was used together with the stone. Besides these mixed materials, some examples also display a mixed wall construction. There are very few examples of timber-frame stone and mudbrick construction mixed together in a wall construction system. At Troia, the superstructure of the walls is mudbrick and supported by a wooden framework. At Bademağacı, the buildings are built on a stone foundation and mudbrick superstructures. Houses at Gordion are built of stone or crude brick, using a half-timber structure (Barnett, 1967). At Gordion, the walls are put together with a wooden framework and filled in-between with masonry screens, forming a skeleton for the construction (Young, 1962). Timber-reinforced mud walling construction can also be seen at Karataş Semayük (Warner, 1994). At Küllüoba, the remnants of wooden planks have been found between the stone foundations and the mudbrick wall above. The megaroid buildings in the region are generally one-storied. Some

ITU A|Z • Vol 12 No 3 • November 2015 • A. Erarslan


are on ground level but some rise above a sub-basement. A few have megaroid structures rising above a barn. The structures built on top of sub-basements are accessed with a few steps of stairs. At Tell Tayinat Building II and XVI, although the structures are on ground level, entrance was gained by means of a stepped porch, flanked by two columns in antis (Harrison, 2012). The annexes used as barns or haylofts in the megaroid buildings in the region are generally attached to the short side of the structure in the back but in only one example, an annex has been added to the long side of the building (Figure 20). In the historical megara, House 66 is the only example in Karataş-Semayük of an annex added to the long side of the building (Warner, 1979). The annex was probably built later both in Karataş and in the region. The annex was not entered from the main building and is separately roofed (Warner, 1979). The megaroid buildings in the region are located in three locations; freestanding, adjacent or abutting another structure and attached parallel to the courtyard wall. But, the historical megara (except the Anatolian settlement siedlungschema) were freestanding structures. 7. Conclusion As can be seen, the megaroid structures in the Phrygian highlands share with the historical megara, the similarities in their plan types, construction systems and materials, point to the existence of this type in the regional memory of rural architecture and to an architectural continuity in the region. How do we explain these similarities exhibited by the historical megaras and those existing in the houses used today by Turkmen groups in the area after all this time has passed? The nomadic Turkmen tribes entered Anatolia in the 11th century. They transitioned into their settled lifestyle only at the end of the 19th century after a long period of living as nomads (Ögel, 1991; Kavas, 2012). With the start of their settled inhabitation, the Turkmen tribes who had made use of yörük tents during their nomadic and semi-nomadic periods were inevitably influenced by in-

digenous Anatolian cultures (Tanyeli, 1996; Köse 2005). They chose the existing house plans of the indigenous people in the region when the time came for them to meet their need for permanent housing. This choice, which is the result of cultural adaptation, creates natural architectural continuity. A look into the plan types, materials and constructions systems Turkish populations preferred in their transition to a settled lifestyle reveal a close similarity with the traditional housing patterns of the regions that they settled in (Tanyeli, 1996; Köse, 2005). The Anatolian culture is not homogeneous. Each subregion contained various cultural traditions. It was because of this that displayed different regional  plan  types. At the same time, new environmental conditions such as climate, geography and topography, as well as cultural interaction were also influential in this choice. They were however adapted to particular needs. We found the megaroid buildings in the region reflecting the regional taste in that the local house plans in the region appear both socially and functionally suitable for the newly migrating nomadic Turks. Many historical house types are still being used in the rural architecture of Anatolia and the Near East can be seen in other plan types as well. Researchers Klinhott, Ragette, Yagi, Cerasi, Kobychev and Robakidze have reported that rural houses constructed in the style of historical plan types such as megaron, bit-hilani, tarma house, riwaq house, iwan house, houses with inner courtyards and houses with a front sofa, are still used in Syria, Iraq, Iran, Afghanistan, Daghistan and the Caucasion region, among other areas (Klinhott, 1978; Ragetta, 1974; Yagi, 1983; Cerasi, 2014; Kobychev, Robakidze, 1969). This is because traditional rural architecture is less prone to the impact of rapid cultural change and has evolved as a result of cultural continuity.  Anatolian traditional rural architecture also has a very rich cultural heritage related to the past. With this as its starting point, research conducted about Anatolia rural architecture constitutes an important resource that will shed light on studies into comparative

Continuity of architectural traditions in the megaroid buildings of rural Anatolia: The case of Highlands of Phrygia


evaluations of the housing architecture to be found in the layers of history hidden in any particular region. The concept of the Anatolian Turkish house has been in interaction for centuries with the Anatolian cultures that have been a part of this region, and has accordingly matured and reached a synthesis. Acknowledgements Firstly, I extend my deepest respects and gratitude to the distinguished scientist Prof. Dr. Jak Yakar, who has contributed extensively to Anatolian archeology and guided me in my work with his most valuable viewpoint and comments. I am also thankful for the kind contributions of Prof. Dr. Timothy P. Harrison in helping me to access some sources that I could not reach. I am strongly indebted to Dr. Caner Göçer, who accompanied me to the district. I am very grateful to Salih Ceylan, who took on the task of drawing up all plans. My deep gratitude goes to Uğur Süleymanoğlu, who drew the maps. I extend a special thank you to all the residents of the villages in Seyitgazi and Ihsaniye, who showered me with warm hospitality and assistance. References Akurgal, E. (2000). Anadolu Kültür Tarihi [The Cultural History of Anatolia]. Ankara: TÜBITAK Popüler Bilim Kitapları, 417. Altınsapan, E. (1999). Tarihsel Gelişim Sürecinde Eskişehir Bölgesindeki Ortaçağ Türk Yerleşimleri ve Yapısal Boyutu [Medieval Turk Settlements and Structural Dimensions in the Eskisehir Region during the Historical Development]. Anadolu Üniversitesi Edebiyat Fakültesi Dergisi [Anadolu University Faculty of the Letters Magazine], 1(1), 1-16. Ambrosini, L. (2009). The Rock-Cut Tombs of the Necropolis of Norchia (Viterbo-Italy): An Important Example of Ancient Architecture that must be Preserved. in Ferrari, A. (ed). Proceedings 4th International Congress on Science and Technology for the Safeguard of Cultural Heritage in the Mediterranean Basin, vol. II, Cairo, Egypt, 217-223. Ambrosini, L. (2012). The Rock-Cut Temple Tombs in the Mediterranean Area. A Study. in: Ferrari, A. (ed). Pro-

ceedings 4th International Congress on Science andTechnology for the Safeguard of Cultural Heritage in the Mediterranean Basin, vol. I, Cairo, Egypt, 229-236. Aşılıoğlu, F., Memlük, Y. (2010). Frig Vadisi Kültür Mirası Alanlarının Belirlenmesi ve Değerlendirilmesi [Determination and Evaluation of Cultural Heritage in the Phrygian Valley]. Ankara Üniversitesi Çevrebilimleri Dergisi [Ankara University Journal of Ecology], 2(2), 185-197. Banning, E.B., Byrd, B.F. (1988). Southern Levantine Pier Houses: Intersite Architectural Patterning during the Pre-Pottery Neolithic B. Paléorient, 14(1), 65-72. Barnett, R. D. (1967). Phrygia and the Peoples of Anatolia in the Iron Age, The Cambridge Ancient History, Volume II. Cambridge: Cambridge University Press. Berndt-Ersöz, S. (1998). Phrygian Rock-Cut Cult Façades: A Study of the Function of the So-Called Shaft Monuments. Anatolian Studies, 48, 87-112. Berndt-Ersöz, S. (2006). Phrygian Rock-Cut Shrines: Structure, Function, and Cult Practice. Leiden, Brill: Culture and History of the Ancient Near East, vol.25. Bintliff, J. (2012). The Complete Archaeology of Greece: From Hunter-Gatherers to the 20th Century A.D. UK: Wiley- Blackwell. Blegen, C. W. (1937). Excavations at Troy, 1936. American Journal of Archaeology, 41, 17–51. Blegen, C. W. (1953). The Palace of Nestor Excavations at Pylos, 1952. American Journal of Archaeology, 57(2), 59-64. Blegen, C.W., Rawson, M. (1966). The Palace of Nestor at Pylos in Western Messenia, I. The Buildings and their Contents. Princeton: Princeton University Press. Burdo, N., Videiko, M., Chapman, J., Gaydarska, B. (2013). Houses in the Archaeology of the Tripillia-Cucuteni Groups. in: Hofmann, D., Smyth, J. (ed), Tracking the Neolithic House in Europe. One World Archaeology, Springer Science+Business Media, New York, 94-115. Castel, C. (2010). The First Temples in antis. The Sanctuary of Tell Al-Rawda

ITU A|Z • Vol 12 No 3 • November 2015 • A. Erarslan


in the Context of 3rd Millennium Syria. in: Becker, J., Hempelmann, R., Rehm, E. (ed), Kulturlandschaft Syrien, Zentrum und Peripherie. Festschrift für Jan-Waalke Meyer, AOAT 371, Ugarit Verlag, Munster, 23-164. Cerasi, M. (2014). The Formation of Ottoman House Types: A Comparative Study in Interaction with Neighboring Cultures. Muqarnas Online, Volume 15, Issue 1: 116-156. David, N., Kramer, C. (2006). Ethnoarchaeology in Action. Cambridge: Cambridge University Press. Davey, C. J. (1980). Temples of the Levant and the Buildings of Solomon. Tyndale Bulletin, 31, 107-146. Deroy, L. (1948). Le Megaron homerique. Recherches d’etymologie grecque. Revue Belge de Philolgie Et D’Histoire, 26 (3), 525-537. Dörpfeld, W. (1902). Troja und Ilion. Ergebnisse der Ausgrabungen in den vorhistorischen und historischen Schichten von Ilion 1870–1894. Athens: Beck&Barth. Duru, R. (2003). Bademağacı Kazıları. 2000 ve 2001 Yılları Çalışma Raporu [Excavations of Bademağacı. 2000 and 2001 Work Report]. Belleten, LXVI, 549-594. Efe, T. (1996). 1995 Yılında Kütahya, Bilecik ve Eskişehir İllerinde Yapılan Yüzey Araştırmaları [Kütahya, Bilecik and Eskişehir Surveys in 1995]. Araştırma Sonuçları Toplantısı [Research Results Meetings]. II, 215-232. Efe, T., Ay-Efe, D. (2001). Küllüoba: İç Kuzeybatı Anadolu’da Bir İlk Tunç Çağı Kenti. 1996-2000 Yılları Arasında Yapılan Kazı Çalışmalarının Genel Değerlendirmesi [Küllüoba: An Early Bronze Age city in northwestern Anatolia. Overview of the Excavations carried out between the years 1996-2000]. TÜBA-AR, IV, 43-78. Efe, T., Sarı, D., Fidan, E. (2011). The Significance of the Keçiçayırı Excavations in the Prehistory of Inland Nortwest Anatolia. in: Bilgen, A. N., Von den Hoff, R., Sandalcı, S., Silek, S. (ed), Archaeological Research in Western Central Anatolia. The IIIrd International Symposium of Archaeology, Kütahya 8th-9th, March 2010, Dumlupınar University, Üçmart Press, 9-29. Erkanal, H. (1996). Early Bronze Age Urbanization in the Coastal Region

of Western Anatolia. in: Sey, Y. (ed), Housing and Settlement in Anatolia Historical Perspective, Türkiye Ekonomik ve Toplumsal Tarih Vakfı Yurt Yayınları, Istanbul, 70-82. Erkanal, H. (2011). Early Bronze Age settlement models and domestic architecture in the coastal region of western Anatolia. in: Şahoğlu, V., Sotirakopoulou, P. (ed), Across: The Cyclades and Western Anatolia during the 3rd Millennium BC, Sabancı University Sakıp Sabancı Museum May 24–August 28, Istanbul, 130–135. Eres, Z. (2013). Türkiye’de Geleneksel Kırsal Mimarinin Korunması: Tarihsel Süreç, Yasal Boyut [Consevation of the Traditional Rural Architecture in Turkey: Historical Process and Legal Aspect], in: Eyüpgiller, K.K., Eres, Z. (eds), Prof.Dr. Nur Akın’a Armağan – Mimari ve Kentsel Koruma [Festschrift to Prof.Dr. Nur Akın. Conservation of Architecture and Urban], YEM Yayınları, 457-469. Fidan, E. (2012). Küllüoba İlk Tunç Çağı Mimarisi [Küllüoba: the Bronz Age Architecture]. MASROP E-DERGİ, 7 [MASROP E-JOURNAL], 1-44. Garstang, J. (1953). Prehistoric Mersin. Yümük Tepe in Southern Turkey. Oxford: The Clarendon Press. Garstang, J., Garstang, J.B.E. (1940). The Story of Jericho. London: Hodder and Stoughton Ltd. Grave, P., Kealhofer, L., Marsh, B., Sivas, T., Sivas, H. (2012). Reconstructing Iron Age Community Dynamics in Eskişehir Province, Central Turkey. Journal of Archaeological Method and Theory, 19(3), 377-406. Harrison, T. P. (2012). West Syrian megaron or Neo-Assyrian Langraum? The Shifting Form and Function of the Tell Tayınat (Kunulua) Temples. in: Kamlah, J. (ed), Temple Building and Temple Cult Architecture and Cultic Paraphernalia of Temples in the Levant (2.– 1. Mill. B.C.E.), Harrassowitz Verlag, Wiesbaden, 3-23. Harrison, T.P., & Osborne, J.F. (2012). Building XVI and the Neo-Assyrian Sacred Precinct at Tell Tayinat. Journal of Cuneiform Studies, 64, 125143. Haspels, C. H. E. (1971), The Highlands of Phrygia. Sites and Monuments. New Jersey: Princeton University.

Continuity of architectural traditions in the megaroid buildings of rural Anatolia: The case of Highlands of Phrygia


Hiller, S. (1986). Early and Late Helladic ‘Magara’: Questions of Architectural Continuity in Bronze Age Greece. in: Hägg, R., Konsola, D. (ed), Early Helladic Architecture and Urbanization, Studies in Mediterranean Archaeology 76, Göteborg, 85-89. Hodder, I (1982). Symbols in Action. Ethnoarchaeological Studies of Material Culture. Cambridge: Cambridge University Press. Hodder, I. (1986). Reading the Past. Current Approches to Interpretation in Archaeology. Cambridge: Cambridge University Press. Hopkins, C. (1968). The Megaron of the Mycenaean Palace. Studi Micenei ed Egeo-Anatolici 6, 45-53. Hrouda, B. (1970). Die “Megaron”-Bauten in Vorderasien. Anatolia, 14, 1-14. Hundley, M. B. (2013), Gods in Dwellings: Temples and Divine Presence in the Ancient Near East. the USA: Society of Biblical Lit. Ilaslı, A. (2004). İlk Yerleşimden Türk Egemenliğine Kadar Afyon [Afyon: From the Prehistoric Settlements to the Turk Sovereigty], in: Uyan, M., Yüksel, İ., Avşar, N. (ed), Anadolu’nun Kilidi Afyon [Afyon: the Lock of Anatolia], T.C. Afyon Valiliği Yayınları, Afyon, 49-62. Işık, F. (1998). Megaron. Özel Bir Yapı Tipinin Kökeni, Gelişimi ve Bir Ege Tapınağı’na Dönüşümü [Megaron: Origin, Development and Transformation in an Aegean Temple of a Special Type], ADALYA, II, 1-49. Işık, F., Yılmaz, H. I. (1996). Structural Links between the Houses and the Tomb in Lycia. in: Sey, Y. (ed), Housing and Settlement in Anatolia Historical Perspective, Türkiye Ekonomik ve Toplumsal Tarih Vakfı Yurt Yayınları, Istanbul, 171-181. Ivanova, M. (2013). Domestic Architecture in the Early Bronze Age of Western Anatolia: the Row-Houses of Troy I, Anatolian Studies, 63, 17-33. Karazeybek, M. (2004). Selçuklu ve Beylikler Dönemi’nde AfyonKarahisar [AfyonKarahisar in the Age of Seljuks and Turks Principalities]. in: Uyan, M., Yüksel, İ., Avşar, N. (ed), Anadolu’nun Kilidi Afyon [Afyon: the Lock of Anatolia], T.C. Afyon Valiliği Yayınları, Afyon, 63-75.

Kavas, K. R. (2012). Türk Konut Mimarisinde Tarihsel Süreklilikler [Historical Continuities in Turkish Domestic Architecture: Central Asia and Anatolia]. Belleten, LXXVI/276, 503-538. Kjelden, K., Zahle, J. (1975). Lykische Graber. Ein vorlaufiger Bericht. Archaologischer Anzeiger, 1975, 312330. Klinhott, M. (1978) Megaron und Bit Hilani. Wohnungsbau im Altertum, Collectif, Deutsches archäologisches Institut, Berlin, 62-71. Knox, M.O. (1973). Megarons and MEGARA: Homer and Archaeology. Classical Quarterly, 23, 1-21. Kobychev, V. P., Robakidze, A. I. (1969). Basic Typology and Mapping of Dwellings of the Caucasian Peoples. Soviet Anthrology and Archeology, Volume 7, Issue 4: 13-28. Korfmann, M. (1983). Demircihöyük I: die Ergebnisse der Ausgrabungen 1975-78; Architektur, Stratigraphie und Befunde. Mainz am Rhein: P. von Zabern. Kortanoğlu, E. (2011). Rock-cut tombs which are located in the Highlands of Phrygia: An analysis within the frame of Hellenistic and Roman imperial periods in context of typology and façade carving. in: Bilgen, A. N., Von den Hoff, R., Sandalcı, S., Silek, S. (ed), Archaeological Research in Western Central Anatolia. The IIIrd International Symposium of Archaeology, Kütahya 8th-9th, March 2010, Dumlupınar University, Üçmart Press, 94-121. Köse, A. (2005). Türkiye’de Geleneksel Kırsal Konut Planlarında Göçebe Türk Kültürü İzleri [Traces of Nomadic Turkish Culture on the Plans of Traditional Rural Houses in Turkey] Afyon Kocatepe Üniversitesi Sosyal Bilimler Enstitüsü Dergisi, VII/2, 165-200. Laszlo, A. (2000). Some Data on House-Building Technıques and Foundatoin Rites in the Arıuşd-cucuteni Culture. Studia Antiqua et Archaeologica, VII, 245-253. Lloyd, S., Mellart, J. (1962). Beycesultan. Volume 1. Ankara: The British Institute of Archaeology at Ankara. Mellaart, J. (1959). Notes on the Architectural Remains of Troy I and II. Anatolian Studies, 9, 131-162. Mellart, J. (1970). Excavations at

ITU A|Z • Vol 12 No 3 • November 2015 • A. Erarslan


Hacılar I-II. Edinburg: The British Institute of Archaeology at Ankara. Mellink, M. J. (1986). The Early Bronze Age in West Anatolia: Aegean and Asiatic Correlations. in: Cadogan, G. (ed), The End of the Early Bronze Age in the Aegean, J. Brill, Leiden, 139152. Metzger, H., Coupel, P. (1963). Fouilles de Xanthos Tome II, L’acropole lycienne. Paris: Klincksieck. Müller, V. (1944). Development of the “Megaron” in Prehistoric Greece. American Journal of Archaeology, 48, 342-348. Naumann, R. (1998). Eski Anadolu Mimarlığı [The Ancient Anatolian Architecture], Ankara: Türk Tarih Kurumu. Neve, P. (1996). Housing in Hattusha, the Capital of the Hitite Kingdom. in: Sey, Y. (ed), Housing and Settlement in Anatolia Historical Perspective, Türkiye Ekonomik ve Toplumsal Tarih Vakfı Yurt Yayınları, Istanbul, 99-115. Omura, S. (1999). The Eleventh Excavation in Kaman-Kalehöyük 1996. in: Mikasa, H.I.H.P.T. (ed), Essays on Ancient Anatolia. Bulletin of the Middle Eastern Culture Center in Japan XI, Harrassowitz Verlag, Wiesbaden, 5193. Ögel, B. (1991). Türk Kültür Tarihine Giriş 7 [Introduction to Turk Cultural History], Ankara: Kültür Bakanlığı/638, Kültür Eserleri/46. Özdoğan, M. (2002). The Bronze Age in Thrace in Relation to the Emergence of Complex Societies in Anatolia and in the Aegean. in: Yalçın, Ü. (ed), Anatolian Metals II (Der Anschnitt 15), Bochum, 67-76. Özgüç, T. (1963). Early Anatolian Archaeology in the Light of Recent Research. Anatolia, VII, 1-21. Ragette, F. (1974) Architecture in Lebanon. Beirut: American University of Beirut. Roller, L. E. (2009). The Incised Drawings from Early Phrygian Gordion, Gordion Special Studies 4, Museum Monograph 130, Philadphia PA: University of Pennsylvania Museum of Archaeology and Anthropology. Rose, C. B. (2012). The Archaeology of Phrygian Gordion, Royal City of Midas. Philadelphia: University of Pennsylvania Press.

Schliemann, H. (1885). Tiryns. The Prehistoric Palace of the Kings of Tiryns. New York: C. Scribner’s sons. Schweitzer, B. (1951). Megaron und Hofhaus in Der Ägäis des 3–2. Jahrtausends V. Chr. The Annual of the British School at Athens, 46, 160-167. Sevin, V. (2007). Anadolu’nun Tarihi Coğrafyası I [The Historical Geography of Anatolia I], Ankara: Türk Tarih Kurumu. Smith, E. B. (1942). The Megaron and Its Roof. American Journal of Archaeology, 46(1), 99-118 Summers, G.D., Summers, F., Branting, S. (2004). Megarons and Associated Structures at Kerkenes Dag: An Interım Report. Anatolia Antiqua, XII, 7-41. Tanyeli, U. (1996). Anadolu’da Bizans, Osmanlı Öncesi ve Osmanlı Dönemi’nde Yerleşme ve Barınma Düzeni [The Settlement Pattern in Anatolia in the Byzantine, Pre-Ottoman and Ottoman Periods]. Tarihten Günümüze Anadolu’da Konut ve Yerleşim [Housing and Settlement in Anatolia from the Historical Periods to Today], T.C. Başbakanlık TOKI II. B.M. Insan Yerleşimleri-Habitat 1996 Konferansı Bildiriler ve Sergi Kitabı, 405-471. Vassileva, M. (2012). The Rock-cut Monuments of Phrygia, Paphlagonia and Thrace; A Comparative Overview. in: Tsetskhladze, G. R. (ed), The Black Sea, Paphlagonia, Pontus and Phrygia in Antiquity. Aspects of Archaeology and Ancient History, BAR International Series 2432, Archaeopress, 243-252. Yagi, K. (1983) Traditional Houses and Living Patterns in Syria, in: Germen, A. (ed), Islamic Architecture and Urbanism, King Faisal University, Dammam, 343-353. Yakar, J. (1985) The Later Prehistory of Anatolia. The Late Chalcolithic and Early Bronze Age. BAR International Series 268. Yakar, J. (2000). Ethnoarchaeology of Anatolia: Rural Socio-Economy in the Bronze and Iron Ages. Tel Aviv: Emery and Claire Yass Publications in Archaeology. Yakar, J. (2005). Ethnoarchaeology in Rural Anatolia. in: Takaoğlu, T. (Ed), Ethnoarchaeological Investigations in Rural Anatolia, Ege Yayınları, İstanbul,

Continuity of architectural traditions in the megaroid buildings of rural Anatolia: The case of Highlands of Phrygia


7-14. Young, R. S. (1960). Gordion: Phrygian Construction and Architecture: II. Expedition, 2, 1-12. Young, R. S. (1962). Gordion: Phrygian Construction and Architecture: II. Expedition, 4, 2-12. Warner, J. (1979). The Megaron and Apsidal House in Early Bronze Age Western Anatolia: New Evidence from Karataş. American Journal of Archaeology, 83(2): 133-147. Warner, J. (1994). Elmalı-Karataş II. The Early Bronze Age Village of Karataş. Bryn Mawr: Bryn Mawr College Archaeological Monographs.

Warner, K. (1993). The Megaron during the Aegean and Anatolian Bronze Age. A Study of Occurrence, Shape, Architectural Adaptation and Function, Studies in Mediterranean Archaeology CVIII. Astroms, Jonsered. Wright, J. C. (2006). The Formation of the Mycenaean Palace. in: DegeJalkotzy, S., Lemos, I. S. (ed), Ancient Greece: From the Mycenaean Palaces to the Age of Homer, Edinburg Leventis Studies 3, 7-52. Wright, G. R. H. (1985). Ancient Building in South Syria and Palestine. The Netherlands: Leiden, Brill.

Kırsal Anadolu’nun megarorid yapılarındaki mimari geleneğin sürekliliği: Dağlık Frigya örneği Megaron, ortasında ocak bulunan uzun dikdörtgen şeklindeki bir ana oda ile bu odanın yan duvarlarının (ante duvarları) uzaltılmasıyla yapının önünde oluşturulan üzeri örtülü bir portikodan/ante odasından (sundurma) oluşan yapıdır. Planın erken evrelerinde megaron, portikosuz/ante odasız tek bir oda iken ilerleyen evrelerde plana bazen ana odanın yan duvarlarının arkaya doğru uzatılmasıyla oluşturulmuş bir arka portiko/arka ante (sundurma) eklenir. Megaron planlı yapılarda örtü genelde beşik çatı iken düz çatı da kullanılmıştır. Dörpfeld, Tiryns ve Pylos daki Miken sarayları için megaron terimini kullanan ilk araştırmacıdır. Megaron türü yapıların kökenine dair farklı görüşler bulunmakla birlikte araştırmacılar en erken megaron türü yapılarla Neolitik Çağ’da Trakya, Makedonya ve Tesselya ovasındaki konutlarda rastlanıldığı konusunda hem fikirdirler. Megaron planın bu ilk örnekleri izole şekilde konumlandırılmış, uzun dikdörtgen şeklinde tek odalı, bazıları düz bazıları ise beşik çatılı yapılardır. Tunç Çağı, megaron planın en yaygın kullanıldığı dönemdir. Bu dönemde Poliochni, Thermi ve Miken saraylarında megaron planlı saray ve bey konutlarına rastlanılır. Megaron plan Anadolu’da da sevilerek kullanılmış bir plandır. Erken örneği Kalkolitik Çağ’da Hacılar II, Yümüktepe XVI ve Beycesultan XXIV da,

Tunç Çağları’nda ise Troia I-II kalelerinde, kıyı Ege’de Beşiktepe, Bakla Tepe ve Liman Tepe VI yerleşmelerindeki uuzn evler, Beycesultan’da ev ve kutsal yapılar, Antalya Bademağacı ve Karataş-Semayük, Eskişehir Küllüoba, Demircihöyük ve Keçiçayırı yerleşmeleri ile Marmara Bölgesi’ndeki Kanlegeçit megaron planın ünlü temsilcileridir. Frig mimarisi ise megaron planla özdeşleştirilmiş durumdadır. Megaron plan Anadolu’da orta Anadolu’ya kadar yayılmış olup bu bölgelerdeki Büyükkale, Kültepe ve Kerkenes dağı yerleşmeleri Geç Tunç ve Demir Çağları’nda megaron planlı yapılar içerir (Neve, 1996; Özgüç, 1963; Summers et al., 2004). Bu alan araştırması, Frigya vadisinin Dağlık Frigya olarak bilinen bölümünün günümüzde kapladığı alan olan Eskişehir Seyitgazi and Afyon İhsaniye ilçelerinin köylerini kapsayan alanda gerçekleştirilmiştir. 2014 yılı Ağustos ayında gerçekleştirilen bu çalışmanın amacı, Frigya vadisinin bu bölgesindeki köylerinde “megaron türü” yapı geleneğinin izlerini arayarak, bölgede kırsal mimaride var olduğu tarafımızca ileri sürülen “yapısal sürekliliğe” işaret etmektir. Bölgede bu amaçla gerçekleştirilen arazi çalışmasında megaron özelliği gösteren, bir ana oda ve önünde portikodan (sundurma) oluşan, düz ve beşik çatılı yapılara rastlanmış ve temkinli olmak amacıyla bu yapılar “megaronumsu”, “megaron benzeri” ve “megaron özellikli” yapılar olarak tanımlanmıştır. Bölgenin mimari kim-

ITU A|Z • Vol 12 No 3 • November 2015 • A. Erarslan


liğini oluşturmada önemli payı olduğu gözlenen bu yapıların bugün bir kısmı yıkıktır. Ayakta olanların bir kısmı ise ya terkedilmiş ya da depoya dönüştürülmüştür. Bölgede bugün in use durumunda hiçbir yapıya rastlanmamıştır. Yapıların kesin yaşları bilinmemekle birlikte kullanıcılarından alınan bilgilere göre ortalama yaş aralıkları 120-140 yıl arasındadır. Bölgede 2 anteli, in antis, tek anteli, prostyle, kapalı portikolu ve arka odalı, olmak üzere 6 megaron özellikli plan tipi tespit edilmiştir. Bu yapılar Anadolu ve Yakın Doğu’daki tarihsel megaronlarla karşılaştırıldığında plan, yapım tekniği, yapı

malzemesi ve çatı sistemi açısından bazı benzerlikler taşıdıkları görülmektedir. Bu durum bölgede megaron benzeri yapılar açısından kesintisiz bir mimari sürekliliğe işaret etmektedir. Tarihsel megaronlarla bölgedeki megaron tarzı yapılar arasındaki benzerlik açıklanmaya çalışıldığında ise bu durum, bölgeye 11. yüzyıldan itibaren gelmeye başlayan Türkmen grupların yerleşik düzene geçiş sürecinde, göçebe ve yarı göçebe dönemin yörük çadırını terkederek bölgedeki yerli halktan gördükleri bu planı kalıcı konut planı olarak tercih ettikleri şeklinde düşünülmektedir (Tanyeli, 1996; Ögel, 1991; Kavas, 2012).

Continuity of architectural traditions in the megaroid buildings of rural Anatolia: The case of Highlands of Phrygia

ITU A|Z • Vol 12 No 3 • November 2015 • 249-265

The reflection of religious diversity and socio-cultural meaning on the spatial configuration of Traditional Kayseri Houses Özlem ATAK1, Gülen ÇAĞDAŞ2 ozlematak@erciyes.edu.tr • Department of Architecture, Faculty of Architecture, Erciyes University, Kayseri, Turkey 2 cagdas@itu.edu.tr • Department of Architecture, Faculty of Architecture, Istanbul Technical University, Istanbul, Turkey


Received: December 2014

Final Acceptance: August 2015

Abstract Recent studies on domestic spaces have demonstrated that social meaning and cultural values are mostly reflected by means of spatial organizations of houses, thus, in that way, different cultures express themselves through different spatial models. In this respect, space syntax and visibility graph analyses arise as the computational approaches to discover the interactions between space and culture. These methods have essentially been constructed through the relationships of permeability based on movement and visibility based on the perception of a moving observer. In the examination of the entire house or it’s certain spaces in the context of introversion and extroversion, they are effective methods which are used to understand the privacy related to spaces, control mechanisms, the level of the internal relations of the household and the relations between the household and the visitors. In this study, domestic space was examined through twenty-seven traditional houses of Kayseri in Central Anatolia, Turkey, where people from different religious beliefs have lived for long years, within the contexts of space syntax and visibility graph analyses. The houses were analyzed by using the Depthmap-UCL software developed by Alasdair Turner. After the permeability and visibility analyses, the study focuses on how socio-cultural meanings are reflected on the spatial configuration of traditional Kayseri houses, what common and/or different characteristics are demonstrated by the houses in terms of the spatial configuration and visibility structure, and the importance of permeability and visibility structures in the spatial configuration of houses. Keywords Traditional Kayseri Houses, Space syntax, Visibility.


1. Introduction One of the important research fields in the relationship between spatial organization and social structure is the domestic space. Studies conducted on the domestic space demonstrate that social meaning and cultural values are mostly reflected by means of spatial organizations of houses, thus, in that way, different cultures express themselves through different spatial models. In this respect, the space syntax claims its place in researches as a computational architectural theory and a morphological analysis method that examines the interactions between the society and all kinds of spatial configurations. The space syntax, proposed by Hillier and Hanson in 1984, implies that the socio-cultural structure and processes exhibit themselves in the space with spatial configuration. According to the theory, the social structure and the space mutually interact with each other. The most essential strategy is to attempt to discover several stable aspects in the spatial configuration and to transform them into cultural human-interaction patterns. This theory, through the departure point that there exists a direct relationship between the spatial organization and the social structure, attempts to explore the ways how people perceive and use the space, depending on permeability in spatial organization and visual fields. A number of computational analysis technique and tool have been developed for the configurational analysis of the space, and they analyze spatial organizations by configurationally defining the entire structure. In addition, Turner proposes the visibility graph analysis method that is based on Benedikt’s isovist concept and the space syntax, and develops the Depthmap-UCL software that is able to carry out this and other spatial analyses within the context of space syntax. İsovist concept initially introduced by Tandy (1967), and was formalized by Benedikt (1979). İsovist is the set of all points visible from a point in the space. The shape and size of the isovist differ according to the observer’s point of view and stance (Benedikt, 1979). According to this method, visibility structure of spaces based on the

perception of a moving observer, along with the spatial characteristics of spaces based on permeability relations, play important roles in the presentation of spatial configuration. The information provided by visual field in the urban environment and in buildings might help the user find his/her way. In addition, it is possible through visual fields to control the information provided to the user within the system. Therefore, permeability and visibility relations reveal the spatial organization of all spatial systems including houses, and the ways the household and visitors perceive the house. In addition, the level of the internal relations of the household and the relations between the household and the visitors are arranged through permeability and visibility structures. The level of privacy (interpersonal interaction) within the house can be determined by defining physical or invisible boundaries. While physical boundaries are the ones that control the visibility, that is to say, the movement, invisible boundaries are the control of the visual knowledge provided through physical boundaries such as the prevention of eye contact. Moreover, the level of privacy differs between different societies and cultures. This difference is shaped according primarily to the family’s social structure and its relations with visitors. Therefore, analyzing the permeability and visibility structures in houses will help us to understand the level of privacy in that culture and, thus, the interactions within the family and between the family and visitors, and the statuses of the functions that belong to the domestic space within the permeability and visibility structures. Space syntax and visibility methods, which enable the determination of permeability and visibility structures, lead us to results concerning the relationship between the space and the socio-cultural structure by considering the space and its configuration through the user’s movement within the space and his/her visual perception. In this study, these two methods will be used in order to demonstrate how social and cultural meanings are reflected on the spatial configuration of traditional Kayseri houses in Central Ana-

ITU A|Z • Vol 12 No 3 • November 2015 • Ö. Atak, G. Çağdaş


tolia where people from different religion and cultures coexisted for long years. 2. Traditional Kayseri Houses Kayseri one of the few Anatolian cities in which a substantial Christian minority lived. Towards the end of the sixteen-century there were fifty Muslim, thirteen Christian and nine-mixed town-quarters (Jennings, 1976). In the seventeen-century Muslim quarters declined to thirty-five, the number of Christian quarters was fourteen, while mixed population town-quarters was twelve (Faroqhi,1987). The proportion of Christian population remained more or less the same until the establishment of the Republic. Two groups, Gregorian Armenians and Orthodox Greeks, lived with the Muslims in a friendly and cooperative way (İmamoğlu, 2006). ‘The traditional Kayseri house until the twentieth century was a living entity, a process, a natural phenomenon, not a finished product on going organic process within a family lot over several generations. Continuous additions or alterations of room or service spaces on the ground or upper levels were considered natural. As a result, overlapping and intermingled volumes, and superimposed walls and planes were common. Houses were divided among brothers and sisters after their parents passed away, independent units being added at the expense of smaller gardens or courtyards. Nothing stopped this organic process in any period. This flexible and dynamic attitude towards buildings has undergone some changes in the 20th century with practices borrowed from Europe, generally by the Christian community, who built

complete and finished rectangular houses. Even these houses, too, were altered either by adding rooms or service spaces, or changing their functions. In short, spaces grew and spaces died just like their owners, but the family lots were continuously in use. Evolution or change within continuity, perhaps reflecting the essence of life, was implicit principles.’ (İmamoğlu, 2006). In general, a traditional Kayseri house is the outcome of a natural and unpretentious building process, inward looking and asymmetrically growing a courtyard or a garden. Houses were generally designed and built by masons, according to custom, as well as, requirements and desires of the owner. Many daily activities of the house are carried out in the courtyard except cold days, rooms are oriented towards the courtyard and they have windows looking at the courtyard. Each room generally carries out more than one function; however, the main function of a room is clear. In rooms, especially in halls (sofa), there exist room-entrances called “seki altı”. “Seki altı” is lower than the main floor (seki). Hall has a different meaning in these houses. Its function to direct the house arrangement and as a space of distribution stays in the background. It has many functions such as being the entrance of the house, a reception place, a guest room, a living room, a prayer place and a bedroom for elders. “Tokana” is a space used as kitchen, winter room and storeroom. A cooker is located in the seki altı part of the tokana. “Harem room” is a private place where strangers are not desired to enter. Family members -especially women and children- spend most of the house time

Figure 1. Examples from Traditional Kayseri Houses and Streets (İmamoğlu, 2006). The reflection of religious diversity and socio-cultural meaning on the spatial configuration of Traditional Kayseri Houses


Figure 2. Examples from Traditional Kayseri Houses, ground floor plan and east elevation of House Öztaşcı House, 19th Century, Muslim house example (İmamoğlu, 2006).

Figure 3. Examples from Traditional Kayseri Houses, ground floor plan and section of İmamoğlu House, late 19th Century, Christian house example (İmamoğlu, 2006).

in the harem room. Kiosks (köşk) are designed as semi-closed sitting units around the garden or courtyard (İmamoğlu, 2006). Traditional houses are generally one or two storied. Generally, upper storey started to be seen in examples from the 19th Century. It is stated in the literature that all units of a house have been rendered private by closing the house to other houses and to the street by surrounding it with high courtyard or building walls. It is of high importance especially for Muslim families to protect privacy within the house and the courtyard. It is apparent that security concerns too, along with privacy, play important roles in this attitude. Another reason to shelter houses is the fact that people leave the city in the summer and go to vineyards. Security concerns are in the foreground in houses built before the last century, however,

such measures started to be loosened. Traditional houses generally have few and small windows. Basically, stone, wood and iron have been used in the construction of houses (İmamoğlu, 2006). It is observed that low- and middle-income families in Kayseri live in similar small houses, regardless of their religion. Among high-income families, on the other hand, houses of Muslim families have a simple plan. One of the important features desired in Muslim houses is the privacy provided to women. High walls surrounding the courtyard, low number of windows on the ground floor and the distant position of the living room from the street can be considered to be a set of design rules in order not only to ensure security but also to distinct women from men. In the 19th Century, the Christian minority started to lead a bourgeois lifestyle. In

ITU A|Z • Vol 12 No 3 • November 2015 • Ö. Atak, G. Çağdaş


this period, Christian houses are twotwo and a half storied, and have very small gardens or green areas. According to the literature, the plan of these houses started to transform from being inward-looking to outward-looking. In early this century, while well to do Muslim houses continued to have incomplete organic constructs, Christian house plans started to become more organized, symmetric, inclusive of all functions and complete rectangles. Rooms are situated around a courtyard or a central hall. All rooms have been constructed in similar manners without considering or highlighting an order of importance. While Muslim houses have two different entrances and courtyards as harem room and selamlık (private and public), Christian houses do not have such a distinction. However, after the establishment of the Republic, Muslim’s understanding of privacy started to change gradually, and modern values both in privacy and interactions with the opposite sex have been mostly embraced (İmamoğlu, 2006). İmamoğlu (2006) explains the difficulty to classify Traditional Kayseri Houses in terms of religion as follows; ‘First of all, since Muslims were in a majority, remaining houses are mostly Muslim houses and the number of Christian examples are limited. This limitedness is more explicit for the period before the 1835 earthquake, because the small number of houses that remain from that time belongs only to Muslims. Another reason is the distribution of population according to income. The majority of Kayseri natives, regardless of their religion, were of low and middle income type, living in modest houses, most probably of similar, if not identical character and layout. However, careful researcher with additional information from the community may be able to distinguish between the houses of well-todo people of different religions. 3. Spatial and visual analyses of Traditional Kayseri Houses Most of traditional Kayseri houses have been destroyed or ruined although they were certified and included in urban protected areas. In this study, a total of twenty-seven houses

- seven houses (House 3, 5, 6, 7, 14, 17, 18) included in Vacit İmamoğlu’s book (2006) and twenty houses included in Gonca Gündoğdu’s master thesis (1986)- were considered. Most of these houses belong to the 19th and early 20th centuries, and only a house from 18th century. Except some of them, religious beliefs of the constructors or owners of these houses remain unclear. Classification of these houses in terms of religion based on information from İmamoğlu’s book (2006) and his book review (2006) on Büyükmıhçı’s book (2005) (Table 1). Most of the examined houses are courtyard-type houses frequently observed among traditional Kayseri houses. However, three of these houses do not have courtyards (House 1, 5, 9). Three of them exhibit a similar spatial configuration: single-storied, symmetrically designed and having a central hall surrounded by rooms. Also these houses are examples of Christian houses. In houses with courtyards, on the other hand, the shape of the courtyard, its size and place within the house differ. Some of the houses with courtyards are single-storied, some have a semi second storey and some are two-storied completely (Table 1). The analyses were carried out on the models of the houses abstracted through the principles of the Depthmap UCL software developed by Turner. In addition, by researching several spatial features of traditional houses, their users, and the era’s cultural, social and economic characteristics; the interaction between space configurations and the socio-cultural structure was explored by establishing relations of causality through these data. In order to display and interpret the depth of social and cultural knowledge on the domestic space, two types of analyses were carried out for all houses in the study. The first type of analysis is convex space analysis that is based on accessibility relations depending essentially on human movement. In this analysis, all spaces are represented as convex spaces in order to see how various functions relate to each other and their positions within the whole. Each house is reduced to fewest and largest convex spaces. Then, convex spaces are related

The reflection of religious diversity and socio-cultural meaning on the spatial configuration of Traditional Kayseri Houses


Table 1. Traditional Kayseri houses which examined in this study – house names, century of built, religion of owners- R (Muslim – M, Christian-C) and plans of houses.

House 16

Haci İbrahim G. House

House 17

Bezircioğlu House

House 18

Camcıoğlu House



House 19

Şükrü Karaca House




House 20

Yapıkçılar House




House 21

D. İzzet House


House 22

H. Ali Yapaner House

House 23

A. Bakkaloğlu House

House 24

Efendi Ağalar House Gazioğlu House

Nuri Sezer House Çalıka House

House 26

Mustafa Hızırel House

House 27

House 11

A. Pastırmacı_ oğlu H.


House 12


House 13

19th /second h.

House 14


to each other depending on the permeability principles among them. After all these stages, the software creates various measurements and maps of these measurements. Obtained measurements are the measurements of the




early 20th

early 20th

early 20th






19th /second h.

early 20th

upper floor


19th /second h.

ground floor

late 19th




19th /second h.

Late 19th


early 20th


19th 20th

House 25

Gavremoğl u House Baldöktü House

19th /second h.

House 6

18th /second h.

Hüseyin Kış House

Late 19th

plans century

Hacı Türkaslan House


House name House 15


19th /second h.

Upper floor

Hoca Haser House


late 19

ground floor

Selçukoğlu Osman

Ahmet Karaca House Körükçüoğlu House



Müftü House

Öztaşcı House


Hacı Ahmet Ağa H.

İmamoğlu House

House 1


House 7

House 3

House 8

House 4


House 9

Muhittin Gürbaz House

plans century

House 10

House 5

House 2

House name


system such as connection, integration, mean depth and controllability. This analysis was carried out in two different types in this study: internal relations of the house were focused in the first analysis without including the

ITU A|Z • Vol 12 No 3 • November 2015 • Ö. Atak, G. Çağdaş


outer space and the relationship of the house with the outer space and changes taking place in its internal structure were focused in the second one by including the outer space. In order to be able to read these results and changes clearly, integration maps of all houses with and without the outer space were obtained and compared. In this map, spaces are colored up from red to dark blue according to their integration levels. Red color represents the most integrated space, while dark blue represents the most nonintegrated space. The second type of analysis is visibility graph analysis. Through visibility graph analysis, houses were compared according to the spaces they include and visual areas they have, to the visual integration scores obtained from visual area maps, and to the integration scores obtained from convex space maps. In addition, their visual perception and movement relations were investigated. This analysis is based on the visual perception of the moving observer. With this analysis conducted from the eye level, measurements and maps of a space such as visual integration, connection, visual mean depth and visual controllability were obtained. In this type of analysis, as is the case in the convex space analysis, a color interval ranging from red to dark blue is used. While red represents the most visually integrated areas, dark blue represents the most visually non-integrated areas. Measurements and maps obtained from these two types of analysis based on human movement and the perception of the moving observer were firstly evaluated individually, and then by making comparisons within each type of analysis and between each other. In this way, based on the housesâ&#x20AC;&#x2122; permeability-visibility structures and relationships with each other; it was attempted to understand the interaction between the spatial configuration of Traditional Kayseri Houses and the socio-cultural structure. 3.1. Spatial analysis findings Hillier and Hanson conceptually define buildings as the regulation of different human categories through a control mechanism. In domestic space, these categories are defined as

the household living in the house and visitors. The domestic space, according to this definition, regulates both the internal relations of the household and the relations between the household and visitors through the control and permeability relations between the internal and external parts of the house (Hillier and Hanson, 1984). In this way, the domestic space is essentially related to the regulation of two types of spatial relations in terms of the domestic space accessibility relations. These are the internal relations of the house and the relations between the internal and external parts of the house. In this respect, in order to understand the internal relations of the house and the relations between the internal and external parts of the house, convex space analyses were first carried out without including the outer space and then repeated by including the outer space. Table 2 demonstrates the ranking of the average integration scores obtained without including the outer space, from the most integrated house to the non-integrated house. In the convex space analysis carried out without including the outer space, the House 1 is the most integrated house with an integration score of 1,192. It is followed by the House 2 and House 3 with integration scores of 1,152 and 1,040, respectively. House 1 exhibits an integrated structure with its integration score within the integration-separation distinction that is based on permeability relations. House 2 and House 3 are situated on the frontier within this distinction. Other houses have non-integrated structures. Therefore, it is possible to conclude that all houses except the first three of them have more tendencies to non-integrated in terms of the composition of their spaces. The first three houses that have integrated structures have completely different spatial compositions. House 1 is one of three houses that have a central hall. House 2 is a double-storied house with a central courtyard, and House 3 is a single-storied house with an organic structure and a fragmented courtyard. Among the other houses, there are examples similar to these three houses. For example, House 5 and House 9 exhibit a spatial composition similar to

The reflection of religious diversity and socio-cultural meaning on the spatial configuration of Traditional Kayseri Houses


Table 2. Convex space analysis results (House names, century, religion of house owners(R), integration without the outer space, integration including the outer space, controllability-C,) (Atak, 2009). Integration-outer space


House Nr.




Integration (HH) Averag Min.





1 2 3 4 5

House 1 House 2 House 3 House 4 House 5

Ahmet Karaca H. Körükçüoğlu House Öztaşcı House Muhittin Gürbaz

….. late 19th 19th /second h.

İmamoğlu House

C … M … C

1,192 1,152 1,040 0,963 0,931

0,616 0,653 0,592 0,469 0,419

2,157 2,611 2,045 1,875 1,679

1,183 1,056 1,021 0,969 0,907

0,619 0,562 0,597 0,475 0,420

2,177 2,312 2,015 1,917 1,610

0,118 0,135 0,152 0,133 0,117

6 7 8 9

House 6 House 7 House 8 House 9

Gavremoğlu House Baldöktü House Hüseyin Kış House Müftü House

M M … C

0,922 0,914 0,907 0,889

0,480 0,551 0,597 0,468

1,811 1,653 2,091 1,579

0,940 1,088 0,976 0,928

0,488 0,650 0,631 0,484

1,883 1,734 2,079 1,686

0,117 0,176 0,114 0,129


House 10


House 11

Hacı Ahmet Ağa H. Mustafa Hızırel H.

12 13 14 15 16

House 12 House 13 House 14 House 15 House 16

17 18 19 20 21

…. Late 19th 18th /second h. 19th /second h. …. late 19th early 20th










19th /second h.









A. Pastırmacıoğlu H Nuri Sezer House Çalıka House Hoca Haser House Haci İbrahim G. H.

early 20th ….. early 20th ….. 19th /second h.

… … M … …

0,853 0,843 0,824 0,804 0,802

0,408 0,514 0,559 0,532 0,584

1,470 1,148 1,210 1,335 1,286

0,849 0,856 0,808 0,807 0,812

0,408 0,520 0,521 0,529 0,576

1,485 1,210 1,214 1,332 1,343

0,156 0,135 0,106 0,126 0,160

House 17 House 18 House 19 House 20 House 21

Bezircioğlu House Camcıoğlu House Şükrü Karaca House Yapıkçılar House D. İzzet House

M C … … C

0,771 0,768 0,757 0,702 0,692

0,413 0,534 0,489 0,396 0,341

1,347 1,278 1,152 1,166 1,125

0,790 0,781 0,785 0,721 0,731

0,425 0,538 0,498 0,401 0,350

1,367 1,299 1,219 1,223 1,175

0,119 0,140 0,135 0,158 0,144

22 23 24 25 26

House 22 House 23 House 24 House 25 House 26

H. Ali Yapaner H. A. Bakkaloğlu H. Efendi Ağalar H. Hacı Türkaslan H. Gazioğlu House

19th 20th Late 19th ….. …... …... 19th /second h. …. ….. early 20th 19th /second h.

… … C … C

0,691 0,671 0,630 0,619 0,612

0,391 0,381 0,387 0,404 0,360

1,042 1,125 1,007 0,827 1,033

0,695 0,679 0,653 0,640 0,622

0,397 0,400 0,389 0,402 0,363

1,040 1,125 1,083 0,948 1,060

0,164 0,145 0,137 0,140 0,142


House 27

Selçukoğlu Osman

early 20th








that of House 1. Although these houses exhibit a similar spatial composition, they have non-integrated structures. Therefore, it is apparent that it is not possible to draw general conclusion on whether these houses have integrated or non-integrated structures departing from differences such as having courtyards or central halls, having complete or incomplete geometrical forms. As Table 2 demonstrates, in the convex space analysis, integration scores of houses differ when the street -that is considered to be the outer space- is included. Integrated ones among these houses (House 1, House 2 and House 3) show a tendency to non-integrated when the outer space is included in the analysis. Therefore, it can be argued that these houses have inward-looking structures. While the integration scores of seven houses decreased after the inclusion of the outer space to the analysis, those of twenty houses increased. Hence, less than one-third of the examined houses have inward-looking structures, and more than two-third of

them have outward-looking structures. Two of three houses that have halls exhibit a more inward-looking structure; the third one has an outward-looking structure and most of the houses with courtyard exhibit outward-looking structures. If a place has a large visual area that is composed of numerous points, the environment can be controlled easily. This measurement in convex space analysis is based on permeability relations. The minimum controllability among the examined examples belongs to House 14 with a controllability score of 0,106 in Table 2, and the maximum controllability score belongs to House 7 with a score of 0,176. It means that House 14 has a weaker composition while House 7 exhibits a stronger structure in terms of controllability. However, it should be noted that there is not a big difference between these minimum and maximum scores. For House 14, it can be stated that the upper storey of this house is used as extensively as the ground floor and, therefore, the con-

ITU A|Z • Vol 12 No 3 • November 2015 • Ö. Atak, G. Çağdaş


Table 3. Convex space integration maps and visibility integration maps (Atak, 2009).

trollability between storeys becomes difficult to achieve. Although House 7 is a house non-integrated as harem room-selamlÄąk and the selamlÄąk part has low integration and -thus- controllability scores, there exist alternative passageways to the other spaces of the house. It can be argued that this situation caused the house to get a high controllability score in terms of acces-

sibility. What is important here is the role of multiple alternative transitions in the controllability measurement that is based on accessibility. In this respect, high controllability scores of the courtyard and central hall are associated with the existence of transitions from many other spaces to these spaces. Table 3 demonstrates the housesâ&#x20AC;&#x2122; convex maps from the most integrat-

The reflection of religious diversity and socio-cultural meaning on the spatial configuration of Traditional Kayseri Houses


ed to the non-integrated and visibility integration maps. The most integrated space is the central hall in houses with central halls (House 1, 5, 9), and the whole or some parts of the courtyard in almost all houses with courtyards. The general conclusion obtained from the spatial analysis based on accessibility relations is that the courtyard plays a key role that holds together all units of the house and connects them to the outer space. In other words, the house that is intensively used is of a character that controls the movement and activities or various spaces in it, and structures these relations. It is observed that this function is carried out by the central hall in houses that have central halls. As indicated before, this measurement in convex space analysis is based on permeability relations. However, visible areas play significant roles for the controllability of a space. When inner spaces have a composition that does not allow them to see each other, it is possible for the controllability score obtained in convex space analysis to yield similar results with the score obtained in visibility measurement. However, it is not clear whether the same results will be obtained or not with the controllability score obtained in convex space analysis in the existence of permeable units such as windows and glass walls as well as doors that allow spaces to see each other. In the examined houses, there exist windows in inner spaces opening to the courtyard or to each other. In this respect, it is expected to obtain significant results by comparing the integration and controllability scores of houses obtained through convex space and visibility graph analyses. 3.2. Visibility graph analysis findings In revealing the social logic lying behind houses, visibility relations play roles as significant as those of accessibility relations. The visibility graph analysis carried out by using the Depthmap UCL software provides several special measurements representing local and global visual characteristics of houses. The measurement scores of traditional Kayseri houses obtained through visibility graph analysis are

presented in Table 4. The table demonstrates the ranking of the houses, from the most visually integrated house to the non-integrated one. It is evident that the visual integration ranking is different from the integration ranking obtained through permeability analysis. The most visually integrated houses are those that have a big courtyard and/or a garden, which can easily be seen from other spaces. Another factor is the existence of high numbers of windows opening from spaces to each other and especially to the courtyard. In addition, the upper storey, if any, of the house has a visual relation with the yard as well. Houses having central halls (House 1, House 5, and House 9) occupy lower ranks in terms of visual integration. These three houses are Christian house samples. This situation stems from the lack of numerous windows in inner spaces of houses that make visual connection, as it is the case in houses with courtyards. Most of the houses have outward-looking structures in terms of permeability relations and inward-looking structures visually. Facades of houses, especially of Muslim houses, do not have many windows and courtyard walls are very high. However, these houses have sides looking at the courtyard and these sides have many windows. Therefore, spaces within the house for the common use such as courtyards have significance in terms of visibility. Therefore, houses’ visual integration scores might yield more sensitive and different results than the integration scores based on permeability. Moreover, while spaces other than doors such as windows are not important in permeability relations, these spaces and their sizes are important in visibility analysis. The visual controllability measurement differentiates the spaces that can be visually controlled easily. If a place has a large visual area that is composed of numerous points, it is possible to define the environment as controllable. The controllable amount of the space decreases towards points with fewer spaces and especially towards doorsides in corridors. On the other hand, controllable spaces are spaces that cannot see other spaces much, but can eas-

ITU A|Z • Vol 12 No 3 • November 2015 • Ö. Atak, G. Çağdaş


Table 4. Visibility analysis results (House Names, Century, Religion of house owners –R,Visual Integration, Visual Controllability-VC, Visual Coefficient Clustering -VCC) (Atak, 2009). Nr 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27

House Nr. House 13 House 17 House 20 House 15 House 2 House 22 House 3 House 12 House 23 House 19 House 21 House 8 House 10 House 4 House 24 House 16 House 27 House 9 House 25 House 7 House 14 House 18 House 6 House 1 House 11 House 5 House 26




Nuri Sezer House Bezircioğlu House Yapıkçılar House Hoca Haser House Körükçüoğlu House H. Ali Yapaner H. Öztaşcı House A.Pastırmacıoğlu H. A. Bakkaloğlu Şükrü Karaca House D. İzzet Efendi Hüseyin Kış House Hacı Ahmet Ağa H. Muhittin Gürbaz H. Efendi Ağalar Hacı İbrahim G. H. Selçukoğlu Osm. H. Müftü House Hacı Türkaslan H. Baldöktü House Çalıka House Camcıoğlu House Gavremoğlu House Ahmet Karaca Mustafa Hızırel İmamoğlu House Gazioğlu House

….. 19th 20th …... ….. late 19th 19th /second 19th /second early 20th …. ….. …... …. early 20th …. …. 19th /second early 20th late 19th early 20th 19th /second early 20th Late 19th 18th /second … 19th /second Late 19th 19th /second

… M … … … … M … … … C … M … C … … C … M M C M C C C C

Visual Integration (int.HH) Average Min. Max. 34,290 5,625 50,958 24,867 6,711 43,980 20,647 7,195 30,195 20,546 4,815 31,670 17,125 5,763 27,967 17,031 5,305 27,685 16,152 5,828 27,009 15,113 4,862 24,490 14,820 5,315 27,215 14,600 5,600 23,410 13,760 5,417 30,416 13,676 5,031 20,844 13,488 4,733 24,540 13,342 4,866 24,119 13,132 6,456 21,155 13,076 5,458 21,866 12,477 5,465 25,703 11,399 5,066 18,488 11,210 3,834 19,153 10,948 4,849 17,188 10,835 3,637 20,090 10,684 6,067 17,386 10,634 4,677 16,742 10,466 1,886 18,108 9,681 4,050 16,190 9,162 3,185 16,796 8,728 3,130 14,459

ily be seen from other spaces. House 13 has the highest controllability in terms of the visual integration score. The lowest score belongs to House 1 with a score of 0,227, which has a central hall and which is ranked first in the integration ranking based on permeability relations. It is seen that other houses that have central halls have low controllability scores (House 5, 9). In short, it is possible to conclude that visually integrated houses and spaces have high visual controllability scores. Coefficient clustering measurement, on the other hand, shows how long the visual area of an observer will remain the same and to what degree the spatial perception of the observer will change at the end of his/her movement from a point to a more distant point. If there will occur a big loss of visual knowledge after the observer becomes distant, the coefficient clustering score becomes low. The highest score (0,855) belongs to House 13, which has the highest visual integration and controllability scores, while the lowest score (0,599) belongs to House 18. The visual integration maps in Table



0,711 0,599 0,563 0,560 0,370 0,521 0,423 0,458 0,491 0,430 0,468 0,435 0,345 0,372 0,440 0,403 0,407 0,249 0,449 0,259 0,346 0,365 0,341 0,227 0,333 0,318 0,495

0,855 0,765 0,767 0,798 0,645 0,750 0,764 0,685 0,756 0,762 0,716 0,714 0,696 0,687 0,648 0,741 0,725 0,602 0,715 0,680 0,665 0,599 0,711 0,663 0,691 0,752 0,772

3, while the most visually integrated points are represented in red on the map, the most non-integrated points are represented in dark blue. It is observed that courtyards, gardens and central halls are mostly more integrated spaces. These spaces are followed by spaces for common use such as halls and kiosks. Spaces such as toilets, storerooms, upper-storey rooms and kitchen (tokana) are more visually non-integrated spaces. Another important situation is that the selamlık part of houses that are non-integrated as harem room and selamlık is rendered visually different from the other units of the house. This situation is also seen in convex space analyses based on permeability relations. In addition, it is observed that more visually integrated spaces have higher visual controllability scores. In addition, isovists were produced from points determined in several spaces within houses. Isovist defines the visual knowledge that an observer obtains by rotating 360 degrees on a vantage point. Firstly, isovists were produced from entrances of all hous-

The reflection of religious diversity and socio-cultural meaning on the spatial configuration of Traditional Kayseri Houses


Table 5. İsovists, (a) from entrance of houses and (b) the center of these spaces (Atak, 2009).

(a) es in order to determine what kind of a visual knowledge is obtained and what spaces are seen as one enters the house. Then, they were produced from the central points of these spaces in order to determine the visual knowledge that is obtained as one arrives at the centers of entrances (Table 5). Isovists obtained from house entrances demonstrated that the first person who enters the house perceives the whole or a large part of the entrance space seen as a courtyard or a central hall. In addition, other spaces can also be seen, although to limited degrees, through openings on their surfaces looking at the entrance. However, this visual knowledge provided through door and window openings in spaces such as rooms is a controllable knowledge. In houses having harem room-selamlık distinction, isovists obtained from the entrances of these parts show that they have a very limited visual relationship with each other. This might be an example of a deliberate visual distinction. Isovists obtained from the centers of

(b) entrances indicate that the whole entrance space and large parts of other spaces are perceived from this space, although to limited degrees. While spaces open to public use such as halls and kiosks are spaces that can easily be seen from this space, spaces such as rooms and kitchen allow a limited and controllable vision through their openings. 4. Discussion and conclusions Space syntax and visibility graph analyses are the methods effectively used in the examination of a house or some parts of a house in terms of inwardness-outwardness, and the determination of spatial privacy, control, social hierarchy within the household and the degree of relations between the household and visitors. These methods have essentially been constructed through the relations of permeability based on movement and visibility based on the perception of a moving observer. Permeability and visibility relations

ITU A|Z • Vol 12 No 3 • November 2015 • Ö. Atak, G. Çağdaş


reveal the spatial mechanisms of houses and the ways the household and visitors experience these systems. In addition, the level of the internal relations of the household and the relations between the household and the visitors are arranged through these relations. The level of privacy within a house can be determined by defining physical or invisible boundaries. While physical boundaries are the ones that control the visibility, that is, the movement; invisible boundaries are the control of the visual knowledge provided through physical boundaries such as the prevention of eye contact. However, the level of privacy ensured through the control of these knowledge areas differs between different societies and cultures. In this respect, in the fieldwork, in order to reveal the spatial configuration of traditional Kayseri houses and the social and cultural knowledge lying behind this configuration, both of these analyses were carried out together; and the houses’ spatial and visibility structures’ characteristics that support each other and their differences were explored. In the sample, the dominant presence of the courtyard-integrated spatial theme is clearly apparent. In two groups from different religious, common house type is courtyard-house type. Courtyard still remains the importance in these centuries for two groups. However in early 20th century, while well to do Muslim houses continued to have incomplete organic constructs, Christian house plans started to become more organized, symmetric, inclusive of all functions and complete rectangles. Rooms are situated around a courtyard or a central hall. The general conclusion obtained from the spatial analysis based on accessibility relations is that the courtyard plays a key role by not only connecting all units of the house to the outer space by holding them together, but also structuring the main spaces of the house. Central halls play this role in a small number of examples having central halls. However, the central hall does not have a function similar to that of the courtyard. Almost all of the examined houses show a tendency of separation in terms of the ways their

spaces come together. In addition, while two-third of the houses exhibit outward-looking structures, only onethird of them exhibit inward-looking structures. This finding demonstrates that traditional Kayseri houses, which are defined as inward-looking houses, actually exhibit an outward-looking structure in terms of accessibility relations. This finding might be correlated with the extroversion trend observed in the social structure in the 19th and 20th Centuries. However, the information that is essentially needed is about the question of in what terms the houses will be characterized as inward looking. In spatial analysis; ground-floor rooms, upper-storey rooms in houses that have upper-storey, kitchens, toilets, storerooms and the outer space are predominantly located in the “non-integrated” side of houses’ integration averages and, on the other hand, courtyards, central halls, arcades, kiosks, entrance halls, halls and seki altı parts of kitchens are located in the “integrated” side. However, it is seen in most of outward-looking houses that the outer space gets a score very close to the integration average and it is sometimes located in the integrated side of the average. It is known that almost all daily functions of traditional Kayseri houses except sleeping are carried out in the courtyard especially when the weather is good. This is a fact that renders meaningful the position of the courtyard as the most integrated space within the system. In addition, hall and kiosk are the other spaces where visitors are received and the daily time is spent. Moreover, doorstep and street are social spaces where relationships are established with neighbors. In revealing the social logic lying behind houses, it is apparent that visibility relations that are based on the perception of the moving observer play roles as significant as those of permeability relations. It was observed that the houses were ranked according to their integration scores obtained in visibility analyses in a manner different than that in the spatial analysis. The ranking demonstrated that houses that have large courtyards, windows looking at the courtyard or at other inner spaces,

The reflection of religious diversity and socio-cultural meaning on the spatial configuration of Traditional Kayseri Houses


and alternative passageways in the inner space are more visually integrated, while houses that have central halls and contain more than one dwelling are more visually non-integrated. Courtyards, gardens and central halls are predominantly more visually integrated spaces. These spaces are followed by spaces for common use such as halls and kiosks, while spaces such as toilets, storerooms, upper-storey rooms and kitchens are more visually non-integrated spaces. However, the point that needs to be emphasized is that while most of ground-floor rooms exhibit a more non-integrated structure in terms of permeability, they exhibit a more integrated structure in terms of visual integration. However, it is necessary to note that this information can be controlled by closing doors and windows that render rooms more visually integrated, that is, privacy can be established by visually separating spaces. Numbers of windows on the façades of traditional Kayseri houses are not high even though they increased in the 19th and 20th Centuries. However, there exist many windows and doors opening to the courtyard. Taking this situation and houses’ visible area structures into consideration, it is concluded that houses are indeed inward-looking in terms of their visibility structures. As indicated before, houses’ integration and visual integration rankings are not in parallel with each other. This situation essentially stems from the fronts and windows that houses have in their inner spaces. In spatial analysis, surfaces that allow visibility but are not permeable are considered no different than other frontiers; however, these openings are included in the analysis in visibility analysis. These two analyses are expected to yield similar results when houses do not have windows in their inner spaces. However, the situation in the sample is different. As a result of this characteristic of the houses, it is apparent that drawing conclusions about the spatial configurations of houses on the basis of only the accessibility relations will be inadequate. It is important to demonstrate the visibility structures of these houses due to the window openings they have in the inner space. Although the spatial

and visual integration rankings do not support each other, it is seen that spatially and visually integrated spaces fit into each other. This situation indicates that the permeability and visibility structures of the inner spaces of houses function together. In conclusion, this study examined the accessibility and visibility structures of the traditional Kayseri houses. It was seen that visibility analyses are more sensitive than permeability analyses since they yield significant results by considering variables such as the openings other than the openings that makes transition possible between spaces and sizes of these openings. As the knowledge obtained through the permeability and visibility analyses demonstrate, in traditional Kayseri houses in particular and in courtyard-type houses in general, exploring the spatial configurations of houses on the basis of only the permeability relations will yield limited and inadequate results. In this respect, it is apparent that visibility analyses have important roles in such studies. It is expected that this study will be a guide to studies to be conducted on similar spatial organizations and to new designs. References Atak, Ö. (2009). Traditional Kayseri Houses in the context of Space Syntax and Visibility Graph Analyses (master thesis), Graduate School of Science Engineering and Technology, İstanbul Technical University, İstanbul. Atak Ö., Çağdaş G. (2010), Traditional Kayseri Houses In The Context Of Space Syntax And Visibility Graph Analyses, XXXVII IAHS World Congress on Housing Science “Design, Technology, Refurbishment And Management Of Buildings”, Santander, Spain. Benedikt, M. L. (1979). To Take Hold Of Space: Isovists and Isovist Fields, Environment and Planning B: Planning and Design, 6 (1), 47-65. Büyükmıhçı, G. (2005). Kayseri’de Yaşam ve Konut Kültürü, Kayseri: Erciyes University Publication. Faroqhi, S. (1987). Men of Modest Substance House Owners and House Property in Seventeeth Century Ankara and Kayseri, Cambridge: Cambridge

ITU A|Z • Vol 12 No 3 • November 2015 • Ö. Atak, G. Çağdaş


University Press. Gabriel, A. (1954). Kayseri Türk Anıtları, Translation: A. Akif Tütenk, Ankara. Gündoğdu, G. (1986). A Research on Civil Architecture Samples in Kayseri Conservation Site (In Turkish-Kayseri Sit Alanı İçinde Yer Alan Sivil Mimarlık Örnekleri Üzerine Bir Araştırma) (master thesis), Mimar Sinan University of Institute of Science Engineering and Technology,., İstanbul. Güney, İ.Y. (2007). Analyzing Visibility Structures in Turkish Domestic Spaces, 6th International Space Syntax Symposium, Istanbul. Hanson, J. (1998). Decoding Homes and Houses, Cambridge: Cambridge University Press. Hillier, B., Hanson, J. (1984). The Social Logic of Space, Cambridge: Cambridge University Press. Hillier, B. (1996). Space is the Machine; A Configurational Theory of Architecture, Cambridge: Cambridge University Press. İmamoğlu, V. (2006). Traditional Kayseri Houses, Kayseri Municipality Culture Ltd.

İmamoğlu, V. (2006). On the book “Kayseri’de Yaşam ve Konut Kültürü”, METU Journal of the Faculty, 23 (1), 83-92. Jennings, R. C. (1976). Urban Popultion in Anatolian Sixteenth Century: A Study of Kayseri, Karaman, Amasya, Trabzon and Erzurum, International Journal of Middle East Studies 7, 21-57. Kırşan, Ç. ve Çağdaş, G., (2004). Ethnicity and Domestic Space ( In Turkish, Etnik Kimlik ve Evsel Mekan), A Symposium on House Evaluation, İstanbul Technical University Faculty of Architecture, İstanbul. Tandy, C. R. (1967). The İsovist Method of Landscape Survey, in Methods of Landscape Analysis Ed. H.C. Murray, London: Landscape Research Group. Turner, A., Doxa, M., O’Sullivan, D. And Penn., A. (2001), From Isovists to Visibility Graphs, Environment and Planning B, 28(1), 103-120. Turner, A., (2003). Depthmap: A Program to Perform Visibility Graph Analysis, 4th International Space Syntax Symposium, London.

Dinsel çeşitlilik ve sosyo-kültürel anlamların Geleneksel Kayseri Evleri’nin mekân organizasyonuna yansıması Kentleşmenin ve kentlerde nüfusun ar Mekân organizasyonu ve sosyal yapı arasındaki ilişkinin en belirgin görüldüğü çalışma alanlarından biri, evsel mekândır. Evsel mekân üzerine yapılan birçok araştırma, sosyal anlam ve kültürel değerlerin büyük ölçüde konutların, mekân organizasyonları aracılığıyla yansıtıldığını böylece farklı kültürlerin farklı mekânsal modellerle kendini ifade ettiğini ortaya koymaktadır. Bu noktada, mekân dizimi ve görünür alan analizleri bina ölçeğinden kent ölçeğine varan her türlü mekân organizasyonu ile toplum arasındaki etkileşimi inceleyen hesaplamalı yöntemler olarak karşımıza çıkmaktadır. Bu yöntemler temel olarak harekete dayalı erişim ve hareketli gözlemcinin algısına dayalı görünürlük ilişkileri üzerinden ortaya konulmuştur. Hillier ve Hanson tarafından 1984’de ortaya konan mekân dizimi, sosyo-kül-

türel yapının ve süreçlerin, tek başına olmasada mekânsal düzenleyim ile kendilerini mekânda ortaya koyduklarını; sosyal yapı ve mekânın karşılıklı etkileşim içinde olduğunu ifade etmektedir. Bu anlamda mekân dizimi yönteminin en temel stratejisi, mekân örüntüsündeki bir takım değişmezleri keşfederek bunları kültüre özgü insan etkileşim örüntülerine dönüştürmeye çalışmaktır. Bunun yanı sıra Turner 2001’de, temeli Benedikt’in isovist tanımlaması ve mekân dizimine dayanan görünür alan analiz yöntemini ortaya koyar. Görünür alan mekânda belirlenen bir noktadan görünen bütün noktaların takımıdır (Benedikt,1979). Bu yönteme göre mekân organizasyonunun ortaya konulmasında, mekânların erişim ilişkilerine dayanan mekânsal tanımlamalarının yanı sıra hareketli gözlemcinin algısına dayanan görünürlük ilişkileri de önemli rol oynar. Kentsel çevrede ve binalarda görsel alanlarla sağlanan bilgi, kullanıcının yol bulmasına yardımcı olabildiği gibi görsel alanlarla

The reflection of religious diversity and socio-cultural meaning on the spatial configuration of Traditional Kayseri Houses


kullanıcıya sağlanan bilginin kontrolü de mümkündür. Dolayısıyla mekân dizimi ve görünür alan analizleri, tüm mekânsal sistemlerin ve özel olarak konutların mekânsal olarak işleyiş ve kullanıcılar yani hane halkı ve ziyaretçiler tarafından deneyimlenme biçimini erişebilirlik ve görünürlük ilişkileri üzerinden ortaya koymaya çalışmaktadır. Konutun tamamının veya belirli mekânlarının içe kapalılık - dışa dönüklük bağlamında incelenmesinde, sosyal anlamda mekâna ilişkin mahremiyet, kontrol, hane halkı ve hane halkı ile ziyaretçiler arasındaki ilişkilerin düzeyini kavramada kullanılan etkin yöntemlerdir. Mahremiyet seviyesi (kişiler arası etkileşim) fiziksel veya görünmez sınırların tanımlanması ile belirlenebilir. Fiziksel sınırlar, erişilebilirliği yani hareketi kontrol eden sınırlar olurken, görünmez sınırlar göz temasından kaçındırma gibi fiziksel sınırlar aracılığı ile sağlanan görsel bilginin kontrolüdür. Ayrıca mahremiyet düzeyi her toplum ve kültürde farklılık gösterir. Bu farklılık başta ailenin ve ait olduğu toplumun sosyal yapısı ve ziyaretçileri ile olan ilişkilerine göre şekillenmektedir. Bu yüzden, özellikle geleneksel evlerde erişebilirlik ve görünürlük yapılarını incelemek, o kültüre ait mahremiyet seviyesini anlamaya, böylelikle aile içi ve ziyaretçilerle olan etkileşimi, evsel mekâna ait işlevlerin erişilebilirlik ve görünürlük yapıları içindeki durumunu ortaya koymaya yardımcı olacaktır. Çalışma kapsamında evsel mekân, uzun yıllar farklı dini inanışlara sahip halkın bir arada yaşadığı Kayseri kent merkezinde yer alan 27 geleneksel ev üzerinden ele alınmıştır. Bu evlerin sadece 14 tanesinin sahiplerinin hangi dini inanışa sahip olduğuna ilişkin bilgi elde edilebilmiş diğerlerine ilişkin ise bilgi elde etmek mümkün olmamıştır. Ele alınan evlerden Müslüman evi olarak bilinen evlerin birçoğu avlulu ve daha organik bir yapı sergilerken, Gayrimüslim evi olarak bilinen evlerin birçoğu merkezi hollü evlerdir. Ancak bunu bir genellemeye dönüştürmek doğru olmayacaktır; çünkü bu genellemeyi bozacak bazı örnekler bulunmaktadır ve ele alınan evlerin ancak yarısına ilişkin bu anlamda bilgi elde

edilebilmiştir. Evlerin analizinde Turner tarafından geliştirilen Depthmap yazılımı kullanılarak her iki analiz türü birden gerçekleştirilmiş, evlerin mekânsal ve görünür alan yapılarının birbirini destekleyen özellikleri ve farklılıkları ortaya konulmuştur. Bu analizler sonucu avlu-bütünleşik mekânsal temanın baskın varlığı net bir şekilde görülmektedir. Geçirgenlik ilişkilerine dayanılarak avlunun, güçlü bir merkezi kontrol noktası olarak konuttaki ana ve yardımcı mekânları bir arada tutarak dış mekâna bağlayan ve bir taraftan da konutun ana mekânlarını yapılandıran kilit rolü üstlenmekte olduğu sonucuna ulaşılmıştır. Merkezi hollü örneklerde ise merkezi hol bu görevi görmektedir. Ancak merkezi hol işlev bakımından avlu gibi bir işleyişe ve kullanıma sahip değildir. Ele alınan evlerin hemen hemen hepsi, mekânlarının bir araya geliş biçimi açısından ayrışma eğilimi gösterir. Ayrıca evlerin üçte ikisi dışa dönük bir yapı sergilerken sadece üçte biri içe dönük bir yapı sergilemektedir. Bu durum çoğunlukla içe dönük bir yapıya sahip olduğu sıklıkla vurgulanan Geleneksel Kayseri Evleri’nin özellikle avlulu evlerin (çoğunluğunun Müslüman ailelere ait olmasından ötürü de daha içe dönük olarak değerlendirilen evlerin), aslında erişebilirlik ilişkileri açısından bakıldığında dışa dönük bir yapı sergilediğini ortaya koymaktadır. Bu durum, evlerin tarihlendiği geç 19. ve 20. yy.’larda, sosyal yapıda görülen dışa dönüş ile ilişkilendirilebilir. Ancak temel ulaşılması gereken bilgi, evlerin hangi açıdan içe veya dışa dönük olarak nitelendirileceğidir. Evlerin, görünürlük analizlerinde elde edilen bütünleşme dereceleri açısından, diğer analizden farklı bir sıralamaya sahip olduğu görülmüştür. Sıralamada, büyük avlulu ve avluda ve iç mekânda pencereleri ve alternatif geçişleri olan evlerin görsel açıdan daha bütünleşik sırada yer aldığı, merkezi hollü evlerin ise daha ayrışık sırada olduğu görülmüştür. Baskın eğilim olarak avlu, bahçe ve merkezi holler görsel açıdan daha bütünleşik mekânlardır. Üstünde durulması gereken önemli bir nokta zemin kat odalarının büyük bir kısmının erişilebilirlik açısından daha

ITU A|Z • Vol 12 No 3 • November 2015 • Ö. Atak, G. Çağdaş


ayrışık bir yapı sergilerken görsel bütünleşme açısından daha bütünleşik bir yapı sergilediğidir. Ancak odaların görsel açıdan daha bütünleşik olmasını sağlayan pencerelerin ve kapıların kapatılarak bu bilginin kontrol edilebileceğini yani istendiği takdirde görsel olarak da daha ayrıştırılarak mahremiyetin sağlanabileceğini belirtmek gerekir. Geleneksel Kayseri Evleri’nin dış cephelerinde yer alan pencere sayıları, 19. ve 20. yy. da artış göstermesine rağmen yine de sayı olarak çok fazla değildir. Ancak avluya açılan pencere ve kapılar çok sayıdadır. Bu durum göz önünde bulundurulduğunda ve görünür alan yapılarına bakıldığında, evlerin aslında görünürlük yapıları

açısından içe dönük olduğu sonucuna ulaşılmaktadır. Daha önce belirtildiği gibi evlerin bütünleşme ve görsel bütünleşme sıralamaları birbiriyle paralel değildir; ancak mekânsal ve görsel açıdan bütünleşik mekânlara bakıldığında, bu mekânların birbirleri ile örtüştüğü görülmektedir. Bu durum, evlerin iç mekânlarının erişilebilirlik ve görünürlük yapılarının beraber işlediğine işaret eder. Bu çalışma mekân dizimi analiz sonuçlarını, geleneksel Kayseri evlerine ilişkin elde edilen tüm bilgilerle yorumlamaya çalışarak sosyo-kültürel anlamların, evlerin mekân kurgusuna nasıl yansıdığını ortaya koymaya çalışmaktadır.

The reflection of religious diversity and socio-cultural meaning on the spatial configuration of Traditional Kayseri Houses

ITU A|Z • Vol 12 No 3 • November 2015 • 267-283

Cross border cooperation in Edirne-Kırklareli border region: New institutionalist perspectives1

Ervin SEZGİN1, Gülden ERKUT2 ervinsezgin@gmail.com • Department of Urban and Regional Planning, Faculty of Architecture, Istanbul Technical University, Istanbul, Turkey 2 gerkut@itu.edu.tr • Department of Urban and Regional Planning, Faculty of Architecture, Istanbul Technical University, Istanbul, Turkey


Received: February 2015

Final Acceptance: September 2015

Abstract This article aims to elaborate the current stage of institutionalization of cross border cooperation programme between Turkey and Bulgaria using the theoretical tools of new institutionalism. The geographical focus is on the Edirne- Kırklareli border region on the Turkish side of the border. It is argued that new institutionalist approaches provide different perspectives that shade light on different aspects of cross border cooperation and at the final stage they help to establish a coherent framework for understanding the institutionalization of cross border cooperation in the region. Drawing on interviews conducted with 49 cross border cooperation related people including public officials, civil society and business representatives and civil servants; this study conducts an institutional ethnographic analysis with a new institutionalist perspective to grasp the institutionalization process of cross border cooperation in the region. The literature review part of the article consists of two chapters: first new institutionalism theory and the different perspectives its three strands, rational choice, sociological and historical institutionalism provide; and second the scientific literature on new institutionalism in the field of cross border cooperation and border regions are elaborated. After the literature review, the article analyses the current condition of cross border cooperation in the Edirne- Kırklareli border region by using the tools new institutionalism theory provides. Finally the conclusion chapter discusses how the three strands of new institutionalism can be brought together to develop a comprehensive understanding of the institutional structure of cross border cooperation in the region. Keywords

Cross border cooperation, New institutionalism, Edirne-Kırklareli border region.


1. Introduction “Well, they are good, but people rarely see their results. They are much more like a game played by some elites in the region” (Personal Interview, 2013). “They are playing neighbourliness” would be a good title for the summary of 40 minutes long interview with one of the administrative officers in the Edirne- Kırklareli border region. The subject of the discussion was cross border cooperation (CBC), a phenomenon existing since 1950’ies in Europe and intensely finding place in the TurkishBulgarian border regions over a period of one decade, thanks to Turkey’s official candidacy for the European Union (EU) membership. The basic premise of CBC in the EU context is to contribute to the regional development of border regions while establishing neighbourly relations among border societies (Ricq, 2006). The expanding scholarly interest on border regions emphasizes their changing condition from “barriers to bridges” by stressing the increasing porosity of borders in the course of globalization that allows establishing economic, social and political links between local communities across borders (Church and Reid, 1999; Niebuhr, 2006; O’Dowd, 2003). Critiques have provided powerful arguments that a “borderless world” is rather a myth than a reality, especially in Europe, where, while internal borders are arguably diminishing, external ones are erected to fortify the “Fortress Europe” that now resembles a gated community (Johnson et al, 2011; Van Houtum, 2007). Nevertheless, cross border interactions among local border communities are not only expanding but also finding political and financial support in the EU that enables them to institutionalize and make CBC a hot topic on the local agenda (Perkman, 1999; Van der Veen and Boot, 1995). Several projects ranging from infrastructure building to people-to-people relations have been realized albeit with some scholarly and practical criticism, such as those above. The interviewed official’s criticism was a common one. During the field research on which this study is based,

49 interviews were conducted with actors related to CBC, including public officials, mayors, and NGO and business representatives. The aim was to understand the existing situation and the emerging institutional structure of CBC in the Edirne-Kırklareli border region after having more than a decade of experience. The official’s criticism was shared by many interlocutors, who in general have argued that the European money coming for such projects contributed for the amusement of a small part of the society instead of local development. Although criticism was plenty, the majority of interlocutors have argued that CBC has created a significant financial resource for local, public and civil society actors and contributed to increasing social interactions between societies among two sides of the border. Moreover with the intensification of CBC and with the involvement of the EU and the central government; a multi-layered governance mechanism unprecedented to the local community has emerged in the region. Now there is room for local actors to engage in international politics and search and compete for financial resources in the European arena. Local administrations as well as civil society organizations are succeeding in raising funds for projects that are not subsidised by the national budget. The context of CBC; the achievements realized through it; the involved and excluded actors and their critiques, from the perspective of this article, are a part of the institutionalization process of CBC. This institutionalization process is not independent from the social- political context in which it takes place, but it is also a place bound phenomenon, directly related to the local social dynamics and power relations. This article uses the tools that different strands of new institutionalism theory provide, to perform a mind exercise for understanding the institutionalization process of CBC. It is argued that this process is a complex structure that cannot be grasped completely by a single- dimensional perspective. Therefore different and sometimes contradicting approaches have to be adopted for apprehending this

ITU A|Z • Vol 12 No 3 • November 2015 • E. Sezgin, G. Erkut


This research is financially and institutionally supported by the ITU- Scientific Research Support Programme; TÜBİTAK- Science Fellowships and Grant Programmes Department; and Radboud University Nijmegen- Centre for Border Research to which the authors are thankful.


Table 1. List of interviewed organizations.        

    272;<:A8/8:.207//*2:;-2:7. :8>27,. !.9:.;.7<*<2>.  272;<:A8/<1.$//*2:;88:-27*<8:8/<1. 9:80:*66.=7-.:<1..9*:<6.7<8/27*7,2*5889.:*<287

272;<:A8/.>.5896.7<.*-8/<1.$.,87862,*7- ;8,2*5,81.;287-.9*:<6.7<


   -2:7.1*6+.:8/#:*-.*7-7-=;<:A.*-8/<1.<1.$ *7-9:83.,<;8//2,.  2:45*:.521*6+.:8/#:*-.*7-7-=;<:A".,:.<*:A .7.:*5  $B=7489:=2;<:2,<-2:7.1*6+.:8/#:*-.*7-7-=;<:A .*-8/<1.<1.$*7-9:83.,<;8//2,.  889.:*<2>.8/254 :8-=,.:;$B=7489:=-2:7. :.;2-.7<  2:45*:.521*6+.:8/0:2,=5<=:. :.;2-.7<  -2:7.1*6+.:8/.:<2/2.- =+52,,,8=7<*7<;  889.:*<2>.8/254 :8-=,.:;-2:7. :.;2-.7<  *+*.;428668-2<A@,1*70.*+*.;422:45*:.52 9:83.,<:.;987;2+5.

  -2:7.8>.:78:;129 *%2,.8>.:78: +"=+2:.,<8:8/2>2;87/8:2;*;<.:;7-.:( :83.,<7<:.9:.7.=: <1.$,88:-27*<287,.7<:.)62@.-;<:=,<=:. :.;987;2+5.<8<1.08>.:78:8/-2:7.<1.6272;<:A8/<1.$

2:45*:.528>.:78:;129 *%2,.8>.:78: +.*-8/<1.<1.$*7-9:83.,<;8//2,. 272;<:A8/-=,*<287-2:7. :8>27,.2:.,<8:*<.

$B=7489:=2;<:2,<-2:7. *8>.:78: +.*-8/<1.<1.$*7-9:83.,<;8//2,. 272;<:A8/272;<:A8/88-0:2,=5<=:.*7-2>.;<8,4 2:.,<8:8/$B=7489:=2;<:2,<//2,. .:2,2;<:2,<8>.:78:-2:7. .;*72;<:2,<8>.:78:-2:7. .62:48A2;<:2,<8>.:78:2:45*:.52 =5.+=:0*B2;<:2,<8>.:78:2:45*:.52 *+*.;422;<:2,<8>.:78:2:45*:.52  #:*4A*!.0287*5.>.5896.7<0.7,A2@.-;<:=,<=:.+=< 6*275A:.;987;2+5.<8<1.6272;<:A8/-.>.5896.7< *-2:7.7>.;<6.7<"=998:<//2,..*-8/<1.//2,. +2:45*:.527>.;<6.7<"=998:<//2,..*-8/<1.//2,.

      ;;8,2*<287/8:.>.5896.7<8/.62:48A .62:48A2:45*:.52 :.;2-.7<  '8=7-*<287/8: :8<.,<2878/*<=:*52/. .62:48A2:45*:.52.*-8/<1..62:48A//2,.  ;;8,2*<2878/'8=70=;27.;;6.727-2:7.  =5.+=:0*B=;27.;;6.7;;8,2*<287=5.+=:0*B2:45*:.52  !$72878/!86*:0*72B*<287;27-2:7.-2:7. :.;2-.7<  2:45*:.52;;8,2*<2878/2;*+5.- .895.:45*:.52 :8>27,. :.;2-.7<

    =B*//.:<*;*A201;,1885$B=7489:=-2:7.2:.,<8:  =5.+=:0*B#.,172,*5*7-%8,*<287*5201;,1885 =5.+=:0*B2:45*:.522:.,<8:

        :80.687;=5<*7,A74*:* :83.,<*7*0.:

     -2:7.=72,29*52<A%2,.*A8: 2:45*:.52=72,29*52<A *#8?75.:4 +.*-8/<1.<1.$*7-9:83.,<;8//2,.

"=5805==72,29*52<A-2:7.*A8: $B=7489:==72,29*52<A-2:7.*A8:

.;*7=72,29*52<A-2:7. **A8: + :83.,<;!.;987;2+5. 2:,*;*521=72,29*52<A$B=7489:=-2:7.%2,.*A8: .62:48A=72,29*52<A2:45*:.52*A8: 8/,*B=72,29*52<A2:45*:.52*A8: %2B.=72,29*5<A2:45*:.52.*-8/<1.<1.$*7- 9:83.,<;8//2,. *+*.;42=72,29*52<A2:45*:.52#8?75.:4 =5.+=:0*B=72,29*52<A2:45*:.52 **A8: +%2,.*A8: ,.*-8/<1.<1.$*7-9:83.,<;8//2,.

complexity; and the new institutionalism theory allows swinging across a variety of perspectives and establishing a more coherent view of institutionalisation of CBC. The three strands of new institutionalism, historical, sociological and rational choice institutionalism are adapted to the CBC context to grasp these multiple perspectives. They are used to approach, observe and interpret the same issue from different viewpoints and reach a comprehensive understanding of this newly emerging institutional structure.

          $72878/.:2,=72,29*52<2.;-2:7.".,:.<*:A.7.:*5  #:*4A*4.7<$72878/=72,29*52<2.;8/*;<&.;<#1:*,. #.42:-*0".,:.<*:A.7.:*5  #:*4*+#:*4A*.>.5896.7<$7287-2:7.".,:.<*:A .7.:*5

       !=5270 *:<A.*-8/<1. :8>27,2*5//2,.27 -2:7.   *27998;2<287 *:<A.*-8/<1. :8>27,2*5 //2,.27-2:7.   *27998;2<287 *:<A.*-8/<1. :8>27,2*5 //2,.272:45*:.52

The methodological approach of the research is institutional ethnography, which aims to â&#x20AC;&#x153;discover the social, rather than theorizing it, beginning with actual people, their doings, and how their doings are coordinatedâ&#x20AC;? to formulate an institutional regime (Smith, 2008, p. 433). This method allows the researcher to establish links between interviewed individuals and their interpretations of institutions and the institutional structure in general. This approach enables mapping a given institutional setting with respect to

Cross border cooperation in Edirne-KÄąrklareli border region: New institutionalist perspectives


its constituent actors, however the impacts of the wider social and political factors remains latent in the interpretations of institutional actors and has to be re-interpreted by the researcher. For this purpose the institutional ethnographic approach needs to be tuned to include a “thick description” (Paasi, 1996) that establishes links, this time with the social and political factors that affect the institutional setting and also with the ways of its interpretation by the institutional actors. As mentioned above, the field research of the study is based on 49 interviews with people related to CBC in the region. The list of interviewed organizations is presented in Table 1. For the sake of confidentiality, the names of persons interviewed are identified neither on the table, nor on the citations in the remainder of this paper. The geographical context of the study was set to cover the north-western border provinces of Turkey, namely Edirne and Kırklareli. Although not every district of these provinces is attached to the border, and even some of them have denser interactions with the internal regions of Turkey, all of these districts are taking part in CBC and contribute to its institutionalization. The remainder of this paper will have the task of exploring the institutional structure of CBC in the EdirneKırklareli border region using three different perspectives of new institutionalism theory. The structure of the paper will be as following: Chapter 2 describes the new institutionalism theory and the different perspectives the three main strands bring. This is followed by a literature review on new institutionalist approaches in CBC and border regions context in Chapter 3. The 4th Chapter focuses on CBC in Edirne- Kırklareli border region and explores the existing situation by looking through the different lenses of new institutionalism. The concluding chapter summarizes the discussion and comments on how these three strands can describe the same picture from different perspectives. 2. New institutional theory- A literature review The organization of social life is done through institutions. The development

of a social order and the evolution of societies, ranging from everyday practices to the highest levels of organization are born of various institutional arrangements; and the evolution of societies is shaped by institutions and their transformation over time. As North (1990, p. 22) argues, 90% of “daily life” activities are made possible by being routinized and regularized by institutions. Institutions are seen as both results of historical processes and powerful subjects capable of influencing the evolution of human society at the same time. As Putnam (1993, p. 7) argues, institutions are shaped by history and yet work to shape politics. This duality has been problematized under the structure and agency dichotomy. From the first point of view, the historical evolution of societies with its various aspects - including norms, practices and power struggles – determines the formation of institutions, since they have emerged as a result of this evolution in order to give a shape to it in a way that reflects these social dynamics. The contrasting approach pays most credit to individuals motivated by rational choice, who use and transform institutions (or create new ones) in order to achieve desired ends. In this respect, institutions are considered to be aggregated outcomes of individual behaviour that serve to reduce uncertainty and limit the choices of individuals in order to enable them to calculate accurately their goals and actions. Contrary to the structuralist approach, this point of view partially neglects the constraining role of structure on the choices of individuals and assumes that any rational behaviour would be repeated in the same way in any kind of structural context. New Institutionalism emerges as a third option in this context, which ascribes a more autonomous power to institutions in contrast to both the structure and agency perspectives (March and Olsen, 1984). Without neglecting both aspects of social life, new institutionalists recognize that institutions move in both directions by imposing structural constraints on the preferences of individual actors on the one hand, and giving access to par-

ITU A|Z • Vol 12 No 3 • November 2015 • E. Sezgin, G. Erkut


ticular actors and collective actions to transform the social structure on the other (Lowndes and Roberts, 2013). The political and social problems in a given setting can be viewed as part of the social structure that also influences institutions themselves. However, the perception of these problems by political actors and the methods chosen to solve problems (i.e. the preferences of institutional actors) affect the future of the public realm and its institutional structure (March and Olsen, 1994). Lowndes and Roberts (2013, p. 28) argue that the distinctive characteristic of New Institutionalism is to avoid the superficial and taken for granted understanding of institutions (as well as structure and agency), and to critically look “at the way in which they embody values and power relationships”. Although the key role of institutions in forming social practices is commonly accepted, New Institutionalism cannot be accepted as a stand-alone theory capable of conceptualizing institutions. Using the basic points of the structure and agency dichotomy, new institutionalists have developed several strands of institutionalism, three of which have grasped the three most common positions regarding the roles of agency, structure and society in institutional contexts. These are not mutually exclusive aspects, separately conceptualizing their own institutional realities; rather, they complementarily square the circle of the agency-institutions-structure triad that allows one to fully grasp the origins of social interactions, their role in the formation of institutions and the emergence of the social structure as something more than the sum of actors’, individual and institutional practices. The study of institutions also enables the adoption of a multidimensional perspective while investigating the institutionalization of a particular policy field. The three dominant schools of new institutionalism are commonly referred to as rational choice institutionalism, historical institutionalism and sociological institutionalism (Hall and Taylor, 1996; Immergut, 1998; Lowndes and Roberts, 2013). These three strands represent the three main streams of institutionalisms, but also

are complemented by various other approaches such as normative, empirical, international, network, discursive and feminist institutionalisms (Lowndes and Roberts, 2013). Rational Choice Institutionalism has emerged from behaviourism and rests on the basic premise that humans are rational beings who act with the aim of maximizing their self-interest through careful cost-benefit analysis. From this perspective institutions are accepted to reflect the desires and habits of the individuals who establish them. Informal relations, political leaders and their followers, interest groups, and ideologies (instead of formal modes of organizations) are among the interest areas of this type of institutionalism (Bolat and Seymen, 2006). The role of institutions according to rational choice institutionalists is to reduce transaction costs in social and economic interactions by avoiding uncertainty and reducing risks in social relations, especially those arising from trust problems (North, 1990, Ostrom 2005). As Lowndes and Roberts (2013, p. 23) explain, from the rational choice point of view, the existence of institutions is a result of cost-benefit analysis. An institution can only exist when the costs of maintaining it do not exceed the benefits gained with it. So, larger institutions, such as the state, represent a more complicated form of social organization that has emerged from the continuous efforts to reduce transaction costs in social interactions. As North (1990) argues, by regulating the exchange relationship between two parties, institutions provide the basis of political/economic systems. Although rational choice institutionalists consider rules that set a wider framework as well, these rules are usually either accepted as given or elaborated at a micro-level only to include those that are set by individuals or groups for the regulation of a particular situation (see for example Ostrom, 2011). Another shortfall of rational choice institutionalism is argued to be the neglect of long-term targets and calculations of actors when taking action, and assuming that the short-term interest is the dominant source of motivation in their actions (Lowndes and

Cross border cooperation in Edirne-Kırklareli border region: New institutionalist perspectives


Roberts, 2013, p. 37). Sociological institutionalism, which is almost opposed to rational choice institutionalism, presumes that behaviours of actors are context-driven; in other words, social structures (e.g. cultural conventions, norms and cognitive frames of reference) determine not only the options and choices of actors, but also actors’ ways of thinking and perceiving the world. Hence individuals’ actions are not shaped by their own decisions taken independently to reach their desired ends, but are dependent on the prevailing social context (DiMaggio and Powell, 1991). Therefore, an actor’s choices related to the same subject will vary greatly under different circumstances, which depend on the social structure in play (Immergut, 1998). Institutions constitute an “infrastructure” for society, providing the reference points for actions of actors (Lowndes and Roberts, 2013). From this perspective, individual choices are socially determined by institutions, which are also bound by practices emerging from the wider social structure. Actors, whether individuals or organizations, are “embedded” in the social structure (Amin and Thrift, 1994, p. 12) which provides a “frame of meaning” that guides their actions (Hall and Taylor, 1996, p. 947). According to March and Olsen (1984, p. 741), sociological institutionalism differs from rational choice institutionalism by conceptualizing actions of individuals as the “fulfilment of duties and obligations” instead of “individual values and expectations”. As DiMaggio and Powell (1991, p. 28) argue, “cultural frames establish approved means and define desired outcomes, leading business people to pursue profits, bureaucrats to seek budgetary growth, and scholars to strive for publication”. Historical institutionalism is related to the long-term evolution of institutions in a wider context. Instead of focusing on individuals or organizations as rational choice and sociological institutionalisms do, the historical institutionalist approach investigates the wider institutional structure (i.e. the nation state or the world system) and its long term, historical dynamics (Hall and Taylor, 1996; Lowndes and

Roberts, 2013). Immergut (1998, p. 16) points to the difference between sociological and historical institutionalism by arguing that while the former is interested in “the ways in which organizational rules and procedures coordinate the action of independent individuals”, the latter focuses on themes related to power and interest. Hall and Taylor (1996, p. 941) stress that historical institutionalism is interested in the uneven distribution of power, and argues that “institutions give some groups or interests disproportionate access to the decision-making process”. The term “path dependence” takes a key role in explaining the historical institutionalist approach. Path dependence argument presumes that when policymakers choose a particular path; the aggregate impact of actions taken in the aftermath produces a snowball effect, which is difficult to roll back or alter in terms of direction. Consequently, a “powerful cycle of self-reinforcing activity” is created that strongly influences the decisions of actors within institutions (Lowndes and Roberts, 2013, p. 39). Putnam (1993) refers to path dependency in his work in the following way: …what comes first (even if it was in some sense “accidental”) conditions what comes later. Individuals may “choose” their institutions, but they do not choose them under circumstances of their own making, and their choices in turn influence the rules within which their successors choose. (Putnam, 1993, p.7).

However, the intentional change or transformation of institutions and institutional structure is not accepted as totally impossible. Institutional change can and does occur throughout revolutionary moments of history, in which the “periods of continuity [are] punctuated by ‘critical junctures’, i.e., moments when substantial institutional change takes place thereby creating a ‘branching point’ from which historical development moves onto a new path” (Hall and Taylor, 1996, p.942). Immergut (1998) stresses that, from a historical institutionalist point of view, particular events during the course of history originate from the free choice of powerful individuals; these choices, however, always bear

ITU A|Z • Vol 12 No 3 • November 2015 • E. Sezgin, G. Erkut


the burden of individual and social histories. The establishment of particular institutions in a specific moment of time, such as recognizing the right to private property during monarchical rule, is a result of other social and historical processes that motivate the transformation of the whole institutional structure. Despite their diversified departure points, these three strands intersect in some common points in their approach to institutions. These intersections consolidate and establish a common ground of new institutionalism that, according to Lowndes and Roberts (2013), represents the “third phase” of New Institutionalism Theory. Lowndes and Roberts (2013) summarize this consolidation under five main topics: (1) It is commonly accepted that structure and agency have a dialectical relationship, in which neither one, nor the other is determinative on the other or on the institutional structure in general. Instead they are in continuous interaction that transforms both of them throughout the course of history. (2) An institutional structure is a complex system that enables institutions to operate through three main modes: rules, practices and narratives. (3) Whether in short or long run, institutions are subjects and objects of social change. Reflecting the dialectical relationship between structure and agency, they can trigger social change or can be affected by the structural forces that either impose or require institutional continuity. (4) Institutions are designed social phenomena and their design process reflects the power relations in society. (5) Even among institutions of the same social domain an institutional diversity always exists. This is a result of different social/ structural dynamics that prevail in the society as well as the influential agents that are in a position to take critical decisions for institutions. These five topics do not only form the common ground of the three

strands, but also offer an exclusive conceptualization of the role of institutions in social life from a new institutionalist perspective. 3. CBC and new institutionalism This section evaluates the scientific literature related to CBC, cross border regionalism and new institutional theory. The aim is to provide examples on the use of new institutionalism in CBC, as well as to establish a ground for discussion the CBC experience in the Edirne- Kırklareli border region with this perspective. Studies that are not directly related to CBC but are coming from regional science related disciplines are also included to extend the scope of the evaluation and provide background for further discussion. A similar new institutionalist analysis approach previously was adopted by Helena Ekelund (2014), who demonstrated the applicability of the new institutional theory in the EU context by discussing the establishment of Frontex, the border patrol agency of the Union. According to her, “the timing of establishment; the decision on the agency’s tasks, role and mission; [and] the decision on a specific institutional design and management” (Ekelund, 2014, p. 110) can be explained by using different new institutionalist strands. From a rational choice perspective, the coordination of border-related activities of member states and the use of expert knowledge can be viewed as transaction cost-reducing activities. On the other hand, the emergence of Frontex cannot be understood solely as a cost-cutting solution. The social and historical contexts are highly influential in the institutional design process. This context is the enlargement of the EU, which requires its own institutional structure to manage the extensive functional areas of the Union. For Ekelund (2014), the enlargement waves that included southern and eastern European states are critical junctures in which crucial decisions were taken to pave the way for the establishment of Frontex. Implications of rational choice institutionalism are most easily observed at the local and regional levels, from which Elinor Ostrom (2005) developed

Cross border cooperation in Edirne-Kırklareli border region: New institutionalist perspectives


the Institutional Analysis and Development Framework, which was based on a large number of local case studies across the world. Through these case studies, Ostrom and her colleagues extensively elaborate how local actors organize around institutions and use institutional practices to overcome trust problems. Extending this debate to the regional scale, Rodriguez-Pose (2013, p. 1037) argues that local institutional arrangements prove to be more effective in generating economic development than nation state governments, since “the national scale can be too distant, remote and detached in order to be effective in mobilizing organizations”. Rodriguez- Pose’s (2013, p. 1042) conclusion is that regional development strategies should be “specifically tailored to the potential of placebounded institutions” in order to make significant interventions for development. However, as a result of the rational choice institutionalist thinking, Rodriguez-Pose’s focus on institutions from a regional development perspective suggests looking away from the institutional characteristics of a region, focusing instead on “institutional arrangements, which represent barriers for the efficiency of other factors influencing economic development” (p. 1043). The management of common-pool resources and transaction cost-reducing roles of institutions provides considerable explanation for the role of institutions during the institutionalization process of CBC. Although in the context of inter-municipal cooperation, which does not include a cross-border aspect, Hilvert and Swindell (2013), Carr and Hawkins (2013), and Delabbio and Zeemering (2013) conceptualize cooperation as common-pool resource management. According to them, cooperation activities emerge from the need to reduce transaction costs, especially in urban management, by sharing and coordinating the joint use of resources, collaborating in infrastructure construction and creating joint institutions for providing services. This is also the case in cross-border regionalism, where grass roots, cross-border initiatives emerge to mobilize common endogenous re-

sources to solve problems and generate regional development. Delabbio and Zeemering (2013) argue that institutional context is a determining factor for collective action. In addition, local officials prefer to take part in collective action when a cost-benefit analysis reveals different aspects of the institutional structure, such as career risks, public resistance, expected outcomes of collaboration and transaction costs. Hilvert and Swindell (2013) also argue that inter-local cooperation becomes satisfactory for decision-makers only when transaction costs arising from cooperation are exceeded by the benefits of collaboration. One of the scholars who first interpreted CBC using a new institutionalist approach is Markus Perkmann. He argues that … CBC has to be seen as an aggregate outcome of various relatively decentralized processes of institution building with strong involvement by non-local actors. Cross-border initiatives cannot be assumed to have single and coherent objectives. Rather, a multiplicity of actors operates in an institutional context of opportunities and constraints that is not predominantly of their own making. As a consequence of their actions, the institutional setting itself undergoes continuous changes resulting in irreversible and historically specific trajectories. Such ideas about the building and evolution of institutions can be extrapolated from a more sociologically inflected `new institutionalism’ recently emerging as a supra-disciplinary paradigm in a range of fields. (Perkmann, 1999, p. 660)

Perkmann departs from rational choice institutionalism and aligns his arguments with a historical institutionalist approach that links institutional change with “strategic action (path shaping) and evolution (path dependence)”. Perkmann (1999, p. 660), rejects the argument of the neutrality of institutions and argues that “they privilege certain actors, certain time and space horizons, and certain strategic agendas over others”. Institutional structure is also influential over the behaviours of CBC actors and since it is predominantly formed within the nation state framework, it remains inefficient in supplying actors with the

ITU A|Z • Vol 12 No 3 • November 2015 • E. Sezgin, G. Erkut


necessary tools for cross-border governance. In tandem with Perkmann (1999), Church and Reid (1999) attempt to conceptualize the institutional framework of CBC between England and France using the new institutional theory. Rather than focusing on the relationship between institutions and CBC in general, they focus on two specific aspects of the institutional structure: its thickness and territorial embeddedness. Their findings suggest that although intensifying CBC activities contributes to increasing institutional thickness, the lack of genuine cooperation results in more intervention on behalf of the EU and national governments in the facilitation and supervision of CBC. With regard to territorial embeddedness, Church and Reid (1999) make two distinct observations. First, cross- border spaces mostly remain imaginary spaces, as wishful thoughts of politicians, perhaps as an impact of external influence without being supported by “economic and cultural transfrontier links”. Second, spaces of CBC usually have a “flexible territorial characteristic” that releases the boundedness of administrative divisions and scales and constitutes its own spatial framework. These two studies align with sociological and historical institutionalisms and evaluate CBC processes as context-bounded phenomena emerging from the historical conditions that both necessitate and constrain it. Putnam’s (1993) longitudinal and extensive study of Italian regional governance institutions reaches the conclusion that mirrors sociological institutionalism, though it also includes elements from historical institutionalism. Putnam (1993) investigates the establishment of local governments in Italy with a longitudinal survey. Within a period of 14 years after their establishment, regional governments in Italy displayed a diversified path of governance, which resulted in success for some and failure for others. Focusing on the reasons of that diversification and by conducting two surveys with a 10 years interval, Putnam argues that democratic traditions are among the most significant factors in determin-

ing institutional success. According to him, “the practical performance of institutions… [is] shaped by the social context within which they operate” (p. 8). These studies demonstrate that new institutionalism theory provides efficient tools to understand the complexity of CBC. However, most of them do not intend to provide a coherent picture of the institutional structure of the subject they investigate and rather remain focused on the interpretation and critique of their problematic, usually by adopting the approach of one of the three different strands. The remainder of this study aims to choose the first path and comprehensively evaluate CBC in Edirne- Kırklareli border region from the perspectives of the three strands of new institutionalism. While doing so, abovementioned five topics in which the three strands converge will be used as reference points for interpretation and for providing a coherent view of the institutional structure of CBC in Edirne- Kırklareli border region. 4. CBC in Edirne Kırklareli border region: New institutionalist perspectives The administrative regions of Bulgaria and Turkey that take part in CBC were part of the Thracian region under the rule of the Ottoman Empire for four centuries. Thrace was inhabited by a heterogeneous society of Muslims, Christians and Jews. The division of geographical units into administrative territories was conducted very vaguely along religious lines, also a major source of identity among inhabitants of the region. Nonetheless, cultural interaction, population movement and economic interaction were also considered a general characteristic of the region (Manos, 2005; Mazower, 2001). The Greek and Bulgarian revolutions resulting in independence in 1832 and 1909 respectively were the roots of the tripartite hostility between these two nation-states and the Ottoman Empire. Modern Turkey, established in 1923, inherited its part of that enmity from the Ottoman Empire, although no war has been fought since the establishment of these nation states. A part of

Cross border cooperation in Edirne-Kırklareli border region: New institutionalist perspectives


that enmity lies in the self-perception of the Turkish nation state as the heir of the Ottoman Empire. In addition, Turks were identified with Islam, which was the source of religious enmity and accompanied political disputes that have made Turkey a “negative point of reference for [the] symbolic construction of the local political communit[ies]” (Gkintidis, 2013). The refugee exchanges between Turkey and Greece and Greece and Bulgaria have caused the homogenization of the previously heterogeneous region (Karakasidou, 1997). After the establishment of new nation states, the long standing hostilities and the exchange of minorities between them, the basic characteristic of the border regions in the Thrace region during the 20th century could be described as alienation, since the dominant narrative of the respective nation states was nationalism accompanied by strong political and administrative dependence on the centre in combination with military fortification along the borders (Gkintidis, 2013; Mazower, 2001). With the end of the Cold War the border milieu started to change slowly but significantly. First, trade liberalization in Bulgaria has increased the volume of trade as well as the cultural contacts between the two countries and bordering provinces. Second, with the official EU candidacy of two countries and Bulgaria’s membership in 2007, the EU has become an important facilitator of bilateral relations. The EU’s external, enlargement, integration and regional policies, programmes and funds have reshaped the context in which the border region is defined. The pre-accession instrument in general and CBC in particular have contributed to this change significantly. Regional policies of the EU and the new regionalist turn in national politics provided new opportunities and perspectives for regional development, and also forced the regions to rely less and less on decreasing central state investments and to look more towards utilizing endogenous resources. The CBC programmes of the EU emerged as one of these opportunities that encouraged border regions to turn toward the other side of the border instead to their capital cities

(Sezgin and Erkut, 2014). The historical institutionalist approach allows interpreting CBC as a part of Turkey’s alignment with EU’s policies as well as a part of the wider economic and political dynamics prevailing within the EU. From the historical- institutionalist lens CBC in Edirne- Kırklareli border region collides with the EU candidacy of Turkey, which can be considered as a critical juncture in country’s history. Although Turkey’s orientation towards EU membership dates back to 1960’s, the official candidacy status granted in 2005 marks a considerable change in Turkey- EU relations, including the rise in the EU funds. CBC in the EU is a part of the Union’s European integration and regional development goals. Turkey’s regional policies also have changed in accordance, shifting from redistribution of resources at national level until 1990’s to entrepreneurialism and competition in 2000’s (Bayırbağ, 2013; Dulupçu, 2005). In the context of border regions, this entrepreneurial position necessitated a transformation in the perception of the other side of the border, particularly at the local level. Enmity and alienation had to be replaced with a partnership perspective, since the neighbouring country and region were no longer negative reference point for national identity but an endogenous resource that holds the potential to enlarge the existing market and compete internationally. The structure within which CBC projects are realized can best be defined with its constituting rules and regulations. The Procurement and Grants for European Union External Actions tendering procedure (PRAG) is the main EU-originated law, which affects CBC processes in the Edirne- Kırklareli border region. This law is a set of rules that define grant and procurement mechanisms of EU funds (EC, 2014). The contents of these rules are outside the scope of this study. Rather, particular importance is given to the contradictory, dual structure that has emerged over the course of the CBC projects’ implementation process. Grants are allocated according to PRAG rules and the use of these grants is also subject to tendering procedures put in place

ITU A|Z • Vol 12 No 3 • November 2015 • E. Sezgin, G. Erkut


by the same rules. More concretely, spending money within a CBC project can only be done in accordance with PRAG procedures. However, local actors, particularly from the public sector, are also obliged to obey the national procurement law. In the event that PRAG rules and the national procurement law contradict each other, it is not clear how project beneficiaries should act. During the field study, it was observed that across the board, all public project beneficiaries were under stress during the tendering process because of this contradiction. On the one hand, complying with PRAG procedures is necessary to ensure project payment, as project spending is done in advance and reimbursement of funds is completed only after inspection by independent controllers. On the other hand, public organizations can be audited anytime by national authorities, looking to ensure that expenditures are in compliance with the relevant public procurement law despite the fact that some items of which contradict PRAG tendering procedures. As a temporary solution, the Ministry of EU Affairs authorities have written official letters to the Ministry of the Interior explaining the situation and asking for the necessary understanding for project beneficiaries (Personal Interview, 2013). Despite informal efforts at the central state level, the absence of legal clarity for some interlocutors is one of the most discouraging factors in terms of participation in CBC Independent from the legal transformation aspect, tendering procedures have complied with PRAG criteria without a significant problem until now, and have had transformative effects at the local scale, even if not always in a positive way. Various interlocutors have described their way of doing business, the “Turkish style”, as focusing on the things that have to be done instead of the procedures themselves. They rely on informal networks and trust to accomplish their missions, and rate legal procedures as being of secondary importance. Therefore, for the sake of efficiency, even in the public sector, they are used to completing the practical work first and setting out to

“filling out the necessary paperwork” later. This is an informal relationship based on personal relations and trust in which local public and private actors openly credit each other and believe that they will act in a way that will not cause any problem in the future. Yet the procedures of CBC do not leave room for the Turkish style of doing business, which, according to most of the project beneficiaries, leads to inefficiency and an increase in the costs of projects. Despite the legal difficulties described above, theoretically, the process of applying for CBC funding is open to any institution located in the border region. However, the laws and regulations defined by both the EU and the Turkish government constrain a large number of organizations that would otherwise participate in CBC. The constraints include first the IPA procedures that allocate 20% of the project budget as an advance and allow for reimbursement only after tenders are completed. This is a process that practically eliminates organizations with low budgets. Second national regulations restrict organizations seeking CBC funding that owe tax and social insurance debts to the state. Third, for many public organizations, such as schools, legally it is impossible to apply for CBC funding, since their annual budget is centrally planned to cover only substantial costs such as salaries and maintenance. However, CBC procedures necessitate the allocation of internal resources, even if only temporarily. During this field study, it was observed that most project beneficiaries have participated in more than one project and regularly apply for CBC funding. These organizations also constitute a considerable share of institutions with sufficient financial and human resources to both prepare and implement projects. Consequently, CBC remains in the hands of a limited number of people and it is difficult for a newcomer to enter the club. Organizations in the region are constrained by laws, regulations and the level of socioeconomic development of the region, and only the forerunners in terms of financial and human resources manage to benefit from CBC funding.

Cross border cooperation in Edirne-Kırklareli border region: New institutionalist perspectives


The legal structure constituted by the EU and Turkish laws and legislations limits the institutional framework of CBC. And its interpretation necessitates the use of the sociological institutionalist lens. Accordingly CBC practices that take place in the Edirne- Kırklareli border region are constrained by the various multi-scalar factors that stretch from the EU to the national levels. At the EU level the PRAG procedures impose their own way of doing business, which interrupts the “Turkish Style” and attempts to change it according to predefined standards. This institutional change also faces considerable resistance as the Turkish national authorities hesitate to take necessary steps for harmonizing PRAG procedures and national legislation. In a similar manner, constrains arising from the national legislation also restrict the CBC actors in the region. At this scale restriction appears as setting the legal limits that while enable some actors to participate in CBC, restrict others. Within this context, sociological institutionalism’s perspective allows interpreting CBC as a limited institutional structure in which only selected actors are allowed to take part. This perspective opposes the rational choice institutionalism by arguing that a particular institutional context is not created by and through the participation of all rationally behaving actors but by only a limited part of them. The legal system in this context represents the social structure that from the sociological institutionalist perspective determines the institutional framework of CBC. However motivations of CBC actors and the practices they have established during their personal CBC histories still contains room for rational choice institutionalist interpretation. It is possible to conceptualize CBC as an imaginary common-pool resource where actors compete and cooperate to use and manage EU funds. Cooperation between partners displays behaviours based on trust relationships. A typical example is the partner choosing process, where organizations prefer to participate in CBC with partners they already know. Moreover, among

project beneficiaries who have managed more than one CBC project, it is common to cooperate with the same partner because of trust relationships based on previous experience. Another common practice among partners is to apply simultaneously to two projects, where one partner becomes leader in one project and the second in the other. This allows them to use their resources more efficiently, reduce the risks of not being awarded by funding and increase their share taken from the pool. There is a general agreement that CBC funding, particularly in terms of infrastructure projects, provides opportunities for municipalities. However, not all of them can be involved in CBC and use EU funding. As mentioned above, any institution that would like to participate in CBC must not owe taxes or social insurance debts to the state. This precondition reduces the number of municipalities participating in CBC. Some of them have no choice but to abstain from CBC, since it is impossible for them to pay their debts. For the others, they face the dilemma of locating enough capital to pay their debts in order to participate in CBC (and also other programmes, such as those of RDAs) or opt out altogether. Usually officials perform a cost and benefit analysis prior to making a final decision. If they see that the benefit from participating in a project exceeds their debts they chose the first option. Moreover, the tendency towards participating in infrastructure-related CBC projects can also be (partially) explained by the same argument, since these projects are argued to be more cost effective than soft projects, which according to some interlocutors, were considered a “waste of time and money”. 5. Conclusion The new institutionalist strands provide explanations for the institutional structure of CBC from their own perspectives. In general this study argues that CBC has emerged as a result of several socio- economic dynamics that have affected the social and institutional organizations of human societies along with CBC. The enlargement, integration and regional policies of the

ITU A|Z • Vol 12 No 3 • November 2015 • E. Sezgin, G. Erkut


EU as well as their preconditioning dynamics such as globalization and the end of the cold war, from a historical institutionalist perspective, have created a path dependent process resulting with CBC. CBC should be understood as an integral part of the social structure of the border region in which it takes place. Although historical dynamics are influential in the formation of the social structure, the place and context based conditions vary from one border region to the other, necessitating developing a unique conceptualization for each case. The sociological institutionalist perspective allows positioning CBC in the Edirne- Kırklareli border region within its context by providing links to the EU and national laws, the contradictions between them, the given bureaucratic structure of the Turkish nation state and CBC practices. This strand of institutionalism allows interpreting CBC as a context bound phenomena reflecting the established social structure in the region. Independent from the social context, however, CBC actors have some common motivations and expectations that the rational choice strand of new institutionalism allows to grasp. Especially the profit maximizing behaviour of CBC participants that orients them to prefer infrastructure projects over the soft ones, and to tend to cooperate with the same partners because of trust related issues are among the best examples of the rational choice point of view in CBC. Apart from the explanations the three strands of new institutionalism individually provide, new institutionalist theory also proves to be useful in understanding the ongoing social change in the region through the five consolidation points of the three strands. The tension between the national laws and EU regulations, for example, reflects the duality between continuity and change. If a separation between the three layers of CBC governance, the supranational, the national and the regional, would be done for analytical purposes, then the national level can be defined as the one that stands for continuity against the tendency to change imposed by the supra-

national and the regional levels. Here change is imposed mainly by global dynamics that necessitate a structural transformation of the economic and political system. This transformation is capitalist in nature, hence strongly related to the global dynamics of capital accumulation. Accordingly, the concentration of political power is shifting from nation states; upwards towards global and supranational institutions and downwards towards subnational regions (Jessop, 2003; Swyngedouw, 1997). Within this context change appears as result of the path dependent processes of capital accumulation. And the resistance of nation state’s institutions is related to taking control over the newly emerging policy fields, such as CBC and their related institutions. The continuity-change dichotomy is also observed in the design of institutions. In the case of CBC, the conflict between the national and the EU regulations is a reflection of this dichotomy. On the one hand CBC in Edirne- Kırklareli has been initiated by the financial and institutional support of the EU; hence it is designed in accordance to its institutional framework, which is also applied with almost the same procedures to other candidate countries. On the other hand, a radical institutional change triggered by CBC is not possible because of the already established national administrative structure that has its own agenda and the lack of institutional capacity in the region. Instead regulation, supervision and even participation to CBC are done by existing public and civil society organizations within the national institutional framework. In this context, from a sociological institutionalist perspective CBC is representing one of the various aspects of public administration, in which existing regulations and resources, especially in terms of human capital are used; and established ways of doing business are applied. Hence actors involved in CBC are constrained by the national institutional structure which imposes its own conditions against the external pressure for change coming from the EU. As a result, although a variety of organizations are involved, an institutional diversity is hardly observable due the structural constrains that

Cross border cooperation in Edirne-Kırklareli border region: New institutionalist perspectives


limit and shape the general framework of CBC in the region. However, through CBC practices a change, albeit slow in pace is initiated in the region. This change is partly a response to new regionalist policies that reduce the amount of investments of the central state and force regional/ local authorities to compete for resources necessary for their basic service provisions. In the lack of the central state investments local institutions tend to increase their profits and perceive CBC as an endogenous resource that contributes to their competitiveness in front of the internal regions and they invent methods for bypassing the institutional factors that constrain them from using it. These methods become the institutionalized practices of CBC in the region. Change is also partly initiated by the will of local society, as a result of reduced central state control and increasing interactions among the two sides of the border. The existence of the border has become a resource, the efficiency of which can be maximized only through its mutual management. CBC, although invented and developed by the EU’s institutional framework, is being used as a catalyser for that purpose. In this sense, as Elinor Ostrom (2005) has pointed out from a rational choice point of view, cooperation, rather than competition emerges as the most efficient way of dealing with the changing political and economic conditions. Practices of cooperation and competition shape the institutionalization process of CBC in the Edirne- Kırklareli border region by large, but they also have emerged as a response to the central state’s control mechanisms of CBC, which in turn are a response to the EU policies that intervene to the established national institutional structure. The institutionalization of CBC in this way comes full circle: CBC has emerged within a global and international context as a result of various economic and political dynamics. Especially at the national scale, it has been applied within the framework of existing legal structure and institutional context established for different purposes; and finally has been shaped by the institutional practices that are most

creatively invented at the local level to overcome the institutional constrains. Acknowledgements This paper is produced from the PhD dissertation entitled “The Impacts of State Transformation Processes on Border Regions: A Reading Through Cross Border Co-Operation” prepared by the corresponding author and with the institutional and financial support of ITU- Scientific Research Support Department; TÜBİTAK- Science Fellowships and Grant Programmes Department; and Radboud University Nijmegen- Centre for Border Research. The authors wish to thank to these institutions as well as to the two anonymous referees. References Amin, A. and Thrift, N. (1994). Living in the global in globalization, In Amin, A. and Thrift, N. (Eds.), Institutions and regional development in Europe, (pp. 1-22). New York, USA: Oxford University Press. Bayırbağ, M. K. (2013). Continuity and change in public policy: Redistribution, exclusion and state rescaling in Turkey. International Journal of Urban and Regional Research, 37 (4), 11231146. Bolat, T. and Seymen, O. A. (2006). Yönetim ve örgüt düşüncesinde kurumsalcılık, yeni kurumsalcılık ve kurumsal eşbiçimlilik, Fırat Üniversitesi Sosyal Bilimler Dergisi, 16 (1), 223–254. Carr, J. B. and Hawkins C. V. (2013). The costs of cooperation: What the research tells us about managing the risks of service collaborations in the U.S., State and Local Government Review, 45 (4), 224- 239. Church, A. and Reid, P. (1999). Cross-border co-operation, institutionalization and political space across the English channel, Regional Studies, 33 (7), 643- 655. Delabbio, D. J. and Zeemering, E. S. (2013). Public entrepreneurship and interlocal cooperation in county government, State and Local Government Review, 45 (4), 255- 267. DiMaggio P. J. and Powel, W. W. (1991). Introduction. In Powel W. W. and DiMaggio, P. J. (Eds.), The new institutionalism in organizational anal-

ITU A|Z • Vol 12 No 3 • November 2015 • E. Sezgin, G. Erkut


ysis, (pp. 1-38). Chicago: University of Chicago Press. Dulupçu, M. A. (2005). Regionalization for Turkey an illusion or a cure?, European Urban and Regional Studies, 12 (2), 99-115. EC. (2014). Procurement and grants for european union external actions, a practical guide. Retrieved 20.04.2014, from http://ec.europa.eu/europeaid/ prag/document.do. Ekelund, H. (2014). The establishment of FRONTEX: A new institutionalist approach, Journal of European Integration, 36 (2), 99-116. Gkintidis, D. (2013). Rephrasing nationalism: elite representations of Greek–Turkish relations in a Greek border region, Southeast European and Black Sea Studies, 13 (3), 455-468. Hall, P. A. and Taylor, R. C. R. (1996). Political science and the three new institutionalisms. Political Studies, XLIV, 936–957. Hilvert C. and Swindell D. (2013). Collaborative service delivery: What every local government manager should know, State and Local Government Review, 45 (4), 240- 254. Houtum, H. van and Pijpers, R. (2007), The European Union as a gated community: The two-faced border and immigration regime of the EU, Antipode, 39 (2), 291-309 Immergut, E. M. (1998). The theoretical core of the new institutionalism, Politics and Society, 26 (1), 5-34. Jessop, B. (2003). The political economy of scale and the construction of cross border micro regions. In Söderbaum, F. and Shaw, T. M. (Eds.), Theories of new regionalism: A Palgrave reader, (pp. 179- 196), Hampshire and New York: Palgrave Macmillan Johnson, C., Jones, R., Paasi, A., Amoore, L., Mountz, A., Salter, M., and Rumford, C. (2011). Interventions on rethinking “the border” in border studies, Political Geography, 30 (2), 61-69. Karakasidou, A. N. (1997). Fields of wheat, hills of blood: passages to nationhood in Greek Macedonia, 1870-1990. Chicago: University of Chicago Press. Lowndes, V. and Roberts, M. (2013). Why institutions matter: The new institutionalism in political science. Hampshire, UK: Palgrave Macmillan Manos, I. (2005). Border crossings:

dance performance and identity border region in Northern Greece. In Wilson, T. M. and Donnan, H (Eds.), Culture and power at the edges of the state (pp. 127- 154). Münster, Germany: Lit Verlag. March, J. G. and Olsen, J. P. (1984). The new institutionalism : Organizational factors in political life. The American Political Science Review, 78 (3), 734–749. Mazower, M. (2001). The Balkans. From the end of Byzantium to the present day, London: Phoenix Press. Niebuhr, A. (2006). Spatial effects of European integration Do border regions benefit above average?, The Review of Regional Studies, 36 (3) 254278. North, D.C. (1990). Institutions, institutional change and economic performance. Cambridge: Cambridge University Press. O’Dowd, L. (2003). The changing significance of european borders. In Anderson, J., O’Dowd, L. and Wilson T.M. (Eds.), New borders for a changing Europe (pp. 13-37). London, Portland, OR, US: Frank Cass Publishing. Ostrom, E. (2011). Background on the institutional analysis and development framework, The Policy Studies Journal, 39 (1), 7-27. Ostrom, E. (2005). Understanding institutional diversity, Princeton, US: Princeton University Press. Paasi, A. (1996). Territories, boundaries and consciousness: The changing geographies of the Finnish- Russian border, West Sussex, England: John Wiley and Sons Ltd Perkmann, M. (1999). Building governance institutions across european borders, Regional Studies, 33(7), 657667. Putnam, R.D. (1993). Making democracy work, civic traditions in modern Italy. Princeton, New Jersey: Princeton University Press. Ricq, C. (2006). Handbook of transfrontier co-operation, Council of Europe, Retrieved 01.08.2012, from http://www.coe.int/t/e/legal_affairs / local_and_regional_democracy/documentation/library/transfrontier_cooperation/tfc_handbookTC2006_EN.pdf Rodriguez- Pose, A. (2013). Do institutions matter for regional develop-

Cross border cooperation in Edirne-Kırklareli border region: New institutionalist perspectives


ment?, Regional Studies, 47 (7), 10341047. Sezgin, E. and Erkut, G. (2014). Avrupa birliği üyelik sürecinde türkiye’de sınır bölgeleri ve sınırötesi işbirliği: Edirne- Kırklareli örneği, (Border regions and cross border cooperation in turkey in the process of european membership: The case of Edirne- Kırklareli), Megaron, 9 (3), 228240 Smith, D. E. (2008). Institutional ethnography. In Given, L. M. (Ed.), The Sage encyclopedia of qualitative re-

search methods (pp. 433- 436). London, UK: Sage Publications. Swyngedouw, E. (1997). Neither global nor local, Glocalization and the politics of scale. In Cox, K. R. (Ed.), Spaces of globalization, reasserting the power of the local (pp. 137- 166). New York: The Guilford Press. Van der Veen, A. and Boot, D.J. (1995). Crossborder cooperation and european regional policy. In Eskelinen H. and Snickars F. (Eds.), Competitive european peripheries (pp. 75- 94), Berlin, Germany: Springer Publishing.

Edirne-Kırklareli sınır bölgesinde sınır ötesi işbirliği: Yeni kurumsalcı yaklaşımlar Avrupa Birliği (AB) Katılım Öncesi Mali Yardım Aracı kapsamında Türkiye ve Bulgaristan arasında 15 yılı aşkın bir süredir sınır ötesi işbirlikleri gerçekleştirilmektedir. Sınır ötesi işbirliği programının uygulandığı EdirneKırklareli sınır bölgesinde, il özel idareleri ve kaymakamlıklar gibi merkezi devlet kurumları, yerel yönetimler, iş örgütleri ve sivil toplum kuruluşları gibi toplumun farklı kesimlerinden aktörler proje faydalanıcısı olarak sürece dahil olmuş; AB Bakanlığı ve Kalkınma Bakanlığı başta olmak üzere farklı ulusal kuruluşlar ve AB’nin ilgili kurumları ile birlikte çok katmanlı bir yönetişim yapısı oluşturmuşlardır. Bu süreç, bölgede, sınır ötesi işbirliği programı etrafında farklı kuruluşların birbirileri ile etkileşim içinde dahil olduğu, kendi mevzuatı ve gündelik pratikleri olan bir kurumsal yapının oluşmasına yol açmıştır. Sınır ötesi işbirliğinin kurumsallaşması süreci, Edirne- Kırklareli sınır bölgesine özgü yerel dinamikler ile birlikte; ulusal ve ulus üstü düzeyde meydana gelen sosyal- politik dönüşümlerin ve tarihsel süreçlerin etkisi altında gelişmektedir. Söz konusu yerel dinamikler, güven ilişkilerine dayalı olarak yürütülen işbirliği projeleri; iki toplum arasında buna bağlı olarak geliştirilen ilişkileri; ve sınır ötesi işbirliklerine dahil olan çeşitli aktörler arasındaki güç ilişkileri olarak ortaya çıkmaktadır. Ulusal ölçekte meydana gelen dinamikler Türkiye’nin 2005’te AB üyeliği için tam

adaylık statüsü elde etmesi ile taçlanan batılılaşma süreci ve buna bağlı olarak hem bölgesel politikalar hem de kamu yönetimi alanlarında yaşanan kurumsal dönüşüm süreçlerini kapsamaktadır. AB ölçeğinden bakıldığında ise birliğin 29 üye ülkeyi içerecek şekilde genişlemesi ve bu genişlemeye bağlı olarak uyum ve bütünleşmeyi sağlayacak politikaların geliştirilmesi gerekliliği en önemli etkenler olarak ortaya çıkmaktadır. Bu çerçevede ulus üstü, ulusal ve yerel/ bölgesel ölçeklerin dahil olduğu yönetişim modellerinin geliştirilmesi olanaklı olmuştur. AB’de sınır ötesi işbirliği de bu bağlamda ortaya çıkmaktadır. Yerel, ulusal ve küresel dinamiklerin etkisi ile şekillenen sınır ötesi işbirliğinin kurumsallaşmasını araştırmak bahsedilen bu farklı ölçek ve süreçlerin belirli bir yerleşim alanı üzerine olan etkilerini anlamak açısından önem teşkil etmektedir. Bu çalışma söz konusu araştırmayı yeni kurumsalcılık kuramını ve onun farklı yaklaşımları olan üç ayrı kolunu, yani rasyonel seçim, sosyolojik ve tarihsel kurumsalcılık yaklaşımlarını kullanarak yapmayı hedeflemektedir. Yeni kurumsalcılığın bu üç ayrı kolu, toplumsal kurumlar konusunda birbirinden ayrışan ve zaman zaman çelişen yaklaşımlara sahiptir. Rasyonel seçim kurumsalcıları insanın, çıkarlarını maksimize etmek üzere hesaplar yapan ve bunun sonucunda hareket eden bir varlık olduğu kabulü ile hareket ederler. Bu bakış açısı ile toplumsal kurumlar, insanlar arasında güven eksikliğinden doğan maliyetleri minimize etmek üzere ku-

ITU A|Z • Vol 12 No 3 • November 2015 • E. Sezgin, G. Erkut


rulmuştur ve bunu başarabildikleri ölçüde toplumda yerleşirler. Tarihsel ve sosyolojik kurumsalcı yaklaşımlar bu bakış açısına farklı yönlerden itiraz etmektedirler. Tarihsel kurumsalcılara göre toplumsal kurumlar uzun tarihsel süreçler boyunca şekillenen toplumsal dinamiklerinin doğal birer sonucu olarak oluşurlar. Örneğin kapitalizm bu tür bir tarihsel- toplumsal dinamiktir. Kapitalizm tarihsel gelişim süreci içinde hegemonik bir güç haline gelmiş ve kendi kurumsal yapısını oluşturmuştur. Tüm toplumsal kurumlar da yol bağımlılığı kavramı ile ifade edilen bir zorunluluk ilişkisi içerisinde kapitalizmin kurumsal yapısının bir gerekliliği olarak ortaya çıkarlar ya da işlevlerini yitirirler. Sosyolojik kurumsalcılık ise tarihsel süreçlerden ziyade bir toplumsal yapıda mevcut güç ilişkileri ve o yapıya özgü dinamiklerin toplumsal kurumların şekillenmesinde etkili olduğunu savunmaktadır. Hem toplumsal kurumların varlığı hem de onların yönetim ve işleyişi bu güç ilişkilerinin birer yansımasıdır. Bu çalışma, Edirne- Kırklareli sınır bölgesinde gerçekleşen sınır ötesi işbirliklerini ve bu doğrultuda oluşan kurumsal yapıyı, yeni kurumsalcılık kuramının bu üç farklı yaklaşımı açısından değerlendirmektedir. Bu çerçevede, tarihsel kurumsalcılık bakış açısı ile Türkiye’nin AB üyelik süreci, soğuk savaş dönemi boyunca süren sınır bölgeleri politikalarını dönüştüren

bir kritik kavşak, fakat aynı zamanda da Cumhuriyet’in kuruluşundan beri hedeflenen batılılaşmanın devamını imleyecek bir yol bağımlılığı süreci olarak değerlendirilmiştir. Sosyolojik kurumsalcı bakış açısı ile AB bölgesel politikalarının bir parçası olan sınır ötesi işbirliği, ulus üstü, ulusal ve yerel ölçeklerden farklı çıkarlara sahip olan aktörler arasındaki güç ilişkilerinin bir yansıması olarak değerlendirilmiştir. Sınır ötesi işbirliklerinin yasal çerçevesi ve proje başvuruları sırasında uygulanan eleme yöntemleri de aynı bağlamda ele alınmıştır. Sınır ötesi işbirliklerine katılan aktörlerin katılma sebepleri ve ortak seçme süreçleri ise rasyonel seçim kurumsalcı bakış açısı ile ve çıkar maksimizasyonu bakış açısı ile değerlendirilmiştir. Çalışma, bu farklı kuramsal bakış açılarının birbirilerini dışlayan ya da birbirilerine zıt yaklaşımlar olmadıklarını, aksine aynı olguya farklı açılardan yaklaştıklarını iddia etmektedir. Bu nedenle bu farklı bakış açılarının sentezi, Edirne- Kırklareli bölgesindeki sınır ötesi işbirliğinin kurumsallaşmasını kavramak açısından önem taşımaktadır. Çalışma bölgede sınır ötesi işbirliği sürecine dahil olan aktörlerden 49’u ile, sınır ötesi işbirliğinin kurumsallaşma sürecini incelemek amacı ile 2013 yılında gerçekleştirilen bir saha araştırmasına dayanmaktadır.

Cross border cooperation in Edirne-Kırklareli border region: New institutionalist perspectives

ITU A|Z • Vol 12 No 3 • November 2015 • 285-306

Modelling road traffic noise annoyance by listening tests

Mine AŞCIGİL DİNCER1, Sevtap YILMAZ2 mineascigil@gmail.com • Department of Architecture, Faculty of Architecture, Istanbul Technical University, Istanbul, Turkey 2 demirkale@itu.edu.tr • Department of Architecture, Faculty of Architecture, Istanbul Technical University, Istanbul, Turkey


Received: July 2015

Final Acceptance: October 2015

Abstract Noise annoyance studies evaluate people’s responses to noise exposure, questioning how much they are annoyed by a certain type of environmental noise. In accordance with Environmental Noise Directive (2002/49/EC) by the EU Parliament and Council, noise annoyance dose-effect relations are determined by noise maps and questionnaires with respondents living in a certain area. The aim of this study is to build a noise annoyance model using listening tests, by examining factors which effect road traffic noise annoyance levels. In this study, listening tests are prepared using sound clips of traffic noises which are listened to in laboratory conditions. Road traffic noises are recorded for each vehicle type, taking into account possible vehicle speeds, traffic flow types, road slopes and road surfaces. Sound clips are formed according to road types and filtered to simulate sound propagation in various city conditions. Sound clips are then filtered with façade sound insulation values to simulate the sounds heard inside houses. Respondents are asked how much they are annoyed when they listen to the sound clips with headphones and imagine they are resting inside their houses. The results are analyzed and responses are investigated to form a road traffic noise annoyance model. This model provides the opportunity to transform raw data (traffic, road and settlement) directly into annoyance. The information on the effects of traffic elements, road properties and settlement types on noise annoyance can easily be used for planning new areas or noise action plans.

Keywords Listening test, Noise annoyance, Road traffic noise.


1. Introduction Environmental noise is unwanted or harmful sound, usually generated by activities such as road traffic, railways, air transport, industry, recreation and construction. People are exposed to environmental noise at various places including their homes, schools or workplaces (Kang, 2007). The potential health effects of environmental noise include ear discomfort, speech interference, aural pain, sleep disturbance, startle and defense reactions, hearing impairment, cardiovascular effects, performance reduction, and annoyance responses (WHO, 2000). Environmental noise annoyance and sleep disturbance effects are taken seriously by the European Union. The main objective of “Assessment and Management of Environmental Noise (2002/49/EC)” Directive (EU Parliament and Council, 2002) is to define a common approach intended to avoid, prevent or reduce the harmful effects, including annoyance, due to exposure to environmental noise. Turkey adapted this Directive (T.C. Çevre ve Orman Bakanlığı, 2010) with the same purposes and is working on implementing it. The term, ‘annoyance’ is defined in the Directive as ‘the degree of community noise annoyance as determined by means of field surveys’. The Directive states that dose-effect relations, that is the relation between annoyance and a noise indicator, should be used to assess the effect of noise on population. Noise indicator for annoyance given in the Directive is Lden, day-evening-night level in decibels. This indicator may be used to assess annoyance for road, rail and air traffic noise, and for industrial noise (EU Parliament and Council, 2002). European Commission Working Groups published dose-effect relations for transportation noise, created from socio-acoustic surveys, made in countries of North Europe, North America and Australia (WG-HSEA, 2002). These relationships do not necessarily apply to other countries. Some dose effect relation studies conclude that, social, psychological or economic factors, are far more important than acoustic or physical factors (Guski, 1997) (Job, 1988). Numerous studies show that the

indicators used, such as A-weighted values or Lden and Lnight, do not reflect many aspects of annoyance (Phan et al., 2009) (Persson Waye & Rylander, 2001) (Kang, 2007). This study is part of a research for composing an approach for developing road traffic noise annoyance prediction model. In the research, noise mapping and socio-acoustic surveying and listening test techniques are used to develop and validate the prediction model for an urban area. In a previous study by the authors of this article, noise maps and socio-acoustic surveys were used to form dose-effect relations for road traffic noise for Besiktas district in Istanbul, Turkey (Badino et al., 2012). Dose-effect relations in Besiktas district proved to be different from relations recommended by European Commission (WG-HSEA, 2002). This divergence could be caused by differences in non-acoustical factors, by differences in characteristics of road vehicles or of built environment, or by inadequacy of noise indicators. In

Start sound recordings

Standards on traffic noise

Test track options s

Determine recording conditions

Statistical data on traffic

Traffic data from noise map

Traffic data from on-site research

Determine types of vehicles and driving conditions

Data from noise prediction model

Traffic sound recordings data

Conduct traffic noise recordings on site No Enough samples?

Yes Sound recordings

End sound recordings



Figure 1.a. Methodology flowchart, Part 1: sound recordings.

ITU A|Z • Vol 12 No 3 • November 2015 • M. Aşcıgil Dincer, S. Yılmaz


this study, with the aim of analyzing the issue further, road traffic noise annoyance listening tests are designed for the same district. People living in the same district listen to traffic sounds at levels which may be heard inside their homes and rate the annoyance they experience. The results are analyzed to understand the factors effecting annoyance levels and a road traffic noise annoyance model is formed using analyzed results. Listening tests are used to evaluate people’s responses to noise in a controlled environment, such as a laboratory. Listening tests may be used for evaluating urban soundscapes or environmental noises such as transportation noise. Rychtarikova and Vermeir (2013) assessed soundscapes by listening tests using binaurally recorded sound in urban public places. Viollon et al. (2002) assessed how listener’s judgments of a set of urban sound enA

Start sound clips

Statistical data on traffic

Traffic data from noise map

Determine road type characteristics

Traffic data from on-site research

Data from Good Practice Guide

Form sound ound clips for road types


Create geometric divergence and atmospheric absorption filters Urban settlement data

Sound clips for road types

GD&AA filter

Determine urban sound propagation characteristics

Sound prop. data

Create urban sound soun prop. filterss

Data on façade elements

Create sound nd insulation ins filter On-site & laboratory sound insulation measurement

Urban prop. filter

Insulation filter

Form sound clips

End sound clips

Sound clips for listening tests



Figure 1.b. Methodology flowchart, Part 2: sound clips.

vironments were affected by visual settings. Trolle et al. (2008), analyzed the auditory perception of environmental noises transmitted through a simulated window via listening tests. Yifan et al. (2008) experimented on annoyance ratings of noise samples with different frequency spectrums but same A-weighted levels. Barbot et al. (2008) investigated acoustic features of aircraft noise which could be improved by aircraft manufacturers from a sound design point of view. Lavandier et al. (2011) used aircraft flyover sounds to rate the level of activity disturbance due to the noise environment when carrying out memory and concentration tasks. There are some listening test studies on certain properties of traffic noise annoyance. Freitas et al. (2012) executed listening tests for road traffic noise, using different road surfaces, car speeds and traffic densities. Trolle et al. (2015) investigated sound unpleasantness due to urban road traffic at crossroads by a listening test and discovered that type of crossroad, traffic lights and heavy vehicle content effect annoyance. Nilsson (2007) executed listening tests on road traffic noise with strong low frequency content and found them to be more annoying. Paviotti & Vogiatzis (2012) investigated pedestrian annoyance from scooter and motorbike noise and found masking effect by general traffic to be effective. Torija & Flindell (2014) examined low height roadside barrier’s effects on annoyance by listening tests. Sandrock et al. (2008) executed listening tests on acute annoyance due to trams, buses and trucks, finding task performance and single pass-by versus realistic traffic flow to be effective in annoyance levels. In this study, listening tests are conducted using sound clips of road traffic noises which are listened to in laboratory conditions. Road traffic noises are recorded for each vehicle type, taking into account possible vehicle speeds, traffic flow types, road slopes and road surfaces. Sound clips are formed according to road types and filtered to simulate sound propagation in various city conditions. Sound clips are then filtered with façade sound insulation values to simulate the traffic sounds

Modelling road traffic noise annoyance by listening tests


heard inside houses. Questionnaire respondents are asked to listen to the sound clips with headphones and imagine they are resting inside their houses. The results are analyzed and responses are investigated to form a noise annoyance model. This model helps to understand the dynamics of noise annoyance. This study focusses on the modelling of road traffic noise annoyance by listening tests. Future studies will be on validating this model, using socio-acoustic survey results to transform it into a reliable prediction model. Environmental noise annoyance studies in EU require the main steps of acquiring data (traffic, road and settlement), forming noise maps via noise prediction models, and executing socio-acoustic surveys in order to establish dose-effect relations. Using an environmental noise annoyance prediction model created by listening tests provides the opportunity of directly predicting noise annoyance from acquired data. This prediction model would exclude noise indicators which have questionable reliability and which do not reflect many aspects of annoyance, determined by works of Phan et al. (2009), Persson Waye & Rylander (2001) and Kang (2007). Because the main purpose of environmental noise control studies is to reduce harmful effects such as annoyance, a direct relation between on-site data and annoyance is valuable. These models can be created for different countries, for different settlements, and for different social and economic zones, taking into consideration Guski (1997) and Job’s (1988) findings on the importance of non-acoustic factors. Annoyance model created by listening tests provides information on the effects of traffic elements, road properties and settlement types on noise annoyance, which can all be used directly in planning new areas or noise action plans. 2. Methodology and theoretical background The methodology of this study brings together various methods used for sound recording, forming sound clips with sound filters and applying listening tests. Figure 1 shows the flow-


Start listening test and annoyance model

Literature on listening tests

Determine listening test conditions

Test site options

Exemplary survey questions

Form listening tests

ISO/TS 15666

Listening test


Listening test results data

Conduct listening tests

Conduct statistical analysis

No Enough samples?

Y Yes Road traffic noise annoyance model

Analyze listening test results

End listening tests and annoyance model

Figure 1.c. Methodology flowchart, Part 3: listening tests and annoyance model.

chart for the methodology in three parts, (a) sound recordings, (b) sound clips, (c) listening test and annoyance model. This chapter also explains the methodology in the same three headings. This methodology may be used for forming listening tests for different countries, for different traffic conditions or for different urban conditions, which will provide different annoyance models. The detailed explanation of the methodology and the theoretical background of the study are given in the following sub-sections. 2.1. Traffic sound recordings For traffic sound recordings, most common types of vehicles were determined by statistical information; driving conditions were determined by data from noise maps, noise prediction models and on-site research. Sound recordings were conducted in a similar methodology to traffic noise measurement standards.

ITU A|Z • Vol 12 No 3 • November 2015 • M. Aşcıgil Dincer, S. Yılmaz


Key minimum area covered with test road surface, i.e. test area microphone positions (height 1,2 m) AA test zone start BB test zone end CC line of vehicle travel through test zone PP line perpendicular to vehicle travel between microphone locations R50 radius of 50 m around the centre of the track NOTE: The shaded area (“test area”) is the minimum area to be covered with a surface complying with ISO 10844.

Figure 2. Test site dimensions (ISO 362-2, 2007).

2.1.1. Traffic sound recording standards There are no guidelines for traffic sound recordings, therefore traffic noise measurement standards were used to guide the recordings. The related standards are, ISO 362-1 (2007), ISO 362-2 (2009) and ISO 10844 (2014). ISO 362-1 (2007) and ISO 3622 (2009) standards are about measurement of noise emitted by accelerating road vehicles of various categories under typical urban traffic conditions. The specifications intend to reproduce the level of noise generated by the noise sources during normal driving in urban traffic. The test track construction and surface shall meet the requirements of ISO 10844 (2014). The test site dimensions are shown in Figure 2. Within a radius of 50 m around the center of the track, the space shall be free of large reflecting objects such as fences, rocks, bridges or buildings. The test track and the surface of the site shall be dry and free from absorbing materials (ISO 362-1, 2007). During the recordings, the geometry provided in the standards were followed. In the vicinity of the micro-

phone, there was no obstacle that could influence the acoustical field and no person remained between the microphone and the noise source. The distance from the microphone positions on the microphone line PP’ to the perpendicular reference line CC’ on the test track shall was 7.5 m ± 0.05 m. The microphone shall was located about 1.2 m above the ground level. The path of the centerline of the vehicle followed line CC’ as closely as possible throughout the entire test, from the approach to line AA’ until the rear of the vehicle passed line BB’. For accelerations and decelerations, the test speed was reached when the reference point was at line PP’ (ISO 362-2, 2007). For fluid continuous traffic flow recordings, test speed was constant from AA’ to BB’. Reference points of road vehicles are defined according to engine positions, which is mostly the front end of vehicles (ISO 362-1, 2007). The test track is a test instrument and shall be protected from damage and be taken care of. The test track should be used only for noise measurements and should be kept clear from loose debris or dust during measurements (ISO 10844, 2014).

Modelling road traffic noise annoyance by listening tests


The background noise was measured before and after recordings. The recordings were made with the same microphones and microphone locations used during the test. The background noise should at least 10 dB below the A-weighted sound pressure level produced by the vehicle under test (ISO 362-1, 2007). ISO 362 standard series recommend vehicle speed and acceleration for the measurement to be determined according to real urban traffic conditions, so that vehicle emission in urban traffic may be portrayed correctly. Inquiries among dwellers along various streets show that noise disturbance happens mainly along urban main streets, and during vehicle acceleration transients (ISO 362-1, 2007). According to ISO 362-1 (2007), the behavior of drivers depends on speed limits (traffic laws), traffic density, road arrangement (traffic lights, corners, etc.), driving purpose (commuting, pleasure, commercial, etc.), enforcement of traffic laws, and the way the vehicle behaves as an acoustical source under these conditions. Annex A of ISO 362-1 (2007) gives the technical background for development of vehicle noise test procedure based on in-use operation in urban conditions. Standard recommends vehicle speed and acceleration for the measurement to be determined according to real urban traffic conditions, so that vehicle emission in urban traffic may be portrayed correctly. 2.1.2. Determining vehicles and driving conditions Available statistical data may be used to determine the most common types of vehicles which may be used in recording vehicle sounds in traffic conditions. Driving conditions were determined by using data from noise maps, noise prediction model and onsite research. The area under consideration was noise mapped for road traffic in a previous study and average speed (km/h) data used in road modelling of noise maps was taken into consideration. As it was advised by the Directives (EU Parliament and Council, 2002) (T.C. Çevre ve Orman Bakanlığı,

2010), NMPB-Routes-96 (1995) was used in this study for traffic noise prediction modelling. In this model, given traffic flow types are fluid continuous, pulsed continuous, pulsed accelerating and pulsed decelerating. The traffic flows are categorized the same way in this study as well, for compatibility. In “Good Practice Guide for Strategic Noise Mapping and the Production of Associated Data on Noise Exposure”, (WG-AEN, 2006) the roads are classified as dead-end roads, service roads, collective roads, small main roads and main roads. The same classification is used in this study for compatibility purposes. Annex A of ISO 362-1 (2007) gives the technical background for development of vehicle noise test procedure based on in-use operation in urban conditions. In the annex, the distribution of vehicle speed in urban traffic is examined and driving behavior is recorded on actual urban routes. Speed, acceleration and gears have been statistically examined in urban driving conditions. Standard recommends vehicle speed and acceleration for the measurement to be determined according to real urban traffic conditions, so that vehicle emission in urban traffic may be portrayed correctly. An on-site study by driving through the area at different times during the day was used to reveal the driving patterns. 2.2. Sound clips The sound clips were formed for the purpose of helping to develop a road traffic noise model. The sound clips each simulated a traffic noise situation possible to hear inside houses in the area under consideration. First, road types and characteristics were determined to create the traffic noise heard 7.5 meters from road sources (ISO 10844, 2014). Then, sound propagation characteristics for the urban area were investigated and used for creating and applying sound propagation filters to sound clips. To simulate the traffic noise heard inside the houses, sound insulation values were applied as sound filters. All of these steps finally created the sound clips to use in the listening tests. The length of sound clips in listen-

ITU A|Z • Vol 12 No 3 • November 2015 • M. Aşcıgil Dincer, S. Yılmaz


ing tests do not have a standard. Parizet et al. (2002) used binaural sound recordings of 10 seconds in various positions in a high-speed train as stimuli in listening tests. Viollon et al. (2002) assessed how listener’s judgments of a set of urban sound environments were affected by visual settings. Each of the various sounds, which were road traffic noise, bird song and sounds due to human presence, lasted 20 seconds. Jeon et al. (2007) worked with various types of refrigerator noise in an anechoic chamber and in a real living environment. A total of 40 noise sources with duration of 5 seconds were presented randomly. Barbot et al. (2008) investigated acoustic features of aircraft noise which could be improved by aircraft manufacturers from a sound design point of view. The duration of all the stimuli were 40 seconds. Sandrock et al. (2008) executed listening tests on acute annoyance due to trams, buses and trucks, with stimuli 6 seconds long. Trolle et al. (2008), analysed the auditory perception of environmental noises transmitted through a simulated window via listening tests. The duration of each generated stimulus was 4 seconds. Yifan et al. (2008) experimented on annoyance ratings of 5 seconds long noise samples with different frequency spectrums but same A-weighted levels. Sound clips were formed with a duration of 20 seconds for this study. The number of vehicles needed for each type of road were distributed as evenly as possible on a 20 seconds long empty sound clip, on the software Audacity.

were also used. The on-site research recommended by Annex A of ISO 3621 (2007) was used to validate traffic conditions in secondary roads. After using all of this data, traffic flow for all roads in the area was determined and grouped. For the use of this data in sound clips, road traffic volumes were adjusted 20 seconds.

2.2.1. Determining road type characteristics Road types in the area were determined for this study. Characteristics which influence traffic noise emission are, traffic volume, types of vehicles, traffic speed, traffic flow type and road surface. These had been determined in detail for major roads in the noise map model prepared for this area. European Commission’s Good Practice Guide for Strategic Noise Mapping (WG-AEN, 2006) proposes some default values for traffic flow volume, these values were adapted to the area under consideration. Statistics of road motor vehicles Urban sound propagation Urban sound propagation research in literature is based on experimental or theoretical works and examples on street canyon research clearly show this. Picaut et al. (2005) experimented on sound propagation in a street canyon, with various source and receiver locations on a street in Nantes, France. Nicol & Wilson (2004) investigated the effect of street dimensions and traffic density on noise levels in urban canyons, by noise measurements in Athens. Walerian et al. (2001) used a simulation program to calculate sound level distribution and ΔL on a canyon

2.2.2. Sound propagation filters The sound clips formed represent different types of roads and traffic flow characteristics recorded at 7.5 meters from road sources, in open space conditions. Some common examples of urban sound propagation are calculated and applied as filters to sound clips at hand, in order to simulate traffic sounds in the city. Filters for geometric divergence and atmospheric absorption were created from literature. Filters for urban condition examples were calculated with noise mapping software. Geometric divergence and atmospheric absorption Filters for geometric divergence and atmospheric absorption have been created from literature. Geometric divergence for line sources is attenuation of 3 dB for doubling of distance. Because the sound recordings were conducted 7.5 m away from source, geometric divergence filter values for double distances such as 15 m and 30 m are used. The same principle is applied for atmospheric absorption using sound absorption values from ISO 9613-1 (1993).

Modelling road traffic noise annoyance by listening tests


street model. Schiff et al. (2010) executed a numerical investigation of sound propagation over multiple street canyons. Experimental research results are bounded by the on-site conditions and dimensions. Results of theoretical studies usually do not provide noise levels on façades and cannot reflect all the different settlement types in the area. Therefore, environmental noise prediction model recommended by the Directive, NMPB-Routes-96 (1995), was used to assess all types of sound propagation in these settlement types. Various noise propagation conditions were simulated in SoundPlan 6.5 Noise Mapping Software, both in open space conditions and in urban conditions. The difference between the two conditions were used to create sound propagation filters, which were used on sound clips, in order to simulate traffic sounds in city conditions. 2.2.3. Sound insulation filters Environmental noise annoyance focuses on environmental noise perceived inside houses. In order to simulate this effect, the sound clips were filtered by façade sound insulation values. Façade elements were determined by one of the on-site survey questions in the area, observation of façades in the area and statistical data on main wall elements. Façade sound insulation to be used for filtering was determined by sound insulation measurements on-site (ISO 140-5, 1998). To validate these on-site measurements, building element laboratory measurements (ISO 10140-2, 2010) were used to calculate sound insulation of composite façade, using Equation 1 (Barron, 2003). Where, TLfaçade: Sound transmission loss of composite façade, dB; TLwall: Sound transmission loss of wall, dB; TLwindow: Sound transmission loss of window, dB; Afaçade: Area of composite façade, Awall + Awindow, m2; Awall: Area of wall, m2; Awindow: Area of window, m2.





2.3. Listening tests and annoyance model For the listening tests, questions and sound clips were prepared, tests were conducted in laboratory conditions and results were analyzed. 20 seconds long sound clips were created to simulate the sound heard inside houses and to evaluate environmental noise annoyance. Different sound clips were created the road types, for compatible speeds, road slopes, surfaces, traffic flow types, and source-receiver distances. Effects of sound propagation in urban conditions were simulated for compatible road types. The effects of time of day, window condition and daily activity were also taken into consideration. 2.3.1. Questionnaire forms Listening test questions were prepared in parts. Pre-criteria questions determined if the participant is competent to attend the survey. The first part of the listening tests included the same questions as the on-site socio-acoustic survey conducted in the area. The second part of the survey inquired into the annoyance of sound clips. Pre-criteria for conducting the surveys were; minimum 12 months of residency in Besiktas District, lack of any hearing problems and being in the age range of 18 to 65. In Part 1 of the listening test, personal information and environmental noise annoyance were questioned. Under the heading of ‘personal information’, gender, age, education level, duration of residence, time and period spend at home during day, noise sensitivity and noise annoyance at workplace were investigated. Under the heading of ‘noise annoyance’, traffic noise annoyance at home, for all day and only night periods were investigated in verbal and numerical scales. Wording of these questions and verbal and numerical scales were given in ISO/TS 15666 (2003). Also under the same heading, most annoying traffic elements and annoyance during dai-

ITU A|Z • Vol 12 No 3 • November 2015 • M. Aşcıgil Dincer, S. Yılmaz



Table 1. Part 2 questions in the listening test. Part 2, Sub-part 1 and Sub-part 2 XX) Imagining you are resting at home, how much does the sound clip you listened to, bother, disturb or annoy you?  Not at all?  Slightly?  Moderately?  Very?  Extremely? XX) Imagining you are resting at home, what number from 0 to 10 best shows how much you are bothered, disturbed or annoyed by the sound clip you listened to? Not at all


0 1 2 3 4 5 6 7 Part 2, Sub-part 3 (the question asks for only one time frame)




XX) Imagining you are resting at home, during day time (07-19) / evening time (19-23) / night time (23-07) , how much does the sound clip you listened to, bother, disturb or annoy you?  Not at all?  Slightly?  Moderately?  Very?  Extremely? XX) Imagining you are resting at home, during day time (07-19) / evening time (19-23) / night time (23-07) , what number from 0 to 10 best shows how much you are bothered, disturbed or annoyed by the sound clip you listened to? Not at all


0 1 2 Part 2, Sub-part 4









XX) Imagining you are reading at home, how much does the sound clip you listened to, bother, disturb or annoy you?  Not at all?  Slightly?  Moderately?  Very?  Extremely? XX) Imagining you are reading at home, what number from 0 to 10 best shows how much you are bothered, disturbed or annoyed by the sound clip you listened to? Not at all 0

Extremely 1





ly activities are inquired using multiple-answer questions. Room positions in regards to main road, open windows during night and main wall elements are also questioned. Part 2 of the listening test inquired about how much the sound clips bother, disturb or annoy the participants in






verbal and numerical scales (Table 1). Wording of these questions were similar to questions given in ISO/TS 15666 (2003). Verbal and numerical scales were the same as scales used in Part 1 and ISO/TS 15666 (2003). Six different tests were created to change the order of the sound clips in each test.

Modelling road traffic noise annoyance by listening tests


Part 2 was divided into four subparts to provide breaks if necessary. In sub-parts 1 and 2, wording of the questions did not change. The question was; “Imagining you are resting at home, how much does the sound clip you listened to, bother, disturb or annoy you?”. In sub-part 3, a time frame was given in each question, such as day time (07-19), evening time (19-23) or night time (23-07). In sub-part 4, the activity changed from resting to reading. A short magazine article about travelling to Mars was read by participants. The article was divided in two parts, first part was read in quiet, while the second part was read with exposure to traffic noise. 2.3.2. Executing listening tests Pilot listening tests were executed with 4 people, to identify the possible problems. Some explanatory phrases and warnings were added to the listening test as a result of this pilot study. The listening tests were conducted as face-to-face interviews with 40 people between the ages of 18 and 65, who live in the related district. The listening tests were executed in laboratory, where background noise was always monitored. Headphones (MESA BMH.I-H42 binaural headset) were used to listen to sound clips. All participants signed a consent form and they were warned to stop the test if they felt any auditory problem. The investigator asked the questions, turned on the sound clips and typed the answers of the participants on a MS Excel worksheet; so that the participants could concentrate on the sound clips. 30 second breaks were given between each sound clip to ensure concentration and a fresh perception. Participants were free to express any opinions they had about the sound clips and the listening test. 2.3.3. Analyzing listening test and building model Listening test results were statistically analyzed; Cronbach’s alpha was computed for reliability and Spearman’s correlation coefficient was calculated for factors affecting annoyance. Verbal and numerical scales were used for sound clip annoyance ques-

tions. These different scales were converted and analyzed on a 100 scale. On the verbal scale, “not at all” was converted to 0, “slightly” to 25, “moderately” to 50, “very” to 75 and “extremely” to 100. On the numerical scale, 0 was 0, 1 was converted to 10, 2 to 20 and so on. For analyzing percentage of people annoyed (%A) and percentage of people highly annoyed (%HA) the cutoff points on a 100 scale are 50 for %A and 72 for %HA (WG-HSEA, 2002). On the verbal scale, cutoff point of 50 for %A referred to points 50, 75 and 100, which were “moderately”, “very” and “extremely” respectively. Cutoff point of 72 for %HA referred to points 75 and 100, which were “very” and “extremely” respectively. %A was associated with the total number of responses for 5, 6, 7, 8, 9 and 10 (from 50 to 100 on the 100 scale) on the numerical scale, whereas %HA was associated with the total number of responses for 8, 9 and 10 (from 80 to 100 on the 100 scale) on the numerical scale. Annoyance levels for each simulated traffic sound clip was examined for number of people annoyed and highly annoyed within the whole group of respondents, in order to calculate percentage of people annoyed (%A) and percentage of people highly annoyed (%HA). Averages of verbal and numerical scale results were used. %A and %HA levels for each sound clip were then compared to others with similar properties. For easy expression and comprehension, some factors which effect annoyance in a similar way were united. 3. Application of traffic sound recordings The road traffic sound recordings were made using the most common vehicles in Istanbul city and the possible driving behaviors in the area under consideration. 3.1. Determining most common vehicles Turkish Statistical Institute is responsible for collecting and disseminating the data which display the social and economic structure of Turkey. The publication “Road Motor Vehicle Statistics 2012”, includes statistics of the

ITU A|Z • Vol 12 No 3 • November 2015 • M. Aşcıgil Dincer, S. Yılmaz


road motor vehicles such as, the current number of the vehicles according to their types, trademarks, fuel type, and model years by the end of year 2012, for each city (TurkStat, 2013b). According to Istanbul city statistics, in all motor vehicles, 52% of vehicles have diesel fueled motors and 46% of vehicles have gasoline fueled motors. Renault is the most common trademark in terms of cars, small trucks, trucks, buses and minibuses. Most common three trademarks and their percentages in the market are; Renault 15.4%, Ford 9.9% and Fiat 9.6%. Honda is the most common trademark for motorcycles with 25.2% market share. Most common engine size of cars is 1600 cc with 38.3%. The most common public transportation bus is Otokar Kent 290 LF with 29.4%. The most common trademarks were used for sound recordings. The cars used for sound recordings were diesel fueled Renault and gasoline fueled Ford with engine size 1600 cc. These cars were also used to record horn sounds. Other vehicles used were, Honda motorcycle, Otokar Kent public transportation bus, Iveco minibus (blue minibus common in Besiktas area) and Renault Midlum Truck. 3.2. Determining most common driving conditions Driving conditions were determined by using data from noise maps, noise prediction model and on-site research. Average speed (km/h) data used in road modelling of noise maps in Besiktas district (Badino et al., 2012) was taken into consideration. The average speed for Barbaros Avenue in north direction received from radars was between 55 and 70 km/h for day, 60 and 65 km/h for evening, 75 and 80 km/h for night. The average speed for Barbaros Avenue in south direction received from radars was between 50 and 80 km/h for day, 50 and 85 km/h for evening, 65 and 95 km/h for night. The average speed for small main roads and collecting roads received from radars was between 40 and 50 km/h. The average speed for service roads and deadend roads determined on-site were between 30 and 40 km/h. Traffic flow types of fluid continu-

ous, pulsed continuous, pulsed accelerating and pulsed decelerating were used as advised in NMPB-Routes-96 (1995). the roads are classified as dead-end roads, service roads, collective roads, small main roads and main roads as advised in Good Practice Guide (WGAEN, 2006). As it was advised in Annex A of ISO 362-1 (2007), an on-site study by driving through the area (Besiktas) at different times during the day was used to reveal the driving patterns. On Barbaros Avenue, traffic flow was mostly fluid continuous during daytime and nighttime, it was mostly pulsed continuous during evening. Pulsed accelerating and pulsed decelerating traffic flows were existent due to traffic lights. For fluid continuous traffic flow, speed during daytime ranged from 50 to 80 km/h, while speed during nighttime ranged from 70 to 100 km/h. Traffic flow during evening hours was pulsed continuous, mostly stopping and starting in traffic. Average speed of heavy vehicles were between 30 km/h and 50 km/h. For roads other than Barbaros, the average speed values from on-site study were consistent with data from noise map models. The traffic flow was fluid continuous for service roads and dead-end roads at all times. For small main roads and collective roads, traffic flow was mostly fluid continuous during daytime and nighttime, it was mostly pulsed continuous during evening. Pulsed accelerating and pulsed decelerating traffic flows were existent due to traffic lights and junctions. Slope of roads were categorized as, h