A Journal of Undergraduate Philosophy Volume IV
Contents ROLE-PLAYING GAMES AND THE ETHICS OF CARE ................................................... 1
Laura M. Bernhardt; Knox College ASCETICISM AND THE VALUE OF TRUTH IN THE GENEALOGY OF MORALS ........................................................................................... 17
Julie Anne Buchsbaum; Beloit College Two ApPROACHES TO ARTIFICIAL INTELLIGENCE:
AN ANALYSIS OF HEIDEGGERIAN AND DREYFUSIAN CRITIQUES .................... 29
Nicholas K. Gracilla; Denison University How NOT TO READ RORTY .................................................................................................. 43
LeBoeuf; University ofMassachusetts, Amherst
NAGEL, PHYSICALISM AND SUBJECTIVITY ..................................................................... 59
Brent Little; TranS1Jlvania University PROPER FUNCTION, RELIABILITY AND WARRANT .................................................... 73
Jack C. Lyons; Valparaiso University
The editors express sincere appreciation to the Denison University Research Foundation, the Denison Office of Admissions, the Denison Honors Program, Pat Davis and Faculty Advisor David Goldblatt for their assistance in making the publication of this journal possible. We also extend special gratitude to the Philosophy Department faculty: Eric Barnes, David Goldblatt, Anna Greco, Tony Lisska, Ronald E. Santoni and Steven Vogel for their constant enthusiasm, support and creative input.
1 ROLE-PLAYING GAMES AND THE ETHICS OF CARE
Laura M. Bernhardt
In dealing with moral philosophy, one productive method of analysis is the examination of the metaphorical presuppositions
behind a given system or approach to ethics, an examination of the author's basic paradigm. In the case of the "ethics of care" or "feminine" ethical approaches, the most commonmetaphoricalback ground for discussion is the "mothering" paradigm, or the model of a productive, growth-fostering positive relationship. I would like to contend that this is pOSSibly in error, in that it further supports certain stereotyped attitudes towards the "feminine" and in some senses may invalidate (or at least severely weaken) the argument for an ethic of care. I would like to propose an alternative metaphor, that of the role-playing game, primarily because role-playing involves, as Ernest Goffman writes, "a cycle of face-to-face social situations with role others, thatis,relevantaudiences" (E, p. 5). This is, inmany senses, precisely what "feminine" moral approaches are all about. I
What does it mean to apply an "ethic of care"? For a nUmbl:'f of years, in the fields of both cognitive psychology and philosophy it meant that the moral agent in question had not yet matured enough morally to use the "higher," more abstract forms of ethical reasoning (decisions based on concepts of "justice"), an assumption based on the work of Lawrence Kohlberg and Jean Piaget in designing pat terns of moral developmental stages. 1 On the Kohlberg scale, people who functioned with care for others as their primary moral guideline scored at level three (out of six, six being the highest), and most of the people scoring three or lower on this scale were women. This, to Carol Gilligan. was indicative of a serious problem in psychological Bern/wdtis ascniorp/lilosophymajol'at Knox CoIlege. Sheisa resident o/W/wa fo 11, Illinois and will be attending Tile Ulliversity a/Illinois ill tile/all to Fltl'slIea doctomtcill pllilosopl1y. 1 The stages of development, according to Kohlberg's Socratic-inspired model are: 1) Punishment und Obedience, 2) Instrumental Exchange, 3) Interpersonal Conformity, 4) Sodal System and Conscience Maintenance, 5) Prior Rights and Social Contract and 6) Universal Ethical Principles.
Episteme • Volume IV • May 1993
research, namely an attitude that created "recurrent problems of interpreting women's development" due to "repeated exclusion of women from the critical theory-building studies of psychological research" (Gilligan, p. 1). (All of Kohlberg' s test subjects were male, a common practice in psychological research). In order to counteract this error, Gilligan did a number of studies of moral attitudes and decision-makingprocesses, using subject-groups ranging from chil dren at different developmental stages to a group of women consid ering abortion. as well as a study of female college students involved ina class on "moral and political choice" (Gilligan, pp. 2-3). What Gilligan discovered, in the course of years of interviews, was a "different voice" in moral behavior, one which, while "not characterized by gender," was nonetheless most often empirically observedinwomen(Gilligan, p. 5). Supportedby NancyChodorow' s work on developmental psychology, specifically her thesis that women, due to their connection and identification with the mother, would more naturally tend towards a "relational" moral framework, Gilligan found an alternative moral standpOint to the morality of justice. It involved the following three necessary elements: Contex tual Relevance (all actions and choices mustbe considered within the context of a concrete Situation). Maintenance of Relationship (prior ity is given to action which allows for the continuation of positive relations with others) and Conception ofIdentity ("I" am a composite of myself and others as a social being and must behave as SUcll) (Gilligan, pp. 25-68). The synthesis of these three elements (contextual relevance, maintenance of relationship and conception of identity) revealed for Gilligan a picture of people who spoke with an ethical voice charac terized as a discourse detached from principle, focused on relations in particular contexts, where this discourse was considered a rna ture, intelligible, productive moral standpoint rather than a juvenile, incomplete approach. The question invariably arises, however, as to where this alternative outlook can £it in to usual moral views given the frequent references made by many of the individuals inter viewed to obligations and values that were not necessarily solely constituted as a matter of "care." While Gilligan, adhering to Chodorow, writes of mother-contact as the fundamental develop mental basis for the "caring" response, it is also productive to examine Gilligan's information in terms of other factors, and it is in
ROLE PLAYING GAMES AND nm E'IHICS OF CARE
these other comments that we can understand the place of care足 ethical directions within the conventional moral realm of description. . One such alternative hypothesis is that the ethic of care may be seen "as a set of circumscribed coping strategies for dealing with sexism" (Puka. pp. 58-82). According to Bill Puka (in attempting to deal with the feminist criticism of Gilligan that shows "caring" as "a sexist service orientation"), there are three developmental levels in Gilligan's hypothesis: self-protection, caring for others (where self足 defense is "selfish" and "irresponsible") and self-balanCing care for others (Puka, p. 59). Progress from one level to another is a matter of growing self-confidence and competence, and since these feelings are not always present, Puka claims, "women progress and regress in care, rather than following an invariant progressive sequence;" women suffer from feelings of vulnerability and impotency in a masculine world, and are thus often forced to regress to a "lower" moral level (Puka, p. 60). In this way, when women are faced with rejection, domination and other damaging behaviors, they have the option of swinging between the levels, depending upon how they feel best able to cope with these "attacks." The difficulty inherent in this, as Puka sees it, is that there will often occur a reversion to the "slave mentality" of the second level, serving only others above oneself, an example of excess in the areas of sacrifice and relationship maintenance (Puka, p. 62). Women, and iri fact anyone dependent upon a care-ethical apptoach of the sort shown by Gilligan, would find themselves constantly faced with the possibility of such a potentially harmful regression to a submissive standpOint. How are we to deal with this very apt criticism? It seems to walk hand in hand with the sentiment that Gilligan's findings, which make a definite differentiation between the justice and care stand足 points based on a noticeable correlation with sex difference, merely enforce the stereotype of the submissive, emotional, "irrational" female, even as Gilligan attempts to defuse this implication of immaturity (Gould, pp. 411-415),2 The danger exists that an ethic of care, posited as a "feminine" institution, can actually be harmful rather than helpful. Thispointmayberefutedinpartbyemphasizing Gilligan's insistence that her "different voice" is a "theme," rather than a necessary empirical characteristic, and that it exists in both 2
This is one of the better of several different articles emphasizing this point.
male and female subjects (Gilligan, p. 6). Amore definitive refutation is to be found, however, in expanding the scope of the analysis beyond sex differences to include differences in age and social p osition as well. In various similar studies of moral orientation, age and status had as much to do with the sexual differentiation as anything else. In one study, for example, it was discovered that adults who were parents showed a definite sex-role differentiation in their moral standpoints, while adults who were not parents did not (Pratt, pp. 373-391). In another study of college students, moral focus was less related to sexual differences and found to be more closely tied to a certain high level of idealistic and non-relativistic thinking (Forsyth, pp. 243-248). When these additional factors are added to Gilligan's original thesis of gender-specific ethical out足 looks, the stereotypical femininity may safely be ignored in favor of more complex circumstances; the caring orientation in answering ethical dilemmas may be seen to be influenced strongly by many factors in addition to gender. Having moved away from specifically "feminine" typecasting for the ethic of care, how then are we to deal with Puka'sproblem of "regression" in care? Gilligan herself provides the framework for identifying a balanced, intelligible ethic of care (Puka's third level) with validity as a "mature" orientation alternative to that Justice足 based ethic, but she does not identify the lower stages flS such-her concern is with the understanding of the mature stage only. just as Kohlberg's work is primarily concerned with the identification and direction of the most mature level of justice-based moral discourse. Keeping in mind the variations in moral "maturity" influenced both by age and by social-relational status (as well as idealism, relativism, etc.), one way to get around Puka's objection is to con足 sider the presence of care as a virtue, as per the teachings of virtue足 theory in ethics. To do this we must move from the descriptive realm to a more normative deSign, as found in the work of Nel Noddings. II
Given the above deSCriptive background, what shape would a normative theory ofethlcal caring take? For Nel Noddings, it would be additionally supplemented by expanded ideas of reciprocation, empathy and specific forms that relationships can take. Noddings
ROLE PLA VING GAMES AND THE Ennes OF CARE
begins her account of caring. as Gilligan did indirectly by invoking Chodorow, with the concept of mothering. It is the mother--child relationship that is the paradigmatic case for her system, with the roles of the mother, or "one--caring," and the child, "cared-for," with each role entailing special obligations toward the maintenance of a caring interrelationship. According to Noddings, the first thing required to establish a properly ethical caring relationship is knowledge of what preCisely this care entails, above and beyond the basic concerns of context, relationship and self-definition. Caring canbe many things, and she differentiates among caring for people, an "aesthetic" care for things and ideas and the ultimately undesirable form of care as a burden or onerous obligation (Nod dings, p. 9). Caring, in the aesthetic sense, is comparable to the idea of preference, or caring for one thing over another, or having an interest ina particularidea or object rather than another person. Caring as concern or burden also deals with objects, in the sense of there existing an obligation to attend to these objects in a certain way (such as "caring" for one's lawn or some other personal possession). It is a mistake, however, to substitute either of these two forms of the concept for pl'operethical caring; for Noddings, ethical care should neither involve an abstract, aesthetic preJerence for certain persons nor an attitude of onerous obligation (such as in caring for one' s elderly, in firm grandmo ther as if she were a burden) (Noddings, pp. 12-13). According to Noddings, the inclination to足 wards caring for people interactively, which is the "mothering" paradigm's primary component, is the basic "premoral" virtue of care upon which most care-ethical considerations should be based (Noddings, pp. 13-15). The ethical ideal that supports this virtue is rooted in "the natural sympathy human beings feel for each other and the longing to maintain, recapture or enhance our most caring and tender mo足 ments" (Nod dings, p. 104). The effod required, first and foremost, is the holding of a "cating attitude," where care is as per the previously stated ideal, rather than a matter of trouble (Noddings, p. 13).3 To show how this attitude should be applied, Noddings divides the caring inten'elation into the roles of "one--caring" and "cared-for." 3 The example of care as burden is twofold, either where care is equated with worry or where caring for someone becomes worry, such as is seen with Noddings' Mr. Smith, who must care for his ailing mother who is in a nursing home.
The role of the "one-caring" is primarily the keeping of this attitude in sucha way that the "cared-for" is the total of all ofthe one caring's attention in those matters in which they must relate to one another. The one-caring is obliged to do whatever is necessary to help the cared-for grow as a person and fulfill herself (Noddings, pp. 60-74). In any relation, the approach of the one-caring is one of engrossment, of an almost complete empathy with the cared-for's best interests and needs. The one-caring is perceived by the cared for as one who "accepts, embraces and leads upward," a presence of ultimate support and belief; in short, it is the role of the mother, and for Noddings, the ideal teacher (Noddings, p. 67). The caring attitude, however, is not something incumbent upon the "one-caring" alone-reciprocation is requ:ired in Noddings' conception of care. In her "mothering" example, the cared-for gives back receptivity of the one-caring's efforts, and it is the capacity to respond in this fashion that delimits where the pOSSibility for a caring relation begins and ends (Noddings, p. 86). The cared-for has, in a sense, a certain obligation to grow and flourish under the one caring's ministrations; this is the only way in which a caring relation can continue, such that the one-caring's absorption remains en gaged. The role of the cared-for as respondent is what sets caring f01' persons apart from caring for animals or material objects 01' ideas none of the latter are really capable of the full responsiveness of a human being, which allows the capability for caring as interaction rather than as concern or inclination (Noddings, pp. 148-170):1 How does one make a decision as one-caring? The root desire for the one-caring (our ethical agent) is attention to the needs (and specifically the possible pain) of the cared-for(s) involved. Then she must consider the situation. All of this is supported by Gilligan and N oddings alike. It is supplemented, however, by N oddings' under standing of care as a virtue and care as a direction based not only on fear of hurt or maintenance of connection or empathy, but also a positive absorption in the situation of the cared-for by the one caring; Gilligan shows empathy, but not absorption. This canbe seen in the way in which Noddings presents care-ethical solutions to ethical dilemmas, and indicates the possibility of needing to involve 4 Our relation to animals and plants can be caring, but only to the limits of their responsiveness, and our relation to objects and ideas is more distant still.
ROLE PLAYING GAMES AND THE ETHICS OF CARE
oneselfin an un-caring response to a problem, which Noddings calls action under a "diminished ethical ideal" (Noddings, pp. 113-115). The primary example that Noddings gives is the case of the abused wife who kills her husband, faced with choosing the mainte nance of a relationship with him and his continuing well being over her own health and safety, as well as that of her children. She is obliged as one-caring to consider the needs of all concerned, and yet she is trapped if she does this-her situation is such that the only way to continue a caring relationship with her children and herself is to cease to behave as one-caring towards her husband (Noddings, p. 114). She cannot (to use Noddings' words) "receive" hun (empathize with him, treat him as one-caring) to change his habits (or so the author would have us assume for the sake of argument), since all of his responses have only an abusive direction to go in. She must choose to act then without reception, and without caring reciproca tion, and is thereby able to kill him to save herself and her children. From the possibility of a "diminished" ethical capacity, we may see the extreme complexity that a care-ethic entails. For the ba ttered wife, there is no simple decision (children and self over husband). As one-carmg, she must attempt to approach all parties involved, the antagonistic husband as well as the children, in a caring. receptive, absorbed fashion; this is comparable to some of the difficulty faced by the women in Gilligan's abortion study, who were forced to approach the felus, their friends, family, lovers, etc., to make the decision appl'opriate to their situations, as well as doctors and other official presences. Around the ethical one-caring, there are "circles and chains" of connected lives and personalities that must be ac counted for in any decision, and there is no simple one-caring/ cared-for relationship to go on; one-caring is also one cared-for in many cases, and the state of the relationship between a person and her/his spouse is different from the same person's relationship to her/his child, or to a doctor or other "official" type (Noddings, p. 27). In light of Gilligan's earlier revelations about the caring voice in moral orientation, what Noddings presents would seem to be pre cisely the proper sort of form for a care ethic to take. It is inclusive of relationship, of the positive nature of continuing relation to others that gives attention to specific contextually relevant aspects of any problem, encourages (in fact, requires) empathy, defines the self in relation to others within an ethical framework and strongly empha
sizes the importance, relevance and strengths of caring as a moral standpoint. It does, however, stand with a flaw that should be dealt with, namely the beginnlng assumption of mothering. This, as in Gilligan, leaves Noddings' work open to the feminist criticism of caring as the slave-mentality of women rather than an alternative method open to many. Noddings' model of mothering additionally provides the difficulty of caring as a superior-inferior relationship model, so that the attitude of the one--caring, although it is possible that Noddings did not intend it to seem so, may be seen as patroniz足 ing orpatemalistic even in its attention to reciprocation. Hermother/ teacher is in many senses much like the classical model in Greek philosophy of the Lover/Beloved distinction, which most definitely posits a superior-to-inferior, subject-to-object orientation.5 It pre足 sents a version of the care ethic that is convenient for exposition, but possibly faulty in practice. The problem of derogatory" femininity" and of superior/subject to inferior/object relation models can, of course, be repHed to by Gilligan's original contention of theme, and by the understanding that Noddings intends to show a more complete reciprocity in simple terms. A better way, however, to get around the sex-based objections to both of these is to remove the caring system's basis from the mothering metaphor and instead examine it in a way that more completely expresses what both Gilligan and Noddings attempted to demonstrate in a care-ethic. This can be achieved by changing to the metaphor of care-ethical action as a role playing game.
III Most of us are familiar with one or another type of role playing game, ranging in complexity from the simple design of "cops and robbers" to more mechanically complex role-playing fantasy games or the type of role-playing exercises that actors use in the everyday practice of their occupations. However different in scope they might be, however, an of these games have some crucial features in com足 mon, and it is these fea lures which, when connected, will provide us with a clearer metaphor for the ethics of care. The first important matter at hand for role playing games, as for 5
Plato, the Symposium.
ROLE PLAYING GAIvl.ES AND THE ETHICS OF
any game, is the question of the rules. In the case of any role playing game, there are some which, no matter what the rest of the game may be about, remain constant and this is particularIy relevant because of what rules serve to do: rules create the world of the game, and, like the laws of nature, determine what is and is not possible within the game itself for the players (E, p. 27). The idea of rules as "laws of nature" is especially important in a role-playing game, however, . because of the peculiar lack of physical limitations (such as a game board and pieces) to the sphere of game-reality. The rules, specifi cally in the sense in which they regulate what the game's "charac ters" may do in a given situation, thus define the context of the situations of play, above and beyond simply designating the means and proper procedures leading towards victory. Games of any kind usually have a goal of some sort for the players to achieve, thus constituting "victory" and the completion of the game. h1 role-playing games, however, it often appears that the game is open-ended-there is no specific moment, at the end of a chain of actions, that constitutes a "win." Take a very simple game, for instance, such as that ever-popular mainstay of childhood, "cops and l'Obbers." The first thing that becomes apparent, within the very simple explicit rules of the game itself, is that there are either many small victories possible (cop catches robber; robber evades cop) such that any protracted amount of play-time consists of many "games" of "cops and robbers," orno real, defined point of "victory" at all. This line of thought aSSlUnes that it is a specific chain of actions towards a pre-arranged goal that players must reach in a particular manner, so as to claim success, which is naturally the way most other sorts of games (chess, solitaire, Parcheesi, etc.) function. In the case of a role playing game though, success is defined, notin terms of a directional goal, but in terms of a developmental one; to be a good role-player, one must develop one's "character" within the framework of the rules of game-reality. Characters, in role-playing games, are the roles inplay which the participants take on and attempt to fulfilL In "cops and robbers," the roles are the "cops and the robbers," obviously, and victory in this game is defined in terms, no t of the achievement of specified success, but in terms of being a good cop, or a good robber. The rules of the game are simple, vague and aimed only at constructing the roles children who play the game often create whole "plots" within which
their character-development can maneuver. In more complex role足 playing games, the idea of "characters" and their interaction with a "plot" is explicitly set forth as the primary directive of play (such as in mass-market role-playing games). The basic strategy ofplay insuch a game, where the character and its development is the ulterior end of all game-action, is something to be approached primarily on the level of constructing an identity in relation to the other players and the collectively created plot, by way of the specific units of play-movement, those being interactive situations. From the beginning of the game, the rules specify the character's general attributes, positions, possessions, etc. (the cops always have guns and radios, and are motivated towards the protec足 tion of the law and all other attitudes that go with that motivation), and her lhis basic situation relative to the other players and their roles. Beyond that, however, prettymuch anything goes. In the more complex role-playing games, a plot is often created to give the players a more explicit framework within which to develop their characters, but plots begin and end. Player-characters go through many plots, until their existence is terminated, either in the course of play or by the player's wishes, and these plots provide situations that allow the characters to become further developed, in both physical and psychological attributes. 6 Again, the point of the game is charac足 ter development, a feature of role playing that carries over even to role-playing exercises used for more businesslike purposes (such as role-playing to learn about sexual harassment, where character development takes on an added meaning). As previously mentioned, the basic "unit" of play in a role足 playing game is a situation; in "cops and robbers," it may be the criminal act that caused the character roles to intersect, while in another game it might be something as mundane as a meeting in a park or as violent as hunting down monsters. Within this unit 6 In fantasy role-gnmes, for example, the player-characters usually begin their "Jives" at Level One, with some few possessions and an understanding of their basic abilities and traits. In the course of play, characters acquire new abilities and knowledge, new possessions and accumulate "experience points" towards the goal of reaching anotherLevel of character development, at which they can acquiremore and more abilities, knowledge, possessions, etc., and are able to go on ever more difficult (and potentially profitable) adventures. Adclitionally, many ofthese games also provide for psychological advancement ("Vampire: The Masquerade" is one of these).
ROLE PLAYING GAMES AND THE ETHICS OF CARE
situation, the characters exercise their roles and their capabilities in reaction to the situational context, and to each other-one can seldom find a role-playing game that one may play alone? Role development depends on types of relationship to external factors. In "cops and robbers," the opposition of the two primary roles is the basic developmental focus; in cooperational, rather than opposi足 tional games, teamwork is grist for the character development mill. The important thing to remember to win (i.e. to continuously create and improveupon the details of the character) is how one's character can function inrelation to others, because character growth is largely dependent upon the outcomes of situational interactions. How does this type of game become a metaphor for care-ethical attitudes and procedures? There are some obvious fundamental similarities to the three elements of Gilligan's care-ethic, such as the importance of contextual relevance and relation to others, but it would appear to lose some of Noddings' weight upon personal absorption, as well as only very roughly approximating Gilligan's concern of personal identity formation. It becomes necessary to find some way in which role-playing can be connected, not only to the basic rudiments of psychological care-procedures. but also to an ideal, a basis for the application of said psychological influences. We are also provided with the opportunity to expand the discussion beyond the realm of masculine / feminine oppositions, as well as the chance to explore and give greater importance to Noddings' "circles and chains" of relation, free of the trap of superior-inferior / subject足 object relational modeling as a rigid standard. To begin with, it is important that we examine the shape the role足 playing metaphor gives to Gilligan's original thesis. This can best be done by bringing forth the theatrical side of our metaphor, as used in the work of Erving Goffman, specifically in the sense that role足 playing as a game and role-playing as theater are fundamentally 7 The only ways that I could imagine to have role-playing without live role others are the "Choose Your Own Adventure" format books, which pre-set your options within a specified, predetermined role, and computer role-playing games. which, depending on the sophistica Han of the program. can either be CYOA forma ts or actually involve a looser, more freely played opportunity. In both cases, however, only the rudimentary mechanical development of character takes place; one cannot, with most computer games for example, undertake a personal change in one's life outside of the "you must find x by surviving the maze and killing off the monsters" sort of controlled situation.
similar in their usage of the presentation of specific roles in a controlled context. When Gilligan's subjects whoused the care-ethic described themselves and their moral lives, they spoke of, as men tioned above, the need to relate to others, via an often externally determined identity, in determinate situational cases. What this means, in Goffman's terms, is that people who use this ethical method can be said to be "on stage" (where the stage is roughly the same as the contextual game-field created by the rules in a role playing game), acting their parts both with other "performers" and with "audiences" (GoHman's examples include relations between shopkeepers and clients and between the staff and passengers on a luxury liner, as well as less formal cases of people visiting each other's homes for tea) (PS). Instead of characterizing interactive behavior in terms of its positive or negative perceived psychological value or gender con nection, GoHman begins and ends his study with observations of people interacting with each other, stated in terms of a theatrical framework; he claims to study "social life" in terms of the theater (PS, introduction). His primary starting point is the "social establish ment," "anyplace surrounded by fixed barriers to perception [places of business, homes] in which a particular kind of activity regularly takes place" (PS, p. 238). Within the social establishment, roles are assumed, such that there is performed behavior and "backstage" behavior; in role-playing as a game, we might say that there is, likewise, a division between character action (activity of the player in-character) and player action (action out-of-role or out of play). Performers and audiences make up "teams," such that everyone is, to some degree, a performer and a member ofthe audience all at once, and performances are interactions between teams taking on both functions (PS, pp. 178-180). In the same way, players of a role playing game interact as "teams" with the situations of play-cops and robbers relate to each otherboth in and out of character, and with different agendas in mind, and fantasy-gaming involves the interac tion of teams of players/characters with non-player personalities portrayed by a "game-master" or "storyteller." Contextual relevance, maintenance of relationship and concep tion of identity take on a slight twist, seen through GoHman's lens. Contextual relevance, in addition to indicating a tendency to deal with concrete situational cases instead of abstractions, can also be
ROLE PLAYING GAMES AND THE ETHICS OF CARE
seen as the practice of functioning in context instead of trying to escape from it, or, metaphorically, acting/moving/performing on stage or in the game. It indicates the great importance of the numer ous, almost overwhelmingly complex parts of any situation that necessarily effect decision-making procedures. Maintenance of rela tionship, too, gains an added meaning, in the sense that it entails not simply the attempt to continue a relation, but a situation as well; it is the practice of staying "in character," and keeping the plot of the game alive, and additionally serves as the major vehicle by which roles are exercised in context. The most powerful and clarifying addition of Goffman to Gilligan, however, lies in the field of determination of identity. Like the student in Gilligan's study who spoke of herself as having layers of self. like an onion, GoHman presents an idea of self that is often externally determined identity-elements combined with more inter nal, personal self-concepts (Gilligan, pp. 67-68). Instead of setting them up in a confusing opposition, Goffman separates the two into character and performer (character and player): "The self, as a performed character, is not an organic thing that has a specific location whose fundamental fate is to be born, to mature and to die; it isa dramatic effect arising fmma scene that is presented ..." (PS, p. 253). Characters and impressions change from situation to situation, as the player plays them, out, and they develop into different things over time and given varying situational experiences, just as players themselves change and grow. Gilligan's student might just as well have expressed her confusion about her identity as a question of where the character and the player aspects of herself begin and end. Based on Goffman and Gilligan, we find an ethical care role player who functions by (as Noddings says) "receiving" the impres sions of others in complex determinate situations, as well as project ing (performing) role-impressions to others. The behavior is meant to maintain the medium ofthe interaction (maintain the relationship, keep the game going), both by using certain aspects of a character and attempting to further develop and perfect that character. If this is in fact the case, however, where does the moral motiva tion lie? It is one thing for a p]ayer, once engaged in play, to wish to continue the game. It is an entirely different motivation that leads a player to begin the game in the first place, and that is where Noddings and Goffmanmeet. Noddings, in providing the idea of the
LAURA M. BERNHARDT
desire to continue positive "caring" relationships with others, gives a viable motivation to become a performer / player in this particular game, although she only relates the fulfillment of this need through the somewhat problematic forms of the mother--child, teacher student relationship models. While she alludes to a multitude of possible roles for ones-caring and cared-for' s, ("circles and chains"). Goffman's contribution is the addition of an understanding of the shifting quality of characters and role-identities ("effect[s] arising from a scene that is presented"), such that, while relationships differ as to when one acts as performer or audience, (one--caring or cared for, ina loose comparison), they are not necessarily superior-inferior in design. The desire to continue the game has more to do. in this case. with the desire to continue and learn to manage character impres sions,tocreatenewopportunitiesforinterrelationindifferentways.8 As an ethical procedure. care as role-playing is based in a virtuous ideal (if we are to approach it as a mature and intelligible moral standpoint in the way that Gilligan and Noddings would advocate) of positive concern for others. and its methods are actions designed to permit the growth and continuation of that care, via role activity in relation to the particular situations of oneself and others. To "win" this game (Le. to be a positive or "good" moral agent in the care-ethical system), one must develop one's role-relations with others ina fashion that fosters further relationship (whenapplicabl e) and self-development on all sides. thus maintaining the game. The action of play revolves around focusing the virtuous caring ideal (seen in Gilligan as the desire to avoid harm to others and shown in Noddings as an empathy intended to be helpful) through the lens of relationship and reception (as performer and as audience). Through the metaphor of the role-playing game, the care-ethic is seen as a moral view incidentally, but not solely, dependent upon the gender of the agent (since people. as social creatures, interact regardless of gender), free of the problematic "feminine" mothering-orientation· s tendency towards stereotypical "female/inferior" interpretation.
The chapter on "Impression Man"'gement" is especially relevant here.
ROLE PLAYING GAMES AND TIm ETHICS OF CARE
Forsyth, Donelson R.. Judith L. Nye and Karl Kelley. "Idealism, Relativism, and An Ethic of Caring." The Journal ofPSljchology, Vol. 122, No.3 (May 1988): 243-248. Gilligan, Carol. In A Different Voice. Cambridge: Harvard University Press, 1982.
[E] Goffman, Erving. Encounters: Two Studies In The Sociologtj Of Interaction. Indianapolis: The Bobbs-Merrill Company, Inc., 1961.
[PS] Goffman, Erving. The Presentation of Self in Everyday Life. Woodstock: The Overlook Press, 1973. Gould, KetayunH. "Old Wine In New Bottles: A Feminist Perspective On Gilligan's Theory." Sodal Work, Vol. 33. No.5 (Sept.-Oct. 1988): 411-415. Kohlberg, Lawrence. The Philosophy ofMoral Development. Essays on
Moral Development, Vol. 1: 'Moral Stages and The Idea of Justic(~. New York: Harper and Row, 1981. Noddings, NeL Caring: A Feminine Approach to Ethics and Moral Education. Berkeley: The University of California Press, 1984. Pratt, Michael W., Gail Golding, William Hunter and Rosemarie Sampson. "Sex Difference in Adult Moral Orientations." The Journal ofPersonality, Vol. 56, No.2 (June 1988): 373-391. Puka, Bill. "The Liberation of Caring: A Different Voice for Gilligan's 'Different Voice' ." Hypatia, Vol. 5, No.1 (spring 1990): 58-82.
17 ASCETICISM AND THE VALUE OF TRUTH IN
THE GENEALOGY OF MORALS
Julie Anne Buchsbaum
From the outset of The Genealogy ofMorals, Nietzsche calls for a sweeping critique of traditional morality. Lamenting the absence of other such attempts, he undertakes an investigation into the specific conditions which have historically genera ted moral values and, most importantly, re-estimates the worth of these values whose goodness has hitherto been assumed to be impervious to doubt. Asceticism is the most sigruficant ethical ideal Nietzsche narrows in on in The Genealoglj ofMorals, and he devotes an entire section to its analysis; indeed, he has good reason to do so since, for Nietzsche, the element of self-overcoming is the backbone of all moral codes (Kaufmann, p. 211-3). Suspecting a pathological perversion of the will and the origin of bad conscience in the Christian treatment of asceticism, Nietzsche traces the manifestations of this ideal and how it has devolved under the sway of institutionalized religion; furthermore, he exposes its interrelation with the will to truth and power, and questions its overall contribution to human existence. In this paper, I examine Nietzsche's genealogical inquiry into asceticism in order to determine whether he rejects this nation as irretdevably tainted through its monopolization by the Christian-moral intel'pretation. Asceticism may nat be inherently good, but is it necessarily and essentially decadent, anti-body and life-negating? I shall argue that Nietzsche does not abandon the ascetic practise and tha t a Nie tzschean re-valuation of the term can promote an opposing form of asceticism which departs from the traditional debilitation of the body and an authoritative position of centered meaning. In The Genealoglj ofMora Is , Nietzsche declares that ethics are bam of blood, cruelty and torture. His analysiS ofWesternetliics takes us as far back as the very geneSis of religious practises, which he perceives to have been founded on a debtor-creditor relation be tween ancient ancestors and their descendents (GM, p. 222). AscetiBuchsbaum is asenior at Beloit College. A philosophy and classics double II/ajar, she lives ill Milwaukee, Wisconsin. Her paper is tlte result of a semester-long independent study of Nietzsc/te'S philosophy.
Episteme • Volume IV • May 1993
JULIE ANNE BUCHSBAUM
dsmhas its origin in the development ofthls relationship. First of all, pain, according to Nietzsche, was of instrumental value in forging a conscience out of the brute instinctiveness of an uncultivated race of humans. Noble rulers carved the awareness of debt and duty into the minds of others through torture and the sacrifice of those who transgressed this code of obligation. In addition, pain often consti tuted. in archaic civilizations, the repayment of a debt. Nietzsche asserts that the creditor was entitled to indulge his lust for power with complete license over an insolvent debtor, who offered up his humiliation and pain to the delight of his tormentor. The debt was thereby erased by the psychological "pleasure of rape" (GM, p. 196). However appalling these revelations may seem to us, it soon becomes clear that Nietzsche is not unearthing the nature of this sadistic contractual arrangement in order to demonstrate how far modern society has come subsequently in repudiating such barbar ity. On the contrary, Nietzsche claims that cruelty pervades and underlies all higher culture-the instinct to punish has not disap peared but only been driven inward; consequently, people are now ashamed to experience these urges which, bottled up inside, poison the once healthy human being with concepts of sin and self-accusa tion. Here then, is the inception of the hypocritical bad conscience which Nietzsche never tires of railing against; what had formerly been experienced with" the naive j oy and innocence of the animal" is now spurned and denaturalized (GM, p. 199). The process which divorces the individual from his or her instinct to domina te inspires the repudiation of the whole realm of bodily functions and a general aversion to life, since life is grounded in these drives and functions. This radical inversion of the value of the instincts engendered the ascetic ideal. In one of his numerous foreshadowings of Freud, Nietzsche declares that "[a]11 instincts that are not allowed free play turn inward" (GM, p. 217); thus, under historical conditions which prohibited the external venting of hostility, humans thwarted and internalized violent urges and ascetic priests ruthlessly terrorized this new "enemy"-their own impulses. Nietzsche deplores this form of asceticism as being an ancillary principle of Christian anti sensual metaphysiCS. This practise encourages us to dichotomize spirit and body, privileging the spirit as pure, active, good and deriding the body as the source of all impurity, passivity and evil. Any moral doctrine, such as that of St. Paul, which sees only the
ASCETIOSM AND TIm V ALUE OF TRUTH
darker aspects of the passions and thus extols their annihilation, provides fodder for some of Nietzsche's most caustic polemics against Christian asceticism (GS, p.189). Nietzsche rejects the ascetic ideal as it is informed by the Christian bifurcation of human nature on the grounds that it institutes an impossibly distorted idealization that can never be realized; he asserts further that its sole purpose is to assure humans of their irremediably flawed natures and to rein force a blind hatred of life. Humankind's sense of indebtedness to God (the most ancient ancestor) is then carried to unprecedented extremes. As Nietzsche writes, with characteristic hyperbolic flam boyance, Christianized man stretched himself upon the contradiction 'God' and 'Devil' as onarack. He projected all his denials of self. nature, naturalness out of himself as affirmations, as true being, embodiment, reality, as God ... as endless torture, as hell, as the infinitude of guilt and punish ment. In such psychological cruelty we see an insan ity of the will that is without parallel ... [the] will to poison the very foundation of things (GM, p. 226). Furthermore, this insane will to power which lies at the heart of the ascetic ideal is disguised by adherents of Christianity; Christian sects deny that the motivating source of asceticism is an internalized impulse to persecute others because this impulse is branded as immoral by the devotees of slave morality. Thus, the ascetic ideal is shroudedin a moralhalowhichprevents it frombeingproblematized or re-interpreted. Following Nietzsche's analysiS in The Genealogy ofMorals, there is a second way in which the will to power is expressed through the ascetic i.deal At the beginning of the third essay, Nietzsche claims that the manifold deployment of this ideal suggests a deeper ten dency of human nature at work-the desire to stave off a void at all costs. He states that "[o]ur will requires an aim; it would sooner have the void for its purpose than be void of purpose" (GM, p. 231). The ascetic ideal is seen to be intimately bound to notions of truth and the whole project of intellectual inquiry, even if the intellectual disavows the more explicit form of sensual asceticism. The ascetic ideal, then, also permeates the will to truth, which in tum actuates philosophical
JULIE ANNE BUCHSBAUM
interrogation. The intellectual mode of asceticism veils the constant sense of a chaotic meaninglessness which lies beneath the seemingly airtight categories of truth, identity and unity-the chaos that threat ens to explode the apparently substantive nature of these notions. Nietzsche maintains, somewhat paradoxically, that even those academics who consider themselves atheists and anti-ascetics are, in fact, the unwitting vehicles of the ascetic ideal in its most purified form. He writes, "the intellectual stoicism which in the end re nounces denial quite as strictly as it does affirmation ... spell[s] asceticism every bit as much as does the renunciation of sensuality" (GM. pp. 287-8). In other words, the philosopher or scholar may disengage him or herself from metaphYSical debates in the realiza tion that traditional categories of meaning fail to accommodate the world of becoming; nonetheless, this philosophical restraint is in spired by precisely the same force (the will to truth) as is the belief in substance and unity. The scholar who prudently withholds judge ment in the face of the now-recognized enigmatic nature of existence still, at bottom, retains the beliefin the diametrical opposition of truth and falsehood, and esteeming the truth ab ove all else, is driven by the desire to avoid error. The relentless will to comprehend the essence of reality yields, at last, the notion that there is no such essence to be comprehended; thus, the Christian exaltation of "tw.th" turns back upon itself and destroys its own ethic (GM, p. 297). The will to truth, then, is another form of the ascetic ideal, and this will mitigates our fear of the void by positing attainable truth as its aim, even when this truth is non-truth or absence of truth-even when we have banished the dogmatic trappings of substance philosophy. Intellectual asceticism is related to priestly asceticism in its determination to impose meaning on human existence. While the ascetic priest tries to eradicate the passions which chain him to the temporal world, this very attempt fortifies those chains byproviding a meaning and direction for his life as well as the life of his commu nity. From his violent denunciations of the earth spring a reason for the continuation of life, much as the scholar's sl.lspension of judg ment in the hopes of attaining truth in non-truth perpetuates his or her lust for truth. In a sense, then, both forms of asceticism serve an instinct for self-preservation. Although the ascetic ideal intensifies suffering and propagates illusions, in doing so it makes existence tolerable because it interprets suffering and masks those illusions as
ASCETIaSM AND THE VALUE OF TRUTII
truths. Thus, the willing of unconditional truth (as a means of sustaining life) seems to be a process Nietzsche might approve of, since he would be the last to counsel a suicidal resignation to despair. However, on the very last page of The Genealogy of Morals, Nietzsche cuts the bottom out of any reading of the ascetic ideal which renders it ultimately life-affirming. He writes, this whole process ofwilling ... this hatred of human足 ity, of animality, of inert matter; this loathing of the senses ... this longing to escape from illusion. change. becoming. death, and from longing itself ... signifies ... a will to nothingness, a revulsion from life, a rebellion against the principle conditions of living (GM,p. 299). Although Nietzsche's genealogical path now seems highly inconsis足 tent (first, the ascetic ideal was seen as the source of bad conscience, then as an expedient survival-mechanism and finally, as a reaction against life), Nietzsche, in the end, is indicating what lies beneath the will to truth and the ascetic ideal; after all, Nietzsche never asserts that life itself is the ultimate goal of the human spirit. Indeed, while asceticism may have the immediate result of warding off suicide by "spreading over existence the blandishments of illusion" (BT, p. 108), if completely realized, its final culmination is death, its deepest motive is the annihilation of all existence. Being a principle which rigidly excludes other interpretive systems, the ideal reveals itself as being the most extreme expression of the will to power. If error is a necessary condition of life (and Nietzsche believes it is). then the search for absolute truths, truths untouched by change, sin or human failing. must be prompted by a more fundamental drive than that of self-preservation-it must arise from the desire for power. To want to gaze on the bare Medusa-face of in-itself reality, to unveil the naked Urstoffe of the cosmos is to desire petrifaction~ it is to desire the impossible-the shedding of one's own particular, finite, subjective point of view and the mind-body aggregate tha t makes thought and curiosity possible. To want to tear ourselves out of the web of becoming to get a glimpse at what lies "beneath" or "behind" its oscillating illusions is to desire that utmost power and the end of all becoming-death. Therefore, Nietzsche successfully fleshes out the claim which launches the third essay; the claim that the will would
sooner have death (the void) as its aim and purpose than lack a purpose. Thus, at the end of The Genealogy ofMorals, are we left to assume that. thanks to Nietzsche, asceticism has been once and for all debunked, devalued, exposed as aninclination one would do well to avoid? Are we justified then in abandoning ourselves to hedonistic dissipation and forgetting about philosophy? Not unless Nietzsche has undergone a complete apostasy, for his works abound in pas~ sages in which he reviles the whole notion of eudaemonistic laisser alles. Repeatedly he heaps derision upon those who would escape from suffering, bury their heads in the sands of ignorance and acquiesce to a lukewarm, "go-with-the-flow" mentality. Letting oneself go and submitting indiscriminately to every whim certainly does not pave the way to greatness for Nietzsche. Nothing worth~ while at any level of culture has ever been attained without the expense of ruthless sacrifice. Furthermore, the will to truth as expressed through the ascetic ideal is also not entirely discarded. Again, throughout Nietzsche's works he makes it clear that he holds the uncompromising search for truth and "the desire for certainh} ... [to be] that which separates the higher humanbeings from the lower" (GS, p. 76). Indeed, if Nietzsche had decided to dispense with reason and with the metaphysical categories of value that are inevitably congealed in. philosophical discourse, he would have ceased writing after his critique of asceti~ cism. That is patently not the case. Though I believe Kaufmann overestimates Nietzsche's regard for rationality. plenty of textual evidence bears out his claim that Nietzsche "proposed to measure power and weakness in terms of man's willingness to subject even his most cherished beliefs to the rigors of rationality" (Kaufmann, p. 232). I believe Nietzsche values lOgiC, consistency and reason, but is also equally aware of their limitations and is unwilling to assign them a superior status in philosophical inquiry. The irrationality Nietzsche objects to is revealed in the will to unconditional truth which automatically rejects deception, inconsistency and instinct and endorses rational consciousness ("their weakest, most fallible organ") as the only key to truth (GM, p. 217). How, then, might one envision a Nietzsche an reappropriation of asceticism that does not fall prey to the Christian-metaphysical interpretation? Is it possible that in Nietzsche's writings ideas that
ASCETICISM AND THE VALUE OF TRUTH
take themselves too seriously by denying the absence of transcen dental grounding "are themselves released from the ideologies and passions that have held them and now lead to thoughts and passions that depart from and contradict the earlier. authoritative structures," as Scott asks (GM, p. 211)? First of all, one simply needs to refer back to The Genealogy of Morals to discover how Nietzsche re-interprets asceticism for the philosopher who affirms existence. Simply put, "[a]sceticismprovides him with the condition most favorable to the exercise of his intelligence" (GM, p. 243). Shed of its self-important piety, the notion of asceticism for the philosopher is severed from concepts of moral scruple. The philosopher who is driven to create from an over-abundance of strength does not tame his or her desires out of a hatred for the bodybut rather because his or her passions are completely channelled into the creative process and thus spiritual izedinto a higher level of humanexpression. According to Nietzsche, only the weak would be so afraid of their impulses that they would seek to eliminate them entirely. Indeed, if anything, the philosopher does not wish to destroy the senses and paSSions but rather to sharpen and deepen them in order to intensify the power of the philosophical project. The priestly form of asceticism, in contrast, aims at the extirpa tion of the impulses because it is driven by a hatred of fles.h and resentment toward life. The ascetic priest ultimately wIshes to inflict suffering. in the form of guilt, upon those who witness his self mutilation. and to reinforce the community's adoration of him. The affirmative philosopher, on the other hand. views his or her self restraint simply as a necessary condition of unrelenting devotion to creativity. The philosopher is indifferent to society's assessment of him or her; the distress of others which may result from this indiffer ence is an inadvertent side-effect, not the primary goal. Thus, philo sophical asceticism is stripped of its romantic ostentation and the self-aggrandizing righteousness that pollutes theChristianizedform. Secondly, in seeking a counter-ideal to Christian asceticism, one might consider Nietzsche's brief comments about art in The Geneal oglJ OfMora Is. Here he makes it clear that the artistic endeavor at least approximates an adverse ideal: "As for art ... it is far more radically opposed to the ascetic ideal than is science. In art the lie becomes consecrated, the will to deception has good conscience at its back" (GM,p. 290). Although Nietzsche rejects the artist as too malleable to
JULIE ANNE BUCHSBAUM
defend this counter-ideal. this brief passage offers at least a sugges tion of what Nietzsche might have been directing us toward. The artist aims at deception in that he or she depicts the surface. the plastic. the superficial. and lives at the level of externality by employ ing images that appeal to the senses rather than trying to plumb the depths of reality and ensnare the object of inquiry with a cold, conceptual net. Whereas the metaphysician wants to maintain that his or her theoretical matrix corresponds directly to in-itself reality, the artist entertains no such notions-the artistis aware thatthe mask is a mask-and thus has a clear conscience when manipulating constructed images. In a sense, then, Nietzsche's own writing is an artistic act in that it refers not to an extralinguistic source of validity but is authorized performatively and with the reflection of its own genealogical configuration. Nietzsche's interrogation of established ideals can be construed as an artistic exercise in the sense that he analyzes different ways in which a term is masked throughout its historical transmutations and at particular historical junctions indeed at the very start of the inquiry he announces his intention to explore the historical conditions which give rise to moral values rather than to probe into the supposed underlying essence of these values. Nietzsche does not succumb to the illusion that his analysis is independent of masks or that it occupies an objective and neutral (rather than perspectival) status. At the same time, however, I believe Nietzsche is also driven by the will to truth in his genealogical investigation and thus is impli cated in a modified version of the ascetic ideal. What is at stake in Nietzsche's writings is a kind of truth-seeking that uncovers and exposes the blind, restrictive and un-self-critical nature of meta physical discourse. It is not that Nietzsche believes that the Christian corruption of asceticism conceals and distorts a more fundamental truth about asceticism which will be offered up to us by means of the genealogical critique. He does not suggest that there exist such "correct" definitions. Instead, Nietzsche's ascetic truths consist in problematizing the very notion of stable truths. On the other hand, Nietzsche does make use of certain descriptive formulations and theoretical models (such as the will to power and the eternal recur rence) and, though he never systematizes these concepts, he does rely on them to an extent that indicates his belief in their applicability and validity. Thus, he is engaged to a certain extent in a continuation
ASCETICISM AND TIlE
V AWE OF TRUIH
of the ascetic project in one of its unpredicted transformations. What alters this ideal in Nietzsche's work is, first, the ubiquitous movement of self-overcoming that constantly destabilizes the au thority of his own position and, secondly. his re-thinking of the object of philosophical inquiry. Nietzsche's own hypotheses are inspired by the will to truth in that they provide an organizing mechanism or serve an explanatory function; at the same time, however, they are undercut by Nietzsche's claim that all such mechanisms are interpretations shaped by one's perspective and thus are exposed as regulatory fictions, models, frames. This strategy of self-overcoming or self-revealing frees Nietzsche's inquiry from thebadfaith of the Christian ascetic ideal which denies its hermeneu tic falsifying and masking. Secondly, Nietzsche re-describes the notion of truth and thereby redirects philosophical inquiry. For Nietzsche, what renders the will to truth so destructive and life negating is its insistence on attaining unconditional truths which can be grasped with certainty in an ahistoricized and universalized manner. To speak of right or wrong, truth or falsehood per se in isolation from a particular context is meaningless. For Nietzsche, the concept of truth is continually destabilized by the reinsertion of terms into the specific context in which they arise and by the refusal of abstract truths. He states, "only that which has no history can be defined" (GM, p. 212); thus, since all truth-claims and values evolve in a temporal medium and are influenced by the historical situation in which they appear, no term, it seems, is definable in any ultimate sense. On the other hand, this realization need not abort the truth seeking project because this project, reconsidered in the Nietzschean manner, does not aim at stable, abiding categories of meaning. The truth-object is, for Nietzsche, always in flux, always subject to continual reappropriation by various viewpoints, yet this lack of fixity does not detract from the affirmative philosopher's project; on the contrary, this instability enriches the meaningfulness ofhis or her interpretations by allowing for the existence and value of other interpretations. Therefore, unlike advocates of the Christian ascetic ideal who perceive the revelation of the unstable nature of truth claims to constitute an impediment to all further inquiry, the affirma tive thinker has enough resilience to absorb this revelation into his or her search for conditional truths. This realization does not exempt us from philosophizing but instead invites us into a process which is
JULIE ANNE BUCHSBAUM
infinitely more challenging, dangerous, life-affinning and fruitful. In conclusion, I believe that one can speak of the possibility of an asceticism which springs from an overabundant affirmation of life and which avoids slipping into the extremes of either an uncritical reverence for truth or an abandonment of the whole truth-seeking project. First of all, in the sensual realm, if one's self-restraint is not promptedby a self-righteous desire to evoke guilt inothers, involves the deepening and sub limating of the passions and combines a spirit of levity with a refusal to capitulate to a single established norm, the ascetic can embrace life as a whole while transfigurmg certain aspects of it. Secondly, if we can avoid baptizing our beliefs as "truths" with deadly solemnity and endure the destabilizing move足 ment of questioning in the absence of solutions, this seems to me to involve a Nietzschean reformulation of asceticism, an aspiration towards the Nietzschean ideal of an "artistic Socrates" in conjoining the awareness of the indeterminacy of the theoretical ground on which we stand with a willingness to commit ourselves passionately to our beliefs. This self-reflexive form of asceticism contains the strength both to confirm that some ideological constructs are mean足 ingfulmsofar as they address humanneeds and to recognize the role that convention plays in their construction. The abstractions oflogic or the rules of grammar, for instance, may facilitate communication and co-existence with other human bemgs, but in certain contexts they are certainly notbinding. In the end, Nietzsche does not want us to abandon either sensual or intellectual asceticism. The brilliance of his reformulated asceticism lies in the fact that it does not compel us to eliminate all masks but to realize that we cannot exist without them and that a mask which sits before our eyes long enough for us to forget it is there must be problematized and historicized until the nature ofitsfluidity, contingency, and open-endedness flows through the fissures.
AscETIasM AND THE VALUE OF TRUTH
Kaufmann, Walter. Nietzsche: Philosopher, Psychologist, Antichrist. New Jersey: Princeton University Press, 1974.
[BT; GM) Nietzsche, Friedrich. The Birth ofTragedy and The GenealogtJ ofMorals. trans. Francis Golffing. New York: Doubleday & Company, Inc., 1956.
[GS) Nietzsche, Friedrich. The Gay Science. trans. Walter Kaufmann. New York: Random House, Inc., 1974. Scott, Charles E. "The Mask of Nietzsche's Self-Overcoming." Nietzsche as Postmodernist: Essays Pro and Contra. ed. Clayton Koelb. Albany: State University of New York Press, 1990.
Two ApPROACHES TO ARTIFICIAL INTELLIGENCE: AN ANALYSIS OF HEIDEGGERIAN AND DREYFUS IAN CRITIQUES Nicholas K. Gracilla
There exist two broad research paradigms in Artificial Intelli gence (AI) which differ radically in their attempts to reproduce human understanding through the use of computers. The dominant paradigm. which I call Traditional AI, has focused on formalizing the process of thinking into rules, symbols and representations of the world. As such, its roots can be found in the philosophical traditions of reductionism and rationalism. The second paradigm, which I call Parallel Distributed Processing (PDP), has focused onusing comput ers to emulate the neurological structure of the brain. Concerned less with formalization than underlying computational structure, its approach has developed from the neurosciences, Gestalt Theory and work in perception. The short history of both paradigms is riddled with fantastic claims and unsupported predictions of success. Hurbert L. Dreyfus was one of the firs t to critically examine these claims-and concluded that Traditional AI was fundamentally and irreparably flawed in its approach. His criticisms, grounded in the works of Martin Heidegger, make an in-principle argument against the pos sibility of formalizing humanin telligent behavior. Traditional AI has approached the formalization by postulating mental representa tions, which both Dreyfus and Heidegger reject. Dreyfus' in-prin ciple argument, however, holds no weight against the non-formal ized, non-representational paradigm which PDP uses, and he is notably less critical towards it because of this. I believe his criticisms of Traditional AI are accurate. When addressing the AI project in its entirety, however, his argument appears to slip toward a different claim. He notes that, currently, "human beings are much more holistic [than PDP networks] ," and in emphasizing this holism, he suggests the minimal unit of analysis of intelligent behavior mayan entire human-like organism in the entire human culture (MVB, p. 39). Dreyfus' holism, which re-emphasizes As a philosophy and computer sciellce double major, G'racilla wrote Utis paperfor his senior compreltensiveprojectat Denison University. A native ofWarrell, Ohio, he plans top!lrsue a doctorate in philosophy at Northwestern University.
Episteme • Volume IV • Mm 1993
"our needs, desires and emotions" as well as the importance of our"... [having] a human-like body with appropriate physical movements, abilities and vulnerability to injury" appears to permit only human organisms to exhibit intelligent behavior (MVB, p. 39). In this essay,lhope to show why Dreyfus' in-principle argument correctly criticizes TraditionalAI, yetmay not address some inherent characteristics of PDP architectures. 1 will also suggest that his broader, holistic argument against AI in its entirety may not be justified. My strategy will be to characterize both paradigms, em足 phasizing their modes of representing information. I will review Dreyfus' criticisms and their roots, as expanded in an analysis of human understanding offered by Heidegger. 1 will show why the critique is applicable to Traditional AI, yet inapplicable to PDP systems. His transition to a holistic claim concerning intelligence will then be evaluated, and hopefully shown to be untenable based on his own arguments leveled against Traditional AI. The computational paradigm used by Traditional AI approaches has been described by Newell and Simon as a physical symbol system. It can be characterized by its use of abstract symbols to represent salient features in a "microworld"-an artificially con足 structed problem domain which simulates a subset of the real world. Syntactic rules manipulate these symbols to reflect the processes and relations which occur in the microworld. The technique is powerful: symbols and rules are capable of representing every fact and process which can occur within the constraints of the explicitly defined problem domain. It is questionable, however, whether a system which uses this approach can replicate human intelligence. Terry Winograd's SHRDLU-a Traditional AI program which processed natural language sentences concerning a microworld of blocks, spheres and pyramids-typifies two primary problems of symbolic representation in microworlds. Within the restricted prob足 lem domain, SHRDLU could correctly respond to questions or commands such as "Can a pyramid be supported by a block?" and "Find a block which is taller than the one you are holding and put it into the box" (WeD, p. 7). Remarkable as this may seem, Dreyfus points out-and Winograd readily admits-that nothing even ap足 proaching an understanding of natural language is modeled in SHRDLU. For example, in reference to "owning," Herbert A. Simon remarks:
Two ApPROACHES TO ARTIFICIAL INTELLIGENCE
... SHRDLU's test of whether something is owned is simply whether it is tagged "owned." There is no intensional test of ownership, hence SHRDLU knows what it owns. but doesn't understand what it is to own something (d.WCD, p.13). SHRDLU cannot understand "what it is to own something" because it is isolated from the context in which "owning" is meaningful. Moreover, SHRDLU is incapable of understanding anything because it is not "in" a context at all-its microworld contains only uninterpreted facts concerning geometric objects and the relation ships between them. A context which provides meaning, however, does not consist of a body of uninterpreted facts and relations. "Owning" is meaningful, for example, in a context of social interac tions and property rights in which one participates. This confusion between a meaningful context and the uninterpreted facts which compose a microworld led Dreyfus to reject the idea that microworlds are "worlds" at all. A set of interrelated facts may constitute a universe, a domain, a group, etc., but it does not constitute a world, for a world is an organized body of objects, purposes, skills and practices in terms of which hu man activities have meaning or make sense (WCD, p. 13). Thus, since the semantics of "owning" are context-sensitive to a human world which SHRDLU is not in, the concept of "owning" is meaningless to SHRDLU. The second fundamental problem of Traditional AI typified by SHRDLU concerns its method of knowledge representation. SHRDLU's microworld consists of explicitly statable facts and rules a kind of knowledge Gilbert Ryle calls "knowing that" (RyIe, p. 28). Thus, SHRDLU can express "that a sphere is in the box" since this is explicitly represented in its microworld. Yet, there is a qualitatively different kind of knowledge which SHRDLU is incapable of repre senting. Ryle calls this "knowing how"-a kind of knowledge which indicates a skill or capability. For example, one might know how to play Bach's preludes, how to swim or how to shoe a horse. Know
how knowledge concerns an active doing, a performance, as opposed to an explicitly stated fact or rule. Such knowledge is obtained through learning from multiple experiences: one learns how to ride a bike by continually getting on the saddle and pedaling. This mode of acquisition reveals the qualitative difference be tween the two knowledge types. One cannot explicitly articulate to a child "how to balance" in any meaningful or helpful way. Learning "how to balance" is not a process of studying and memorizing facts concerning one's center of gravity, the effects of motion on objects and so forth. Even after coming to know "how to balance," one cannot easily articulate such facts. Dreyfus remarks "the fact that you can't put what you have learned into words means that know-how is not accessible to you in the form of facts and rules" (MOM, p. 16). Since SHRDLU is designed as a physical symbol system, its micro world necessarily consists of explicitly statable facts and ru1es. Thus, any knowing-how one wishes SHRDLU to have access to must be converted into knowing-that-a deeply problematic undertaking. Considering the efforts made inAI during the 1970sand beyond, it is dear that the criticisms concerning microworlds and the repre sentation of knowing-how have become influential in Traditional AI theorizing. New proposals, such as Minsky's frame system or Schank's scripts, used complex representational structures in an attempt to address these issues. Consider Schank and Abelson's script system: A script is a structure that describes appropriate sequences of events in a particular context. A script is made up of slots and requirements about what can fill those slots. The structure is an interconnected whole, and what is in one slot affects what can be in another. Scripts handle stylized everyday situations. They are not subject to much change, nor do they provide the apparatus for handling totally novel situations. 'Thus, a script is a predetermined, stereotyped sequence of actions that defines a well-known situation (Schank. p.41). .
Scripts attempt to enrich microworlds by representing human world interactions, and attempt to capture the kind of common sense know-how humans use in everyday situations. Schank cites
Two ApPROACHES TO ARTIFICIAL INTELLIGENCE
the following short story as evidence of this: "John went to a restau足 rant. He asked the waitress for coq au mn. He paid the check and left" (Schank, p. 38). Human understanding embodies much more infor足 mation than presented in the story: we understand that John ate the coqaumn, for example, thathesatata table, ate the meal with utensils and so on. Human knowledge of how one eats at arestaurant allows us to understand this story in ways Traditional AI systems, which did not model human practices or common-sense, could not. Do these richer representational schemes offer any significant improvements ? Scripts offer a significant advantage in their ability to use expec足 tations, in the form of unfilled or default data slots. One such slot, for example, could contain information about John's sitting position at the table. Thus, if the story had an additional line, such as "When the gun fired, John hit his knee on the table," the script could account for John's unfortunate reflex by already having information concerning his leg placement beneath the table. We imagine that such a script, in order to handle the incredible amount of information encountered in a restaurant, would be qui te complex. Minsky, commenting on his frame system (similar in many regards to a script) notes, ... the list [of facts] is not endless. It is only large, and one needs a large set of concepts to organize it. After a while one will find it getting harder to add new concepts, and the new ones will begin to seem less indispensable (WeD, p. 11). Minsky's approach of decomposing the common-sense knowledge of, say, "how to use a spoon" is characteristic of AI's information processing model: the use of a spoon is a conglomeration of a huge number of actions and rules-the degree of tension the fingers must use to hold the spoon, the proper angle to hold the spoonso that food will not slide off and so on. Moreover, Jerry Fodor, another Traditional AI theorist, ques足 tions the importance Ryle and Dreyfus place on the distinction between knOWing-how and knowing-that. He remarks, there is a real and important distinction between knowing how to do a thing and knowing how to
explainhow todothatthing. n. But what has this to do with the relation between knowing how and know ing that" (Fodor, p. 71)? In refuting one's inability to articulate knowing-how as evidence of a qualitatively different kind of knowledge, Fodor offers a distinc tion between mental competences or skilled abilities and mental traits-like intelligence or sensibility. Knowing-how to do some thing is evidence of a competency, but not necessarily a trait like intelligence. Moreover, traits like intelligence are not dependent on competencies. By drawing this distinction, he suggests that, if John is intelligent, there is no
specific activity he
need be good at ... being intelligent is not a matter of doing something ... [since] "Being intelligent" and
"being stupid" do notname actions or types of actions (Fodor, p. 72). He suggests that knOWing-how appears to have the character of a qualitatively different kind of knowledge only because humans have no conscious access to it: we must, subconsciously, rapidly process large amounts of knowing-that knowledge in every action and ability. Thus, having larger amounts of information in richer repre sentational schemes presumably addresses both problems of impov erished microworlds and the representation of commonsense know ing-how. This technique, however, has met with serious difficulties. In the attempt to "bolster" the information a script can contain, Traditional AI theorists hope to work upwards from isolated, constrained prob lem domains towards the world of human knowledge and experi ence. Yet, at every tum, more and more information must be explic it! y represented within the script. The magnitude of the project does not go unnoticed; Minsky later (1975) comments: Just constructing ~~ow ledge base is a major intellec tual research problem .... We still know far too little about the contents and structure of common-sense knowledge. A "minimal" common-sense system must "know" something aboutcause--effect, time, purpose,
Two ApPROACHES TO ARTIFJOAL INTELLIGENCE
locality, process and types of knowledge ... (Minsky, p.124). An obvious solution would be to construct a machlne which could move around in the world and learn to create its ownrepresen tations. Yet this approach has encountered a serious paradox. Richer knowledge representations require advances in robotic movement. vision and interaction to learn from the environment: yet such advances in robotics first require advances in knowledge represen tations in such fundamental areas such as representing the robot's ownbody, the solidity of objects, the effectsofmovement onperspec tive and more (WeD, p. 46). Dreyfus does not consider enriched representational schemes any kind of advance towards machine understanding at all. The problem lies in an unjustified belief con cerning human ability in the world: why would one consider, as Minsky and Fodor do, the explicit representation of human practices to be formalizable? This makes sense onIy in the context of the highl y constrained microworld in which a program operates, and reveals serious discrepancies between microworlds and the real world of human experiences. Indeed, in ananalysis of the attempt to represent the knowledge of even a small part of the world we live in, Dreyfus concludes that microworlds are completely unlike the human expe rience of the world. He suggests that we may work in subworlds, such as the university or the theater, but they are not related to each other in an isolated mode of "composing" a larger, shared world as microworlds are. Human subworlds instead presuppose a larger unified whole, and work as local elaborations of it (WeD, p. 14). The attempt to gain machine understanding through enriched representational schemes of the world has, so far, met with failure. Yet Dreyfus' arguments have indicated an even deeper problem with the approach: the question does not concern the degree of complexity a representational scheme must have, but rather whethe r human understanding involves representations at all. To further develop this, I turn to Heidegger's analysis of human understand ing. Dreyfus' criticisms of AI's use of micro worlds, and his concern for the humancontext in general, can clearly be traced to Heidegger' s analysis of human existence (Dasein) Being-in the world. "Being-in" conveys a sense of "in" entirely different from the way objects may
be "in" other objects. A sphere, for example. may be "in a box" in the sense thatit is surrounded on three or four sides; but this sense of "in" is an unengaged one: the sphere, Heidegger says, is really "along with" the box (Heidegger, p. 79). Humans, on the other hand, are very different: we are engaged in the world; we dwell in a familiar and involved way in it. Heidegger notes "there isno such thing as the 'side-by-sideness' of an entity called 'Dasein' with another entity called 'world'" (Heidegger, p. 81) to emphasize that Being-in is not like an "object inside an object." Indeed, the world is not a thing at all, nor is it a composition of things. Instead, the world is a context, a background for which entities have always already been in. Entities in the world can be encountered by Dasein in two ways. In use, an object is ready-to-hand (Heidegger, p. 98). Heidegger's examples of ready-to-hand entities typically involve skilled activi ties, such as hammering. The hammer, when actively used, is unno ticed: "an entity of this kind is not grasped thematically as an occur ring Thing" (Heidegger, p. 98). Thematic grasping of an entity qua entity requires detached contemplation, a way of revealing objects as present-at-hand. Thus, a hammer could be revealed as present-at hand-if it is sitting on a table and Dasein is analytically examining it, or if its head suddenly breaks when prying a nail and Dasein attempts to repair it. But typically entities are known in their use, as ready-to-hand. The distinction between use and detached contemplation clearly corresponds to the distinction between knowing-how and know ing-that. Heidegger's analysis, which shows that Dasein is always in-the-world, sets the ready-to-hand encountering of entities as the fundamental, typical way Dasein understands. This understanding is knowing-how-encountering an entity as ready-to-hand in its use. Revealing an entity as present-at-hand in detached contempla tion yields an entity qua entity. This is knowledge of the knowing that sort. concerning facts and information about objects distinct from Dasein. Thus Dreyfus notes that Being-m-the-world cannot be understood solely on the model of a relationship belween subject and object, because such amodel dpes not account for understanding a thing as ready-ta-hand (BW. p. 45). Heidegger, also refuting the subject/object model of understanding. insists that "... the perceiv ing of what is known is not a process of returning with one's booty to the 'cabinet' of consciousness after one has gone out and grasped
Two ApPROACHES TO ARTIFICIAL IN'IELllGENCE
it ..." (Heidegger, p. 89). He refutes the traditional representational theory of mind, which hold that we form meaningful mental repre sentations of the world and manipulate them when thinking. Heidegger does not deny the possibility of mental phenomena: he does, however, reject the idea that such phenomena create "internal meanings" of the world. Traditional AI has had, as its primary focus, an analysis of the way humans-as-subjects "grasp" objects in the world and interpret them in an internal, mental sphere. This is the attempt to analyze understanding as a collection of knowing-that knowledge. But this "knowing" is only knowing the present-at-hand: it involves sym bolic representations of the world and the rules needed to meaning fully manipulate them. It completely neglects understanding as primarily understanding entities as ready-to-hand. Fodor's earlier argument that intelligence is not a skill like hammering makes this fundamental mistake. This is the attempt to formalize understand ing as something distinct from the way Dasein is in-the-world-an impossible project, since the world revealed ready-to-hand cannot be represented by a set of context-free elements. The use of a hammer, for example, is nested in the context of a human social world with purposes and roles, which need not be represented as a set of facts (MVB, p. 29). Formalizing understanding to gain commonsense knowledge is at an impasse because Heidegger's commonsense understanding everyday know-how-does not consist of procedural rules, but rather an unformalizable knowing-what-to-do in everyday situa tions (MVB, p. 33). Dreyfus suggests that a child comes to know what-to-do by constant exposure to the world, and that "the same might well be the case for the social world. If background under standing is indeed a skill and if skills are based on whole patterns and not on rules, we would expect symbolic representations to fail to capture our commonsense understanding" (MVB, p. 33). The Heideggedan perspective provides a useful background to Dreyfus' criticisms of Traditional AI. The danger, however, lies in the ease at which one can overemphasize the holistic nature of human understanding. That "one cannot build up the phenomenon of world out of meaningless elements" (BW, p. 119) does not neces sarily imply that human understanding is dependent on the entire human culture, as Dreyfus does. To show this, I will first show how
NICHOLAS K. GRACILLA
the PDP app roach to AI satisfies Heidegger' s and Dreyfus' criticisms of Traditional AI, and then critically examine Dreyfus' "wholer than holism" criticisms. In theirmost general case, PDP systems are simulations of neural networks found in the brain. They consist of large numbers of individual processing units connected together in varying degrees of complexity. Individual units typically perform simple computa tions; they process information by sending excitatory or inhibitory signals to other units in varying degrees of intensity, dependent entirely upon the signals of the units simultaneously connected to them. Such a network will have two primary edges of multiple connection lines: the first can be considered as an input edge, where received information can pass through the network of connections and computation units to an output edge. A PDP network is not programmed with explicit rules nor does it create representations of the world to manipulate. Instead, a network is repeatedly exposed to "input" information concerning the world and "output" expected responses. By adjusting its internal connections, the networ k learns to associate the expected response to the situation. For example, one might "train" a network to predict weather patterns by presenting facts about the current weather conditions and what followed from them. The trained network could then associate similar future conditions to what had happened. More importantly, since there are no explicit rules concerning barometl'ic pressures, wind patterns, etc., the network can generalize to new conditions based on past experience. Salient features of new experi ences can be associated with past experiences, allowing for re sponses to conditions which the network had not seen before. The method of knowledge retention, too. is non-explicit and non-representational. Note well that Heidegger does not deny inter nal psychical entities or mental states-he merely denies that they are "internal meanings" or representations. The same holds true for PDP connectionist networks. In the context of weather recognition, the value of an individual node at some position is meaningless. No individual or group of nodes "represent" a rule which might state "if the barometric pressure is high, it is likely to be a nice day," nor is the value of some individual processing unit a meaningful representa tion of a feature in the world. Indeed, a trained network is only meaningful when considered
Two ApPROACHES TO ARTIFICIAL INTELUGENCE
as a whole in the context of the world. Its internal adjustments, are entirely dependent on the infonnation presented. No previously structured system akin to a script is used to deconstruct and process particular salient features of the problem. Instead, problem situa足 tions are presented to the networ k, which independently determines which are, and which are not, useful features. PDP networks are consistent with Dreyfus' and Heidegger's accounts in several ways. First and foremost, they do not create internal representations in the spirit of a representational theory of mind. Activity certainly occurs between nodes, but this activity cannot be meaningfully related to external phenomena. An indi足 vidual node's value is meaningful only in the context of the entire network. Secondly, the network is not independent of the context of the situation. Its only rule-which might be stated as "adjust to suit the expected response"-is entirely dependent on the information presented as well as the expected response. Any rule-like behavior which a network appears to follow can not be the result of the formation of rules, since rules cannot be represented in the network. Such behavior must be said to be emergent: a complex activity gained through the interaction of processing units which are not explicitly concerned with the more complex overall goal (Wallich, p. 128). The problem of machine representations of knowing-how appears to be readily addressed by the emergent quality of PDP systems. Just as, for humans, the acquisition of such knowledge requires repeated practice and development, parallel acquisition in connectionist sys足 tems require repeated training sessions. Dreyfus recognizes the compatibility of PDP systems with his and Heidegger's in-principle pOSition concerning knowing-how representations and Being-in-the-World. Yet Dreyfus is still critical of network systems. He comments "Intelligent behaviol' requires as a background the totality of practices whichmake up the human way . of Being in the world ... [yet] the capability for providing such a background is, at present, beyond the horizon" (FMK, p. 132) and" ... human beings are much. more holistic than neural nets. Intelligence has to be motivated by purposes inthe organism and goals picked up by the organism from an ongoing culture" (MVB, p. 39). It is this increasing dependence on an argument based on more and more holism which I find incompatible with his earlier, consis足 tent views. We say of a child, who clearly has not gained the "totality
of practices which make up the human way of Being in the world," that she is still intelligent despite this deficiency. On D:ceyfus' ac足 count, at what point could we determine that a person was intelli足 gent? How many human social practices would one have to know? It would be absurd to think that traveling to a country whose social practices were not known in their entirety would render a person unintelligent. Increased holism appears to be a quantitative argument. Yet Dreyfus refuted Traditional AI's attempt to use a similar argument. At that point, the Traditional AI approach was to include more and more information into highly structured representational systems. Dreyfus had shown that the quantity of information a Traditional AI system had was irrelevant: its representations had no know ledge of the ready-to-hand. Yet now that PDP systems meet such criteria, Dreyfus reverts to the quantitative argument he refuted: a PDP system must now have access to an entire human culture, with innumerable goals and purposes. I donot believe that PDP systems are the final answer to the many questions involved in modeling human intelligence. Yet I have shown the significant advances they do offer: networks are capable of exhibiting the non-formalizable behavior both Dreyfus and I believe are vital to human understanding. This capability renders Dreyfus' in-principle argument against them inapplicable.
[FMK] Dreyfus, H. L. "A Frameworkfor Misrepresenting Knowledge." cf. M. Ringle, ed. Pltilosophical Perspectives in Artificial Intelligence. New York: Humanities Press, 1979.
[BW] Dreyfus, H.L. Being-in-the- World: ACommentaryon Heidegger's Being and Time, Division I. Cambridge: The MIT Press, 1991. [MOM] Dreyfus, H.L. and S.B. Dreyfus. Mind Over Machine. New York: The Free Press, 1986. [MVB] Dreyfus, H.L. and S.B. Dreyfus. "Making a Mind Versus
Two ApPROACHES TO ARTIFICIAL INTELUGENCE
Modeling the Brain: Artificial Intelligence at a Branchpoint." d. Daedalus, (winter 1988): 15-43. [WCD] Dreyfus, H.L. What Computers Can't Do. New York: Harper and Row, 1979. Fodor, J. Representations: Philosophical Essays on the Foundations of Cognitive Science. Cambridge: The MIT Press, 1981. Heidegger, M. Being and Time. trans. John Macquarrie & Edward Robinson. San Francisco: Harper & Row, 1962. Minsky, M. "A Framework for Representing Knowledge." d. J. Haugeland, ed. Mind Design. Cambridge: The MIT Press. 1975. Newell, A. and H. Simon. "Computer Science as Empirical Inquiry." d. J. Haugeland, ed. Mind Design. Cambridge: The MIT Press, 1976. Ryle, G. The Concept of Mind. Great Britain: The Mayflower Press, 1949. Schank, RC. and R.P. Abelson. Scripts, Plans, Goals, and Understanding. Hilladale: Lawrence Erlbaum Assoc., 1977. Wallich, P. "Silicon Babies." Scientific American, (Dec. 1991): 124-134.
How NOT To READ RORTY David E. leBoeuf
University of Massachusetts. Amherst
One often hears talk that contemporary philosophy, in the most general sense, is broadly divided into two rather adversarial camps that of the "analytic" tradition and that of the "Continental" tradition. The former termusually refers to those philosophers concerned with logical analysis and conceptual clarity, philosophers who have inher ited many of the concerns of the Vienna Circle and who continue to try and formulate a successful post-positivist epistemology. TIle latter term usually refers to the more speculative, metanarrative oriented philosophers such as Heidegger and Foucault, who seek to radically historicize all human inquiry, particularly Western science and the various attempts by many philosophers to secure some form of first-order privilege for such scientific activity. In recent years, much has been written not only about the split between analytic and Continental philosophy, but also about closely related issues such as "modernism versus postmodernism," "objec tivism versus relativism" and Richard Rorty's schema of "foundationalism" versus "antifoundationalism,"l but for readers unfamiliar with the complexities of all these competing schools of philosophy, and also with the nuances that existinside the logics and doctrines of each of the major individual thinkers, grouping such a diverse array of figures and theories under a simple "analytic/ Continental" or "objectivist/relativist" dichotomy can often be more of a disservice than an elucidation. As undergraduates becoming initiated into the rich and detailed history of, for example, epistemology, we are bound at some point to oversimplify matters. For instance, with regard to how some may interpret Rorty's wholesale attack upon the epistemological tradi tion, one undergraduate philosophy instructor has remarked:
LeBoeuf is a s8nior philosophy Imd economics double major altelldil1g the Ulliversity of Massachusetts at Amherst. A resident afThree Rivers, Massachusetts, lie plans ta pursue a doctorate in phiIosopTlY. I See, for instance, Bernstein, Beyond Objectivism and Relativism: Hollis und Lukes; Margolis; and Putnam, Reason, Truth and History.
E isteme • Volume IV • May 1993
Teaching Rorty is difficult. Students respond favor ably, but superficially, to his critique. They consider it iconoclastic and exciting, but few of them have had the time to feel the grip of what he rejects. They may appreciate in an abstract way that it is unproductive to do epistemology but few canfeelliberatedby Rorty' s critique because they have not been captives of Bernstein's "Cartesian Anxiety" (Prado: d. Malachowski, pp. 365-6). Depending on whatever vague and underdeveloped theoretical orientation we uphold, we may be hasty to stereotype others' theo ries as "Platonist" or "relativist," or we may hold a set ofbeliefs which inhlndsight are blatantly inconsistent. It is my belief that much of the precociousness of our formative undergraduate years can be attrib uted to our misreadings of (and natural inability to yet adequately digest) various metanarrative "histories of philosophy"-i.e., those works which attempt to situate 2000 years of philosophy into a simple and neat framework. From Nietzsche' s Genealogy ofMorals to Reichenbach's The Rise ofScientific Philosophy, such texts, when read by individuals without adequate training in philosophy, are bound to receive rather perverse interpretations. And it is not just first-year philosophy undergraduates who are guilty of such interpretations. If it is, in fact, the case that many philosophy undergraduates engage in certain types of simplisms at early stages, then for obvious reasons, the same can be said for individuals in other academic disciplines who "dabble in philosophy." 2 For instance, the influence of Derrida in English departments and Kuhn in history departments is quite significant. My point is not whether some philosophy undergradu ates are wishy-washy, but rather that we, who have noticed signifi cant changes in our own philosophical perspectives, use the lessons we have learned about ourselves to stem similar misinterpretations by others, especially as certain philosophical doctrines with poten tially radical consequences gain in popularity amongst individuals outside of p,rofessional philosophy.3 The eclectic and controversial "pragmatism" of Rorty stands as a 2 See, for instance, Bernstein, Beyond Objectivism and Relativism; Hollis and Lukes; Margolis; and Putnam, Reaso1t, Truth and History. 3 I have in mind here underdeveloped interpretations of thinkers such us
How NOT To READ RORTY
prime example of a metanarrative-oriented philosophy capable of easy misrepresentation. Cited as anauthoritative philosophical source by numerous individuals in the humanities and social sciences (but also scathingly attacked by a multitude throughout academia), Rorty is arguably the most talked about philosopher in the western world. 4 Since the publication of his landmark book, Philosophy and the Mirror of Nature, the language of Rorty's subsequent essays has become increasingly accessible. Rorty's lucid writing style, with his injec tions of humor and his continual abandonment of technical analytic terminology, make him increasingly readable, while his ambitious attempts to synthesize a wide range of writers, both teclmical and literary, give his call for the end of philosophy a certain authoritative and rhetorical, if not seductive appeal to many non-philosophers. It has been noted by Alan Hobbs that:
The Mirror of Nature is designed as an exorcism of ghosts. For students to profit from the book, their minds must therefore first be haunted. Without suit ably muscled ghosts with which to do battle, the excitement of the exorcism is missing (I-Iobbs: d. Malachowski, p. 366). Without "suitably muscled ghosts," the excitement may be missing but a real danger is quite present-namely, a grossly underdevel oped (or just plain wrong) interpretation of Rorty and all the unfore seen consequences that go with it. By briefly examining the core of Rorty's philosophy and by reflecting upon a few of the prevailing criticisms of his work, I hope to demonstrate (however indirectly) the potential that Rorty' s mass appeal may have in inadvertent! y fueling naive versions of relativism and irrationalism. This concern is com pounded tremendously, of course, if Rorty's philosophical pOSition canbe shown to be ultimately incoherent and unworkable, as many Nietzsche, FOLlC<lult, Heidegger, Denid<l, Kuhn, Rorty and various nea-Marxists. Though I focus in this paper on the philosophy of Richard Rorty, the S<lme rule of thumb would <lpply equally well to, for example, Chomsky or cert"in sociobiolo gists. 4 For <l free-market economist' Stotal embrace of Ror ty' Sphilosophical position, see McCloskey; for <l similar embrace of Rorty's philosophy by two neo-Marxist economists, see Resnick and Wolff; for an English professor's complete dismiss" I of Rorty's view of language, see Steinmann.
of today's most prominent philosophers maintain. I
What turned so many heads with the 1979 publication of Philoso
phy and the Mirror a/Nature was the pronouncement, by a very able philosopher not only trained in the rigors of contemporary analytic philosophy but also conversant with the major Continental thlnkers of the day, that philosophy itself, as it has been traditionally con ceivedof,hadrunits course. As is well known, this is not the first time that someone had proclaimed the end of philosophy, but what is striking abou t Rorty is his status of being an "insider" to the technical arguments of cutting-edge analytic philosophy who is also sympa thetic to some of the most controversial ideas that have emerged from the Continental tradition in philosophy. Rorty does not, how ever, see himself as "taking sides" in either the analytic / Continental or objectivist/ relativist split. He is equally critical of thinkers onboth sides of the Atlantic who subscribe to what he sees as an outmoded philosophicalenterprise-thatof "epistemological foundationalism." Rorty applies a grid of"foundationalism versus antifoundaticn alism" to 2000 years of Western philosophy, and he uses this grid to divide the history 0 f philosophy into those who believe that genuine foundations to knowledge can, in principle, be discovered or expli cated, and those who believe that all knowledge (including attempts to legitima te "genuine" know ledge) is his torically contingent. Under Rorty's schemata, foundationalism has been, and continues to be, all the various systematic attempts "to underwrite or debunk claims to knowledge made by science, morality, art or religion" (PMN, p. 3). Foundationalist philosophers belieye that by correctly spelling out the nature of the human mind and its relationship to what is outside the mind, we canhave knowledge of reality as it exists in and of itself. Rorty characterizes foundationalist philosophers' main conviction: To know is to represent accurately what is outside the mind; so to understand the possibility and nature of knowledge is to understand the way in which the mind is able to construct such representations. Philosophy's central concern is to be a general theory of representation, a theory which will divide culture
How NOT To READ RORTY
up into the areas which represent reality well, those which represent it less well and those which do not represent it at all (despite their pretense of doing so) (PMN,p.3). Such conviction is, according to Rorty, the central tenet that unifies all foundationalist philosophers-be they analytic or Continental. One can turn to nearly any page (if not any paragraph) of a Rorty article and hear the same leitmotif, though perhaps played in differ ent keys: epistemology-centered philosophy goes back to the Pla tonic urge to find universals, and though no one still believes in Platonic ideas, we have nonetheless inherited the idea that true knowledge can be ascertained-that, in one form or another, lan guage or the mind acts as a "mirror of nature." Universality, neces sity, rationality, correspondence theory, representationalism, objec tivism, cOgnitivism, essentialism, logocentricism, structuralism, re alism-all these terms (as well as others) are, ifwe are to follow Rorty, members of the same family, as they all presuppose the feasibility of the epistemological project. Wittgenstein, Heideggel' and Dewey are the figures whom Rorty regards as "the three most important philosophers of our century" (PMN, p. 7), fOl' each was, in his early years, a foundationaHst trying to ground knowledge systematically-but in time, [e]ach of the three came to see his earlier effort as se 1f deceptive, as an attempt to retain a certain conception of philosophy after the notions needed to flush out that conception (the 17th-century notions of knowl edge and mind) has been discarded ... , Thus their later work is therapeutic rather than constructive, edifying rather than systematic, designed to make the reader question his own motives for philosophizing rather than to supply him with a new philosophical pro gram (PMN, pp. 5-6). II
In an early favorable review of Rorty's Philosophy and the Mirror of Nature, Richard Bernstein notes that some readers of Rorty who
"may not be acquainted with the latest subtleties" of issues and arguments in analytic philosophy may have felt that somehow philosophy tooka wrong tumwith the analytic movement. Theymay feel some satisfaction that Rorty has written the type of critique that could onlybe writtenby an "insider," and that he has shown that the emperor has no clothes-or at least is scantily clad. If only Anglo-American phi足 losophershad taken a different turn ... then we might have avoided the tangled mess which has consumed so much technical competence. But if this is the way they have read Rorty, then they have misread him and they have missed the real sting of his critique (PCM, pp.38-9). Rorty, after all, is formulating his ideas within the tradition of analytic philosophy-using the techniques and arguments of ana足 lytic philosophers to buttress his metanarrative hopes. (Whether his moves are valid is, of course, an entirely different issue). Rorty describes the antLfoundationalist "therapy" he is offering as "para足 sitic upon the COllstl'Uctive efforts of the very analytic philosophers whose frame of reference I am trying to put in question" (PMN, p. 7). I-Ience, most of ROfty'S particular criticisms of various attempts at foundationalism 4 are, in a sense, borrowed from various analytic philosophers such as Quine and Davidson: I am as much indebted to these philosophers for the means I employ as I am to Wittgenstein, Heidegger and Dewey for the ends to which these means are put. I hope to convince the reader that the dialectic within analytic philosophy ... needs to be carried a few steps further. These additional steps will, I think, put us in a position to criticize the notion of "analytic philoso足 phy" and indeed of "philosophy" itself as it has been understood since the time of Kant (PMN, pp. 7-8).
'For Rorty. the foundationalist urge took its distinctively modern turn with Locke, Descartes and especially Kant. See "The World Well Lost."
How NOT To READ RORTY
The question we must ask ourselves is precisely how many "steps further" we have license to take withoutbecoming incoherent. Rorty's radical antifounda tionalism leaves absolutely no room what soever for even a trace of the idea that "the world," in some way and to some degree, "determines" our knowledge. According to Rorty, any such hope can only be Kant's "noumena" dressed up inmodem garb. "If you start out with Kant's epistemology," writes Rorty, "in short, you will wind up with Kant's transcendental metaphysics" [WWL, p. 16). Rorty sees Kant's attempt to find apriori principles and structures (that Kant believed were presupposed for genuine knowl edge-especially with regard to the success of science) as the imme diate forerunner of analytic philosophy and the general fmm of its traditional concern with epistemological foundations. Rorty is pre pared to classify today's sophisticated versions of scientific realism5 as no more than latent Kantianism, pure and simple. Adopting an argument of Davidson's, Rorty writes: All that can be done with the claim that "only the world determines truth" is to point out the equivoca tion in the realists' own use of 'world.' In the sense in which" the w o rid" is just whatever tha t vast majority of our beliefs not currently in question are currently thought to be about, there is of course no argument ... The notion of "the world" as used in a phrase like "different conceptual schemes carve up the world differently" must be the notion of something com pletely unspecified and unspecifiable-the thing-in itself, in fact (WWL, p. 14). Hence, everything from the more explicit versions of scientific realism to Quine's "naturalized epistemology" (which reserves a special determinate role for stimuli-triggered "observation state ments") depends ultimately on what amounts to a Kantian defense. As with most philosophers he chanicterizes as "foundationalist," Rorty portrays Quine in a pejorative manner, arguing that 5 For lending defenses of scientific realism, see for instance, the workofRichm:d Boyd, Larry Laudnn, Ernnn McMullin nnd Clnrk Glymour. Rarty's most direct address to these and other renUsts is his essay, "Is Natural Science a Natural Kind?"
other philosophers followed Quine in falling back into dogmatic metaphysicS, decreeing that the vo cabulary of the physical sciences "limns the true and ultimate structure of reality." It is significant that Quineconc1uded that "the unit of empirical inquiry is the whole ofscience," when one might have expected, given the drift of his argument, "the whole of cul ture." Quine, and many other holists, persisted in the belief that the science-nonscience distinction some how cuts culture at a philosophically significant joint (NS, pp. 46-7). One can still, however, subscribe to a holism that does regard "the unit of empirical inquiry" as "the whole of culture" and yet not be
forced to accept Rorty's conclusions. One such holism is that of Davidson,6 who notes his point of departure from Rorty: Wherewe differ, ifwe do, is onwhether there remains a question how, given that we cannot "get outside our beliefs and our language so as to find some test other than coherence", we nevertheless can have knowl edge of, and talk about, an objective public world which is not of our own making. I think this question does remain, while I suspect that Rorty doesn't think so (Davidson, p. 123), Hilary Putnam's "internal realism," which he backs up with the notion of a transcendental rationality. is another holistic line we can take that is very different from Rorty's. As a causal theory of reference, it is one of the last foreseeable defenses of the very notion of "reference" (that there is some determinate linkage between "our world" and "the worId"). Putnam counts Rorty as a "cultural relativ 6 Davidson's philosophy resls heavily on Quine' 5 theoryof meaning. If Qui ne' s indeterminacy of translation thesis ultimately has no empirical explann tory power (since different translation manuals, or different "conceptual frameworks," all square with the stimuli of "the world") then Davidson's argument against "the very idea of a conceptual scheme" has even less empirical explanatory power, and consequently (and in a rather paradoxical way) seems to square ever more so with common-sense realism (ala Tarski). One then wonders how Rorty feels he has the license to deny (or at least qualify in a very strange way) the primary roles of "reference" and "truth" and the concept of reason they presuppose (PFD).
How NOT To READ RORTY
ist" because of Rorty's insistence that convincing one's peers is all there is to the game of "truth." In the Peircean tradition, Putnam writes: What I am saying is that the "standards" accepted by a culture or a subculture, either explicitly or implic足 itly, cannot define what reason is, even in context, because they presuppose reason (reasonableness) for their interpretation..., Reason is ... both immanent (not to be found outside of concrete language games and institutions) and transcendent (a regulative idea that we use to criticize the conduct of all activities and institutions) (Putnam, p. 234). According to Rorty, using the term "reason" (with its implicit notionofanidealconvergencepointof"truth")isreallynomorethan convincing others to accept your assertions-there is no role being played by "the world," as the term is really just a confusion with language. Putnam, however, stresses that we cannot appeal to a consensus definition of reason "because consensus among grown足 ups presupposes reason rather than defining it (Putnam, p. 240). Reason, as Putnam describes it, is normative. ASCribing to it a transcendentalnalurecan be viewed as a Kantianmove,butsowhat ? One will only see this as something "bad" if one subscribes to Rortyism, which (given the ideals, scientific achievements and "foundationalist" accounts of the Enlightenment) certainly must bear the burden of proof. III This brings us to the issue of whether Rotty is a relativist. But first we must be eleen' about what we mean by the term "relativist." In its most ancient form, relativism is Protagoreanism, the self足 contl'adidory view that accompanies a s tatement such as, "there is no truth." How are we to judge sucha statement? For if it were true that "there is no truth," we would have a reflexive paradox-as there wouldbe at least one true statement, namely, that "there is no truth."7 7 For a nice introduction to the self-referential problem and discussion of how it affects the philosophies of Nietzsche, Heidegger and Derrida. see Lawson.
Responding to anticipated charges that he subscribes to this type of naive relativism, Rorty writes: "Relativism" is the view that every belief on a certain topic or perhaps about any topic. is as good as every other. No one holds this view. Except for the occa sional cooperative freshman, one cannot find any body who says that two incompatible opinions on an important topic are equally good .... If there were any relativists, they would, of course, be easy to refute. One would merely use some variant of the self referential arguments Socrates used against Protagoras. But such neat little dialectical strategies only work against lightly-sketched fictional charac ters. The relativist who says that we can break ties among serious and incompatible candidates for be lief only by "nonrational" or "noncognitive" consid erations is just one of the Platonist or Kantian philosopher's imaginary playmates, inhabiting the same realm of fantasy as the solipsist, the skeptic and the moral nihilist (PRI, pp. 166-7). But if naive Protagorean relativism8 is an "imaginary playmate" for "Platonist or Kantian philosophers," cannot we turn the tables on Rorty. arguing that "naive Platonism" (as Rorty often seems to caricature most of Western philosophy) is Rorty's own imaginary playmate? Charles Taylor writes of the certain suppositions [that] seem to be made in the various invocations of [Rorty' s] argument: that the only candidate for a general account of truth is in terms of correspondence; that correspondence is to be understood in a rather simpleminded way. approach ing at times a picture theory; that believers in the correspondence theory are Raving Platonists. Under lying all of this is a continuing imprisonment in the 8 Joseph Margolis warns us that, "relativism should not be construed as, or as reducible to, any form of skepticism, cynicism, nihilism, irrationalism, anarchism or the like, although it may be that a well-defended relativism would lend comfort to doctrines of these sorts" (67).
How NOT To READ RORTY
model basic to the whole epistemological tradition which understands thinking in terms of representa tion.... By Raving Platonism, I mean the view that Rorty often invokes to ridicule his adversaries, such as that "the final vocabulary of a future physics will somehow be Nature's Own," or "that a vocabulary is somehow already out there in the world." We should consider this just as a rhetorical flourish. Rorty can't really believe that hard-faced scientific realists who think that mechanistic materialism is literally true, subscribe to Raving Platonism (Taylor, pp. 268-9). In a similar vein, Richard Bernstein (whose criticisms of Rorty have become harsher and more pointed over the years) writes: Why does Rorty think that philosophy (or "Philoso phy") amounts to little more than the worn-out vo cabulary of "bad" foundational discourse? So much of his recent writing falls into the genre of the "God that failed" discourse. There seems to be something almost oedipal-a form of patricide-in Rorty's ob sessive attacks on the father figures of philosophy and metaphysics (OSF, p. 557). It certainly does seem that Rorty paints far too much of a human face
on the West's apparent predisposition to engage in epistemological foundationalism. Rorty is an intellectual historian and, as such, he is prone to "lay the blame" of our foundationalist urges (if not explic itly, then implicitly) on Kant, and before him, Descartes, going all the way back to Plato. 9 Rorty's d1aracterization of how we arrived at our present modes of scientific and philosophicalinquiry appears much more black and white than Peirce's "cable versus chain" account of knowledge, where instead of picturing our reasoning as a linear and isomorphic "chain" from, say, "facts" to "states-of-affairs-in-the-world," we instead picture our fallible knowledge, at any given moment, as the many different types of evidence, modes of argumenta tion, hunches, 9 Cornel West believes tha t Rorty' s "thin historicism rests can ten t with intellec tual historical narratives and distrusts social historical narratives ... his narrative needs a more subtle historical and sociological perspective" (West, p. 270).
etc., that, individually, are only weak strands, but. collectively. form a strong cable. We never have knowledge that perfectly corresponds to "the world," but "the world" does guide, in a certain determinate sense, where our inquiry moves. Of Rorty's caricature of Western philosophy as being simply a series of illusory "chain" pictures, Bemstein echoes Taylor: According to this story. the real villain is Plato-at least the Plato identified (mythologized) by Platonism ... Iwant to maintain that this narrative is itself rapidly becoming a blinding prejudice that obscures more than it illuminates. What was once a stinging critique is becoming a bland, boring cliche. One begins to wonderifthereeverwasa "foundationalist"thinker at least one who fits the description of what Rorty calls "foundationalism." Even Plato-the Plato of the Dialogues-fails to fit this description ... Rorty's char acterization and caricature of the history of philoso phy is rapidlyrunnmg itself into the ground ... what is now needed is to demythologize this narrative of the invidious fallelmess of Platonism. For it is only to the extent that we still accept some version of Rorty's mythologizingaboutwhat philosophy and metaphys ics are, and what "philosophic justification" must be that his playful skepticism has any sting.... It is time that Rorty himself should appropriate the lesson of Peirce, "Do notblock the road to inquiry ..." (OSF, Fp. 558-60). Bernstein, like Rorty, believes it is time to go "beyond objectivism and relativism," that for too long we have been suffering from "Cartesian Anxiety"-Le., the fear that unless we emerge from our Cartesian skepticism with some kind of secure and objective Archimedean point for our knowledge, we will be in a state of epistemological nihilism, having no objective reasons for the beliefs we hold to be true and perhaps not even knowing what to believe. Bemstein sees traditional epistemology as offering us a false di chotomy. a false "Either lOr," to choose from-in which we eithel' believe in objective and stable bedrock foundations for our knowl edge, or else everything is relative. But Bernstein sees in Rorty yet
How NOT To READ RORTY
another grand and equally untenable Either JOr-that of foundationalism versus antifoundationalism. It is, after all, only a grid, only one of the many possible ways to quickly summarize 2000 years of human inquiry. As ametanarrative, it is hard to see how Rorty's philosophy can escape rela tivism and, ultimately, a forced position of irrationalism. And, as Steinmann notes, "the intellectual price of accepting it is far too high, and, worst, it is incoherent ... Why should we renounce the Enlightenment and its (still quick) heirs for unspecified benefits" (Steinmann, p. 47)? Despite the controversies, counterarguments and uncertainties that surround Rortyism, he seems to be gaining in popularity, particularly with disciplines outside of professional philosophy and with the lay public. Any philosophy student who, at one time or another, found him or herself uncritically swallowing (hook, line and sinker) a doctrine such as Rorty's (and I count myself as one) must take it upon him or herself to engage others in exercises that will sufficiently muscle their ghosts and haunt their minds.
[BOR] Bernstein, Richard. Beyond Objectivism and Helativism: Science, Hermeneutics, and Praxis. Philadelphia: University of Pennsylvania Press, 1988.
[OSP] Bernstein, Richard. "One Step Forward, Two Steps Backward: Richard Rorty on Liberal Democracy and Philosophy. " Political
Theory, Vol. IS, No.4 (1987): 538-563. [peM] Bernstein, Richard. "Philosophy in the Conversation of Mankind." Philosophical Profiles. Richard Bernstein. Philadelphia: University of Pennsylvania Press, 1986. Davidson, Donald. "A Coherence Theory of Truth and Knowledge." Reading Rorty. Oxford: Basil Blackwell, 1990. Hobbs, Alan. Letter. d. Alan Malachowski, "On Teaching Rorty." Reading Rort1j. Oxford: Basil Blackwell, 1990.
Hollis, Martin and Steven Lukes, eds. Rationality and Relativism. Cambridge: MIT Press, 1982. Lawson, Hilary. Reflexivity: The Post-Modern Predicament. LaSalle: Open Court, 1985. Margolis, Joseph. Pragmatism Without Foundations. Oxford: Basil Blackwell, 1986. McCloskey. Donald. The Rhetoric of Economics. Madison: University of Wisconsin Press, 1985. Nietzsche, Friedrich. The Birth ofTragedy and the Genealogy ofMorals. New York: Doubleday Anchor, 1956. Prado, C. G. The Limits of Pragmatism. d. Alan Malachowski, "On Teaching Rorty." Reading Rorh). Oxford: Basil Blackwell, 1990. Putnam, Hilary. "Why Reason Can't be Naturalized." Realism and Reason: Philosophical Paper. Volume 3. Cambridge: Cambridge University Press, 1983. Reich.enbach, Hans. The Rise ofScientific Philosophy. Berkeley and Los .Angeles: University of California Press, 1951. Resnick, Stephen and Richard Wolff. Knowledge and Class. Chicago: University of Chicago Press, 1987. [NS] Rorty, Richard. "Is Natural Science a Natural Kind?" Objectivity, Relativism and Truth: Philosophical Papers, Volume 1. Cambridge: Cambridge University Press, 1991. [PFD] ROlty, Richard. "Is There a Prob lem with Fictional Discourse?" Consequences of PragmatiS11'1: Essays 1972-1980. Minneapolis: University of Minnesota Press, 1982. [PMN] Rorty, Richard. Philosoplly and the Mirror ofNature. Princeton: Princeton University Press, 1979.
How NOT To READ ROR1Y
[PRI] Rorty, Richard. "Pragmatism, Relativism and Irrationalism." Consequences of Pragmatism. Minneapolis: University of Minnesota Press, 1982. [WWL] Rorty, Richard. "The World Well Lost." Consequences of Pragmatism. Minneapolis: University of Minnesota Press, 1982. Steinmann, Martin. "Rortyism." Philosophy and Literature, Vol. 12, No. 1 (1988): 27-47. Taylor, Charles. "Rorty in the Epistemological Tradition." Reading Rorty. Alan Malachowski, ed. Oxford: Basil Blackwell, 1990. West, Cornel. "The Politics of American Neopragmatism." Post足 Analytic Philosophy. John Rajchman and Cornel West, eds. New York: Columbia University Press, 1985.
59 NAGEL, PHYSICALISM AND SUBJECfIVITY
The twentieth century has seen the demise of nonphysical theo ries ofmind. The Cartesian tradition of mind-body dualism accord ing to which "the natures ofmind and body are acknowledged to be not only diverse but even, in a manner of speaking, to be the contraries of one another," (Descartes, p. 8), today seems quaint to most thinkers. Although a few contemporary theorists continue to embrace Descartes' view that mind is a nonphysical substance that interacts with the (physical) human body,! arguments against dual ism "have moved most ... of the professional community to embrace some form of materialism" (Churchland, p. 21). "Materialism," or "physicalism," is the position that takes our modern scientific worldview seriously. The position assumes that humans fall unproblematically into the biological world of trees, flowers, fish, monkeys and amoeba. The biological world is in tum explainable in terms of the dances of elementary bits of matter and the forces which play roles in their interaction. "The message of the last 300 years of science is that ultimately we--and all else-are nothing but swarms of particles" (Sterelny, p. 2).2 Thus phYSicalism forces us to conclude that human minds-like human bodies-fall squarely in the domain of our physical theories. The phenomena we call "mental" are on an ontological par with the activities of squid, radia tion and bricks; the entities are different onIy in organization and complexity.3 However popular among contemporary philosophers and scienUttle is a junior majoring in philosophy alld psychology at Transylvania University. He plans to pursue a doctorate ill plJilosophy and teach at the ulliversity level. Namely John Eccles (1979, 1980) and Karl Popper (d. in Eccles 1980). One should remember. though, that the terms of physics change with empiri cal and tileoretical developments. (d. ].J.e. Smart: "By 'materialism' I mean the theory tilat there is nothing in the world over and above those entities which are postulated by physics (or, of course, those entities which will be postulated by future and more adequate physical theories)" (p. 159, italics mine). 3 But the "only" here should not be read as diminutive; these differences should not be taken lightly. The human brain is host to over 50 billion neurons and trillions of interneuronal connections. For more on the importance of the boggling power of the human organ of thought, see Dennett's "Fast Thinking" (IS, pp. 323--337). I
Episteme • Volume IV· May 1993
tists, the dominant strains of physicalism have been resisted by some non-dualist theorists as patently wrong or, at best, incomplete. 4 These philosophers fear that the materialist position leaves out something important about our nature and in so doing presents an incomplete picture of the world. These thinkers challenge the notion that contemporary science is able fully to explain the universe, for, they claim, the concepts ofmodem science are in principIe unable to subsume certain facts that we know or intuit to be true-facts which center on the mlnd-body problem in philosophy. This essay deals with a prominent exponent of this hesitancy toward physicalism, Thomas Nagel. In a series of essays and books, Nagel has developed his position around the issue of subjectivity. The problem is one of opposition between subjective and objective points of view. There is a tendency to seek an objective account of everything before admit足 ting its reality. But often what appears to a more subjective point of view cannot be accounted for in this way (MQ, p. 196). My work here focuses on evaluating the import for phYSicalism of Nagel's concerns about subjectivity. My treatment of the issue breaks into two tasks: first, identifying and describing the various senses of subjectivity that Nagel finds problematic for physicaJism, and second, sketching a possible salve for the quandary elicited by each sense. I hope in this way to answer the following question by the end of the paper: Does subjectivity pose an intractable problem for a
modern physicalist worldview? My approach relies on the notion that one can make progress on even seemingly big problems when steps are taken to identify manageabIe parts. By taking those steps in this paper, I hope at least to point the way toward an ultimate reconciliation between Nagel's notion of subjectivity and physicalism-a useful enterprise even if the reader fails to accept my own formulations for this feat.
4 I shall not deal here with contemporary dualists. I take their positions to be obviously untenable for reasons any cursory look at the materialist literature will reveaL
NAGEL, PHYSICALISM AND SUBJECTIVITY
The first sense: Intraspecies Subjectivity On Nagel's view, we miss something when we describe a person in the objective terms ofphysics. The thing that physical descriptions leave out is "the internal element ..."which "remains, even if ignored, as the true source of persistent dissatisfaction with all physical or other external theories of the mind" (MQ, p. 202). Nagel thinks that this internal feature of subjective experience "cannot be captured by the purified form of thought suitable for dealing with the physical world that underlies the appearances" (VN, p.lS). Subjective expe rience for Nagel includes such phenomena as "raw feels" and "inten tional mental states" (VN, p. 15). So, on a first pass, Nagel seems concerned about what philoso phers call "qualia."5 The debate about qualia-the intrinsic proper ties of our sense experience which physical descriptions somehow cannot capture-is familiar to all of us. In fact, many ofus discovered the "qualia problem" in childhood. If you ever wondered whether what you call "red" might be seen as "green" to me, internally, then you hit on the puzzle. Or, if you ever found yourself troubled that what I call "pain"mightnotfeel the same as what you call "pain," you were uncovering a similar difficulty. Even though we might point to the same things when we say "red," or we might say we feel "pain" in aU the same circumstances (perhaps whenever we have our teeth drilled), it seems impossible to say whether my "red" or "pain" is the same-has the same "qualia"-as your "red" or "pain." But if our qualia are not fixed by reference to our behavior, how can physics, the science of behavior par excellence, ever capture what is it to see "red" or feel "pain"? Even if we possessed full neurophysi ological descriptions of the brain states involved in the instantiations of these qualia, how would we say in physical terms what it was like to "see red" or "feel paln"? In fact this seems what Nagel has in mind in asserting tha t
5 Here I group "raw feels" and "intentional mental states" together. Strictly speaking, the term "quale" is used in the literature to refer only to "the intrinsic qualitative nature ... that is revealed in introspection" (Churchland, p. 24) of sensations. However, I think that much of what seems irreducible to physical terms in the notion of "intention" might be characterized in terms of qualia.
the subjective character of experience ... is not analyzable in terms of any explanatory system of functional states, or intentional states, since these could be ascribed to robots or automata that behaved like people although they experienced nothing (WLB, p.436). The subjectivity problem, then, is first of all an intraspecies problem: How can we say anything about the character of our subjective human mentalstates-like feeling pain and seeingcolor足 by using the predicates of physics? How could a description like "neuron 800,456 is firing- you are seeing red" capture anything close to what I mean when I say I see "red"? We could answer this question in at least two ways. A first possibility is to take the position that, because we cannot say much about qualia (try describing what "red" looks like), they prove too flimsy and informationally poor to matter to physics (or artificial intelligence, by the way). A subset of this view might hold that the idea of qualia as entities, as something real, is misguided-there really are no such things as "painness" or "redness" left over whenwe explain your "pain feeling" or "red feeling" in terms of your physical design and functioning (Dennett takes this view in his brilliant work on consciousness (eE, pp. 369-411損. These stances on quaHa are ingenious, plausible and perhaps true. But they are also counter足 intuitive and controversiaL Nagel seems correct in saying "our origi足 nal concept [of pain} already picks the thing out by an essential feature ..." (VN, p. 47) Ifhe is correct, then we must account for qualia in some way, and primajacie we might have to accept Nagel's "a priori da im that the mental cannot be reduced to or analyzed in terms of the physical" (VN, p. 48). As a second way of approaching the qualia problem, suppose that Nagel is right about the irreducibility of phenomena like "pain" and "redness." Does it follow ilia t physicalism is false or incomp lete? The answer seems to be no, and it is helpful to tum to Nager sown defirution of physicalism to see why: "I mean by physicalism the thesis that a person, with all his psychological attributes, is nothing over and above his body, with all its physical attributes" (P, p. 214). The tenability of the thesis seems to depend on the class of the "physical." If we could expand this class to include such things as
NAGEL, PHYSICALISM AND SUBJECTIVITY
"pains"and"redness,"wewouldbeabletoassertaphysicalismwhlch covered many of the "psychological attributes" Nagel worries about. And why not? Many phenomena these days are "physical" which our scientific ancestors would have rejected as patently non足 physical or even supernatural. Nagel himself notes this: Electricity and magnetism could not be analyzed in terms of mechanical concepts of matter in motion. ...The shift from the universe of Newton to the uni足 verse of Maxwell required the development of a whole new set of concepts and theories ... specifically devised to describe and explain these newly explored phenomena (VN, p. 52). Explaining qualia like pain and redness would require new equa足 tions which would explain under what physical conditions the qualia-themselves physical-would appear, but this should strike one as no more difficult than describing the phenomenon of, say, life. Ofcourse, it is a challenge, to understate things a bit, to describe what physical conditions produce life. But we have good theories in this arena. We know, for example, that mammals require oxygen, water and some form offood to supply the raw materials for Jiving; we also know much about the .processes through which animals use these . raw materials and excrete waste products. In addition, we possess elaborate cellular theories of the conception find development of these organisms. "But none ofthat explains 'life,'" rhear a critic say, "It just tells us the conditions for life." The last part seems to me a true statement, and exactly why science has (or has begun to) "explain" life. We "understand" life in a "scientific" way because we know what physical conditions must hold for a thing to live, just as we under足 stand ocean waves in a scientific way because we know what physical conditi.ons must be satisfied in order for frothy waves to roll onto the beach. One could claim that these "explanations" really do not get at the "oceanness" or "waveness" of the ocean waves, or the "Hfeness" of Hving things, but that would be to misunderstand the mission of science. Nagel even seems to see the possibility for admitting qualia into the world of physics, but he cryptically rejects it:
Some may think there is nothing to prevent mental phenomena from eventually being recognized as physical in their own right. ... It seems to me more likely. however, that mental-physical relations will eventually be expressed in a theory whose funda mental terms cannot be placed clearly in either cat egory (WLB. pp. 449-50). And what would these new fundamental terms be? Scientific revo lution after scientific revolution, we have not failed to call new concepts "physical"; why should mental phenomena be party to a different fate? I do not have the space here for a detailed discussion of the philosophy of scientific explanation. Instead. I offer at least a plau sible diagnosis ofNagel'shesitancywhenitcomes to "physicalizing" qualia. We tend to fall into what I call the "roll, bump. thud" characterization of physiCS. According to this view. the identifying characteristic of the "physical" is that it is made of "stuff"-stuff that goes "roll. bump, thud" or is composed of components which fit this description. Tables. bricks. water, atoms andbrains all£allneatlyinto this designation of "physical." But the problem is that "physics" seems to use a quite different conception of "the physical." Instead of the above commonsense notion of a "category of composition" (according to which "physical ness" is a matter ofw hat something is made of). physics seems to rely on a "category of predictability" when postulating or describing physical objects. Following the boundaries circumscribed by this category, physics admits entities into the category "physical" which best explain phenomena under observation. For example, X-Rays were posited as a physical entity before anyone knew "what they were made of" (better: "how to fit them into our present theoretical framework or how to change the framework to admit the new phenomena"). In this way. it seems that we might indeed admit qualia like "redness" and "pain" into our physical worldview. All we would need would be ways of ordering or systematizing the phenomena which we give these labels. Although this would perhaps be a humbling task, we already possess the beginnings of this theory (for example, I can say with certainty that the quale "pain" will take place
NAGEL, PHYSICALISM AND SUBJECTIVITY
in fellow humans with normal nervous systems if a lighted match is
held next to their skin. 6 Of course, what makes this an intraspecies problem is that we get the data of our new physical pain and redness theories by being members of a type of biological class the members of which possess similar physical structure. This similarity of physi cal structure allows us to generalize from our experience of "pain" to the quality of the experience of another member of the species. At this point, it might prove helpful to step back and summarize the analYSis so far. First I claimed that Nagel's problem with subjec tivity on the intraspecies level has something to do with the apparent futility of describing qualia in physical terms. Next I showed how even if we accept the commonsense notions of "pains" and "color qualities" as "real" instead of dismissing them as misguided or trivial-qualia can plausibly be "physicalized. "7 But if qualia pose no threat to physicalism, what is left of the intraspecies subjectivity problem? Something does seem to be left. After looking at my analysis of qualia, someone will say, "But surely my experience as a human is not composed just of qualia-these are too simple to describe the richness ofmy subjective experience. How could a physics-€ven one that admitted qualia-explain all that richness"? I will attempt to shake the intuitive significance of this residual richness by introducing two further concerns linked to intraspecies subjectivity: the first deals with the nature of human subjective ex perience, and thesecondisanotherpoint about scientific exp]ana tion. The first point begins wi th an explana tion of a phenomenon I cn11 the "intermodal fallacy."s Although it may be surprising, the degree to which our sense experience seems subjective is related to which Q I know this because even Nagel, who hilS doubts about the complete/JeSs of physicalism, can contradict physicalism only ilt the risk of absurdity. That is, he would not want to claim that structurally congruent physical entities subjected to identical stimuli could harbor significantly different reactions. To claim that they could would be to take stock in nn intervening variable-and whnt would it be? "Mind stuff" is not an a ttrnctiv€ answer for NageL 7We mightnotwantto take this route, of course-the demands mightprove too taxing. I just want to point out the possibility for this avenue, which is more than Nagel will allow. S Recently I found that Dennett makes an observation along the Silme lines (CE, pp. 380-81); however, he makes the point in passing and seems not to give the phenomenon the credit I think it deserves.
sense modality we talk about; some sensory channels seem to give us more "objective" data than others do. For example, imagine visiting an art gallery with a friend. On the wall hangs a large oil painting by Salvador Dali. You talk with your friend about the symbolism of the wilted trees and fanciful human-like figures, and discuss everything from the role of shading in Dali's presentation to the importance of the play of various types of angles in the work. But then your friend has an idea. "No one is around, so why don'twe see how the painting feels?" The ensuing conversation is markedly impoverished.
- "It feels ... well, it feels 'rough' here and 'smooth' here." - "No, it feels more 'grainy' than 'rough.'" - "Yes, well, how do you know what 'rough' feels like to me?" Far from treating Dali's "aesthetics of touch," your conversation
never gets off the ground. And, curiously, whereas your discussion about what you see runs only into problems of differences in artistic judgment, your discussion of tactile impressions turns into a full足
blown metaphysical quagmire; in fact, it is the problem of qualia all over again. One sense modality seems more objective than another, I buthowisthispossible? Are not all the senses on one ontological par? ' Why do they seem to admit different degrees of subjectivity? Many explanations suggest themselves, but I will deal only with two of the most obvious. First, human vision has a finer "grain" than human touch. Each normal human retina is host to about 132 million photoreceptors linked in an elaborate network of connecting cells (Foley and Matlin, p. 59). However, the human hand has the benefit of only several thousand receptors for touch (Foley and Matlin, p. 370). Second-and this is probably a corollary to the physiology of vision and touch-the human visual vocabulary seems much larger than the human tactile vocabulary. These possibilities corroborate a larger point: perhaps our sense of subjectivity exhibits an inverse relationship with the richness of the sensory data to which we have access. If this is the case, the intraspecies subjectivity problem might have to do more with weak sensory vocabularies and apparatuses than with an ontological dilemma. An analysis of the intermodal fallacy thus poses the following I
NAGEL, PHYSICALISM AND SUBJECTIVITY
interesting idea: if lack of a rich sensory vocabulary is much of the intraspecies subjectivity problem, we might be able to eliminate it altogether by admitting our sense qualia to the physical world and developing sophisticated new "vocabularies"-in the form of com足 plex descriptions of the physical conditions for the phenomena. This program would indeed seem possible. However, one still might object that this program of systematiz足 ing intra species qualia must in principle still miss much of the complexity of human perceptual experience. "Surely," one might say, "the richness of my experience is not captured in my vocabu足 lary-even in my relatively rich visual vocabulary." This critic would of course be absolutely correct, but with no Significant conse足 quence for a physical theory of perceptual experience; explaining this point requires a discussion of another interesting observation about scientific explanation. When it comes to talking about our perceptions, we seem to employ two senses of "experience," E(l) and E(2): E(l): A clear yet incomplete notion
E(2): A rich, complete, yet vague notion
When we look at a "red" car and see "redness," when we see a clock and notice that it appears "round," or when we hit our thumb with a hammer and feel "a throbbing pain," we are employing the concept "experience" in the first sense. Tha t is. what we notice and are able to say about the phenomenon is (relatively) clear, but does not exhaust the "wholeness" of the experience-we cannot seem to say enough to do justice to what we "experience" in the second sense of experience. E(l) is the sense of experience that the preceding analysis of qualia proposes to physicalize through gradual physical vocabulary building. E(2) is the sense of experience not fully covered by even the best physical explanation of the phenomena of perception. We should be ready to admit that experience in the E(2) sense actually exists;9 however, does this spell the demise of the possibility of a physical explanation of our experience? 9 But see Dennett (CE, passim) for plausible evidence thatwe are mistaken abou t the richness of out perceptual experience. Trying to concede as much as possible to the pretheoretical notion of experience, I do not argue Dennett's hard line here.
No. Not all physical objects are the objects of physics. As men足 tioned above, physics does not propose to explain the "oceanness" of ocean waves, or the "lifeness" of living things. If this were the case, science would have on its hands quite a task: in order to prove itself "complete" to its critics, it would have to describe the uniqueness of every possible combination of physical phenomena in the world足 everything from "a ruffle-feathered cardinal perched atop the oak writing desk in the second room on the left" to "the pitted plastic button on the shirt I wore when I visited the huge mansion on Fourth Street." Modem physical explanations of these examples-posed in terms of arrangements of atoms and forces which hold between them-do not describe the phenomena's richness, but this is some足 thing that we can live with. Physics attempts to explain phenomena in terms of their physical parts, not to describe them in their entirety. In fact, the simplicity of physical explanation seems to be one of its most powerful traits. Although rich, vague experience in the E(2) sense might exist, the fact that it is difficult for physicalist science to describe this experience takes nothing away from phYSicalism. Sense two: Interspecies Subjectivity
We have investigated a plausible route for the solution of the intraspecies subjectivity problem. Strangely, although Nagel seems at times apprehensive about such a physical, objective account of human mental states, at times he seems amenable to the idea. Infact, he even claims that There is a sense in which phenomenological facts are perfectly objective: one person can know or say of another what the quality of the other's experience is [But], They are subjective ... in the sense that even this objective ascription of experience is possible only for someone sufficiently similar to the object of ascrip足 tion to be able to adopt his point of view ... (WLB, p. 442).
Remember that the above solution for the intraspecies problem hinges on the fact that we can know about other members of our species' experience in virtue of our structural (physical) similarities.
NAGEL, PHYSICALISM AND SUBJECTIVITY
It seems that Nagel, despite his caveats about the problematic nature
of intraspecies subjectivity, has been ready all along to admit the possibility for physicalizing human experience in this way. Nagel's stronger case rests on the problem of interspedes subjectivity. Without similar physiological structures, it is impossible for one organism to know "what it is like to be" the other organism, in Nagel's terms. In perhaps his mostfamous article, "What is it Like to Be a Bat?" Nagel argues that because humans and other creatures足 bats, for example-have wildly different types of perceptual appara足 tus, humans cannot in principle know "what it is like to be a bat." From this observation, the philosopher goes on to argue that This bears directly on the mind-body problem. For if the facts of experience-facts about what it is like for the experiencing organism-are accessible only from one point of view, then it is a mystery how the true character of experiences could be revealed in the physical operation of that organism (WLB, p. 442). An explanation of bat experience seems to be aU that stands in the way of a full-blown physicalism, We can approach the interspecies problem in several ways, Perhaps the most plausible answer to Nagel's concern that physics cannot describe "what it is like to be a bat" is to say that physics actually does describe bat experience in a rough way, This approach is stressed by Dennett: [Physiological and behavioral] investigations would show us a great deal about what a bat could and could not be conscious of under various conditions, by showing us what provisions there were in their ner足 vous systems for representing this and that, and by checking experimentally to make sure the bat actu足 ally put the information tousein the modulation of Us behavior (eE, p, 444). Of course, one could read this approachlO as building directly on 10 Although Dennett would not-remember. he is not a "realist" about our pretheoretical notions of experience.
the framework for dealing with intraspecies subjectivity that we erected in the last section: because humans know what "sensing" is like-even if humans do not possess a sonar sensory modality humans can at least understand the rudiments of bat experience. In this way, humans can use the structural descriptions of bats pro vided by the sciences of biology and physics to extrapolate from human experience to bat experience. But, surely, we still do not know exactly what it is "like" to be a bat-evenifwe have a rough sense of what it mightbe like. But is this a problem for physicalism? Is there something in the bat point of view which physics-as we now understand it-is unable to ex plain? The answer is both yes and no. At first glance, there is nothing about physicalism that blocks it from including bat experience. In fact. deSCriptions of bat physiol ogy, as Dennett points out, are pretty good rough descriptions of what it is like to be a bat. However, there is still a difficulty, but the problem is nothing more (orlessl) than this: Ahumancannotbeabat. In the end, Nagel's argument against physicalism seems to turn on this one statement. Indeed, if one has to be a bat in order to know what it is like to be a bat, and there is no way for a human to be a bat, we are stuck. But this seems strangely tautological. Ofcourse a human cannot be a bat! Can a square be a circle? Can a mountain be a lake? Can a car be a train? In short, Nagel's observation that we cannot directly know "what bat experience is like" is presupposed by the very physicalism he thinks problematic. It is a commonly known axiom of physics (or of logic), after all, that one type of thing cannot be another at the same time; this is just obvious. Thus, when we stop to take a good look at Nagel's seemingly startling observation that in trying to imagine bat consciousness "I am restricted to the resources of my own mind, and those resources are inadequate to the task" (WLB, p. 439), we see both that this statement is tautological and that the best way to explain why our human
minds are inadequate is to use a physical, structural explanation. Conclusion: A Toast to Physicalism Through a careful analYSis of Nagel's arguments for the inad equacy of physicalism, we find that the "problem of subjectivity" involves not one but several problems, involving various assump
NAGEL, PHYSlCAUSM AND SUBJECTIVITY
tions about the character of perceptual phenomena, the claims of physicalism and the nature of physical explanation. We see that once our intuitions about the problem of intraspecies subjectivity are assuaged, the interspecies subjectivity problem seems either dimin足 ished or completely misguided. Thinking of our universe-all of it-in physical terms seems to hold the most promise for explaining even the most mysterious phenomena. I hope that this paper serves to help point the way for physicalism to surmount the difficulties posed by Nagel's analysis of subjectivity. Here's to the continued success of physicalist explana足 tions of the world. ll
Churchland, Paul M. Matter and Consciousness. Cambridge: The MIT Press, 1990. [eE] Dennett, Daniel C. Consciousness Explained. Boston: Little, Brown and Company, 1991.
[IS] Dennett, Daniel C. The Intentional Stance. Cambridge: The MIT Press, 1990. Descartes, Rene. Meditations on First Philosophy, Donald A. Cress, trans. Indianapolis: Hackett Publishing Company, 1979. Eccles, John C. The Human Mystery. New York: Springer Internati onal, 1979. Eccles, John C. The Human Psyche. New York: Springer International, 1980. Matlin, Margaret W. and Foley, Hugh Boston: Allyn and Bacon, 1992.
J. Sensation and Perception.
11 My thanks go to professors Jack Furlong, Rick O'Neil and Nancy Slonneger for generous use of their private libraries. '
[MQ] Nagel, Thomas. Mortal Questions. Cambridge: Cambridge University Press, 1979. [P] Nagel, Thomas. "Physicalism." The MindlBrain Identity Theory. C.V. Borst, ed. Bristol, Great Britain: Macmillan and Co., Ltd., 1970. [VN] Nagel, Thomas. The View From Nowhere. New York: Oxford University Press, 1986. [WLB] Nagel, Thomas. "What is it Like to Be a Bat?" Philosophical Review, 83 (1974), pp. 435-450. Searle, J o hnR The Rediscovery oftheMind. Cambridge: The MIT Press, 1992. Smart, J.I.C "Materialism." The Mind/Brain Identity Theory. CV. Borst, ed. Bristol, Great Britain: Macmillan and Co., Ltd., 1970. Sterelny, Kim. The Representational Theory of Mind: an Introduction. Cambridge: Basil Blackwell, 1990.
73 PROPER FUNCTION, RELIABILITY AND WARRANT
Jack C. Lyons
Warrant: The Current Debate and Warrant and Proper Function
constitute Alvin Plantinga's recent effort to refute contemporary theories of warrant and to establish his own. 1 The assault against the current theories in Warrant: The Current Debate is undertaken by using bizarre counterexamples such as the Case of the Epistemicall y Serendipitous Brain Lesion or of the Epistemically Inflexible Climbe l', as well as forays into Alpha-Centaurian worlds which include Cartesian demons turned Star Trek and middle-aged radioactive invisible elephants. The majority of these creative counterexamples are produced to show that the standard accounts of warrant fail because they wither when confronted with abnormalities and mal functioning cognitive faculties. While I am not in complete agree ment with Plantinga's consequent theory in Warrant and Proper Function, I do think that both his attack on the other theories of warrant and his own formulation are very illuminating and might pl'Ovide some useful insights into the nature of warrant. By discuss ing what I take to be the shortcomings of his proper function the~)I'Y' I hope to show the general direction in which we might find a fail' initial approximation of warrant. The resulting theory wi 11 be large Iy descriptive, rather than normative, but it is difficult to keep a strict distinction between the two in these matters. Plantinga's positive theory is laid out in Warrant and Proper Function:
a belief has warrant for me only if (1) it has been produced in me by cognitive faculties that are work ing properly (functioning as they ought to, subject to no cognitive dysfunction) in a cognitive environment that is appropriate for my kinds of cognitive faculties, Lyol1sis a selliol' philosophy major at Valparaiso UniVersity. A resident of Vnlparais(I, Illdialla, Jw will be attending The University of AdzOllfl il1 tlte fall wit/I tlw illtel/Moll of getti11g a Mctomte in philosophy. 1 Since they are still forthcoming, all citations will be of the form "WPF, lI.I,C, p. 31," denoting chapter, section, subsection and pllge number of the manuscript version.
F,11i.'lteme • Volum
(2) the segment of the design plan governing the production of the belief is aimed at the production of true beliefs, and (3) there is a high statistical probabil足 ity that a belief produced under those conditions will be true (WPF, II, II, p. 58). Although the above cited formulation of warrant states that I am warranted "only if" (1), (2) and (3), I think that the rest of the book makes it dear that he intends these criteria to benot only individually necessary, but also jointly sufficient (or very nearly so) for warrant. I do not think, however, that they are either individually neces足 sary nor very near Iyjointl y sufficientfor warrant. Criterion (3) nearly makes the first two seem superfluous, for as long as there is a high probability that the belief-producing mechanism functioning under a certain set of circumstances yields true beliefs, what difference could it make whether they are in the right environment, or that they are properly designed or properly functioning, or that they are designed with the intention of producing true beliefs? Provided that we grant (3), what need is there for us to struggle with (1) and (2)? This, I think, is the question that must be addressed. If we consider Plantinga's Case of the Epistemically Serendipi足 tous Lesion, it seems that a minor alteration will show that proper function does not contribute to warrant in the manner that he has suggested. In this instance, a person has a brain lesion tha t makes hel' believe all sorts of ridiculous things,and one of those beliefs is that she has a brain lesion. Surely she does not know that she is suffering from a brain lesion, even though it is true. Thus, she is not warranted in believing it (WeD. IX, II, B, p. 256, ff.). This case is invoked to show that a causal type of reliabilism (such that I am warranted in the belief that s is F iff s's being F causes or sustains that belief) is false.* However, we need only change this to the Case of the God-Given Epistemically Serendipitous Lesion to show that pl'oper function is also false. In this case, however, the brain lesion does not produce a large number of beliefs; it only produces one, that the pel'son has a brain lesion. Further, if we stipulate that God gives someone that lesion so thatshe would truly believe that she has a brain lesion, then .. Throughout this paper the term "iff" will be used to designate "if and only if" in logical propositions. -ed.
PROPER FUNCTION, RELIABILI'IY AND WARRANT
Plantinga must concede that she has knowledge, even though it is clear to the rest of us that she does not. None of her other faculties are disrupted by the lesion; it adds to her cognitive apparatus without taking anything away. God. working in mysterious ways ashe does, designs a brain lesion that will always produce this one true belief. This is the purpose of the lesion, and God creates a faculty. rather than a deformity. Since this new lesion-sensing faculty was devised and implanted by God, it is clear that (1) the belief in question is the result of a properly functioning faculty in an appropriate environment (God took environment into consideration when devising the leSion), (2) the relevant segment of the design plan is aimed at true belief and (3) the design is a good one (the belief it produces is always true). This case fits perfectly with Plantinga's notion of warrant, yet surely this person is not really warranted in believing that she has a brain lesion.2 What relevant difference would it make if the lesion were produced by God and not by some injury? A similar example can be borrowed from an old episode of The Twilight Zone. In this episode, a man tosses a coin and it lands on its edge. For the rest of the day he is able to read people's minds. By the end of the day it would seem as if he might have some warrant for the beliefs produced by this new and amazing faculty. since he would have the opportunity to check what he "hears" with other sources, conferring with people he could trust, and so forth. An important point to note, however, is that whether we decide that his beliefs are warranted or not, we do so before discovering whether his new faculty is a gift from God or the result of some freak accident. We can modify this scenario a bit and suppose that God gives our coin-tosser his mind-reading faculty. but that this faculty works only once every six hours and 42 minutes. Two or three times a day our friend hears a voice in his head which seems to be coming from someone else's head and he believes that he is hearing someone 2 Even if the lesion also produces a vast number of ridiculously false beliefs (as in Plantinga's original case), this counterexample should still hold, for Planting" stipulates that only the relevant segment must be aimed at the production of true beliefs. If the lesion's recipient ill/ers tha t she has a brain disorder because she notices that she suddenly has a large number of ridiculous beliefs, then we are suddenly dealing with a different cognitive means (inference rather than the lesion). For Plantinga's theory to hold, she must be warranted prior to inference, since proper function etc. is supposed to be sufficient for warrant.
JACK C. LYONS
else's thoughts. However, the faculty works only for the duration of a single thought. and he is not around anyone with whom he can speak in order to test the verity of his new faculty. Once again, God has designed a splendid, truth-yielding, cognitive faculty, even if it is not always at the disposal of the subject>swill (Plantinga does not claim that it must be); the faculty does exactly what God intended it to do, in the environment for which God intended it. Six hours and 42 minutes after tossing that fated coin, our protagonist accidentally bumps into a scowling oldbiker and "hears" his first thought: "Watch where you're going, you silly and brutish oaf." But our friend certainly does not know that "silly and brutish oaf" was the precise appellation that the biker had formulated in his mind; surely he is not warranted inbelieving that. (In fact, given the subject's past experience with scowling old bikers, he is probably warranted in believing that "silly and brutish oaf" was not the appellation used in this particular case). As time progresses. our friend may learn that these voices come to him exactly every six hours and 42 minutes, and he may sit down and test this faculty with his wife or a friend, thereby learning that it does in fact yield true beliefs. It would seem tha t at this point (and not until this point) our protagonist may be warranted in his beliefs. But if we stipulate that this faculty is the result of some freak accident and not the work of God or evolution, then we are forced to choose between our coin-tosser's warrant and Plantinga's theory, for Plantinga would have to hold that the man's beliefs are not war足 ranted, since proper function, etc. is a necessary condition for war足 rant. Until our friend has somehow ascertained the reliability of his new faculty, it would seem that he is not even nearly warranted. Whether he is a fortunate recipient of a divine gift or a hapless individual who just stepped into the Twilight Zone, the beliefs produced by this brand-new faculty simply are not warranted. Examples like this can be easily multiplied. We can imagine any number of adventitious faculties that simply defy explanation. In every case, I think. we will say that the recipient is not (at least not immediately) warranted inbelieving wha t these faculties induce her to believe, no matter where they have come from. The mere fact that they are new and bizarre implies that they are not to be trusted prior to investigation. What I think these thought experiments show is not only that
PROPER FUNCTION, RELIABILITI AND WARRANT
there is something wrong with proper function theory but also just what is missing from simple reliabilism. What hinders warrant in these cases, the God-Given Lesion and the Part-Time Mind Reader, is that although the cognitive apparatus is completely reliable and functioning in a proper environment, the cognizer has little or no reason to believe that it is. Reliability is an extremely important aspect of warrant, but not just any reliability will sulfice, it must be reliability that is recognized as such btJ the cognizer. This seems to invite immediate problems. but it need not. Reli足 ability as I conceive it is an objective, external property of a cognitive faculty or process. I think that Plantinga is right in thinking that an externalist account of warrant is more promising than an internalist one. I may be doing my very best to believe all and only propositions that are true, but if I am suffering from some type of cognitive malfunction I will not be warranted in many of these otherwise responsibly formed beliefs. However, as my intended counterexamples to proper function theory show, this alone is not sufficient. The cognizer must also have some reason to believe足 some assurance-that this faculty or process is, in fact, reliable. Before I try to explain how this might be had, though, I should like to set out a perfunctory explanation of what a faculty or a process is, as this has caused some problems in the past. First of all, I would like to lump them together under the title "cognitive means." (The way in which I use this term should obviate any objections; for the present purposes, the difference between a faculty and a process will not be significant). A cognitive means is anything that will suffice as a reasonable answer to the question, "How do you know that?" There will be unreasonable answers like, "A little bird told me," but these should be fairly easy to weed out without a great deal of controversy. Reasonable answers would include such responses as, HI remember it," or "I saw it," or "I multip lied the numerator and the denominator of the second fraction by two and added the numerators of both fractions," or "I read it in the newspaper." Perhaps we could even say, "Every six hours and 42 minutes I have the ability to read minds for the duration of a single thought, and I 'heard' you think that." Inany case, the mere fact that we use the label "cognitive means" does not entail that we automatically assume reliability. Simply answering the above question will not always give us a clear-cut delineation of cognitive means, but it is not obvious that a
clear-cut delineation is always necessary. We can and should, how足 ever, make more precise distinctions where there is a difference in reliability between different species of the same generic means. For example. we might initially want to view memory simpliciter as a cognitive means, until we have some assurance of differences in reliability between different kinds of memorial functions. I might have some strange memorial disorder which prevents me from remembering things that happen on Mondays. In this event, the rest ofrny memorial cognitive means might be reliable, although when it comes to things occurring on Mondays, I cannot be warranted in believing that I remember them. By way of comparing reliability we come to distinguish different cognitive means. Thus the distinctions can and should vary from person to person, as far as concerns the individual's assessment of her own equipment. All instances of seeing are not the same for a person with poor night vision. For this reasonitisneithernecessarynordesirablethatwealwaysmakethese distinctions clear--cut. 3 Having touched on this, I think I can explain my conception of warrant in slightly more concise terms: 5 is warranted in believing p iff (1) the cognitive means (c) that produces or sustains 5's belief in pis reliable, and (2) 5 has some proper assurance that c is reliable. As my objections to Plantinga are intended to show, mere reliability (or as Plantinga calls it, "a high statistical probability that a belief produced under those conditions will be true"), whether produced by God or not, is not enough. The cognizer must also have what I call proper assurance that the relevant cognitive means is reliable. If proper assurance were merely warrant under a different name 3 Although perhaps not always clear-cut, the distinctions will have to be very narrow when it comes to Gettier cases, my response to which will be to suggest Ulat certain very particular cognitive means in very particular circumstances are not reliable. I think I can do this because I have included within the designation "cognitive means" both faculties and processes. In normal circumstances the inferences made may be completely reliable, but in these rare situations they are not Gustasvision is not reliable under certain lighting conditions); even though we may be warranted in thinking that the inference is reliable, it might be the case that the inference itself is not reliable and that it is therefore not warranted.
PROPER FUNCTION, RELIABILITY AND WARRANT
it would be very difficult for us to know anything, for this would yield an infinite regress. Instead, I mean to include something like a particular, weaker form of justification which applies only to beliefs concerning the reliability of cognitive means. This should become more clear as I continue. In the above examples, I indicated that the adventitious faculties did not produce warrant but that they could and often would if the agent were given enough time to test their reliability. How we determine this reliability can be seen by observing how we do this with the faculties we already possess, how we come to some feeling of assurance concerning our normal faculties. This seems to be accomplished through coherence. The type of coherence I have in mind here is very similar to that which David Burne invokes to explain the origin of our ideas of an external reality. Burne's theory is that we come to believe in the endurance of objects beyond our immediate perception, because to do otherwise would fly in the face of our experience. Whenwe sit and watch objects, we conSistently find thatthey do not simply disappear and reappear. [My] observations are contrary, unless I suppose that the door still remains, and that it was open'd without my perceiving it: And this supposition, which was at first entirely arbitl'my and hypothetical, acquires a new fOl'ce and evidence by its being the only one, upon which I can reconcile these contradictions (Hurne, p. 196-7). I think that Hume is right in believing that the only evidence we can find for the reliability of our senses is some degl'ee of coherence. 4 We cannot directly perceive objects without the interposition of our faculties, and so we cannot simply compare the sense impressions to the things themselves. A foundationalist might want to claim that the evidence of our 4 I do not want to draw Hume as <1 coherentist, although I think that his contribution here to coherentism is often neglected. Hume certainly does not think that this argument entailed a proof that we have sensory knowledge (in his senses of the terms, "proof" and "know ledge"). He does, however, seem to think tha t this comprises a justification in some broadly deontological sense.
senses is properly basic, but I hardly think any philosopher would believe that the cognitive productions of my intermittent mind足 reading capacity are. We must be skeptics here, in regards to these supposed adventitious faculties, and if skepticism is to be taken seriously, the only reasonable resolution I can conceive is some fonn of coherence. This will not be an unmitigated and egalitarian coherentism. There will be certain beliefs that must enter into the equation, beliefs about simple necessary truths and about our experience, a "founda足 tion," if you will. 5 Ourpsychological/ epistemic goal here is to reduce incoherence, at least enough so that it does not bother us. To contra足 dict the facts that a thing cannot be red and green all over at the same time or that I am currently appeared to redly cannot ease any of the tension between my beliefs. for these beliefs cannot and should no t be gotten rid of. At every tum, necessary truths and truths of appearances will force themselves upon us. Therefore, my idea of proper assurance can be stated as follows: 5 has proper assurance of the reliability of c iff: 5 is warranted bysome other, previously established cog足 nitive means (c,,) in believing that cis reliable, or (1) 5' s assessment of c takes into account the purported testimony of c, (2) 5's belief in the reliability of c does not contradict any obvious necessary truths or truths of appearances for 5 and (3) 5's experience with regard to c has been fairly regular and coherent and fits with the informationgiven5by his other faculties. The first conjunct above implies only that 5 need not rely upon personal experience to test the reliability of the cognitive means in question. Provided that 5 is warranted in, say, believing a doctor's testimonythathis(S's)hearingisworkingperfectlywell,hewillthen have proper assurance of the reliability of his hearing (if he did no t already). It is also important to note that this is only one possible way 5 This is to protect us from cases like Plantinga's Epistemically Inflexible Climber, who suffers from a cognitive malfunction that inhibits the production of new beliefs such that he still (coherently) believes tha twhich was true several hours ago but is now false. If we did not hold that belief in appearances was mandatory. our cognitive agent could simply devise coherent but ridiculous belief systems out of sheer perversity or a misguided effort to relieve the tension of incoherence.
PROPER FUNCTION, RELIABILlTi:' AND WARRANT
of achieving proper assurance; if it were the only way, proper assurance would lie forever beyond our grasp. Implicit in this definition of proper assurance is the notion that S is aware of the cognitive means that produces or the sustains a particular belief. Therefore, victims of the machinations of some Cartesian demon (or Plantingan Alpha-Centaurian) would not be warranted in the beliefs thus produced, for typically we would think that these people do not know that their beliefs are being produced in this manner. These people cannot assess the reliability of this particular cognitive means (the demon or Alpha-Centaurian), for they are not even aware of its existence. I feel that my description of proper assurance captures the way in which we do assess our faculties. I use the term "assurance" in order to imply that this is internal. Thus, my theory is a combination of internalism (proper assurance) and externalism (reliability). Just as Plantinga causes a great deal of problems for intemalism by providing cases in which a person is doing everything right but whose cognitive faculties are malfunctioning, so too, I think, have I made things more difficult for the pure externalist by suggesting adventitious but veridical faculties. Pure internalism seems better suited for epistemic obligations than for knowledge, while pure externalism seems better suited for consistently true beliefs than for know ledge. I hope that this theory can enj oy the better aspects of each of the other two. There are several consequences of this theory, and I would like to take note of a few of them. First of all, in the case of the God-Given Brain Lesion, its recipient could never be warranted in the belief that this lesion produces. The lesion functions as a perfectly reliable faculty, but since it only produces a single belief there is nothing else with which that true belief can cohere. It is difficult to see how she could ever gain any assurance of reliability. Also,she could not possibly fulfill my criteria as she has no idea why she has the beliefs she has. She cannot answer the question, "how do you know that?".6 With the Part-Time Mind Reader we will see a very different case. In this instance our protagonist will not at first be warranted in 6 As I mentioned in footnote 2, above, the beJiefthat one has a brain lesion couid also be produced by other means, means that are warranted. If the victim's doctor tells her that she has this lesion, she could, of course. be warranted in that belief, but the lesion alone still does not produce warranted belief.
beliefs produced by these new and untested cognitive means. How- . ever, with time and experience, he might come to have warrant if he can recognize (through introspection perhaps) the workings of dis足 crete faculties and can achieve some assurance of the reliability of these faculties through the type of coherence mentioned above. Our Part-Time Mind Reader can subject himself to experiments, relying a great deal on the testimony of others ("You're right! That is exactly what I was thinking!"), or the coherence of his readings with other facts, perhaps one day achieving warrant. A consequence ofthis theory is that babies and Adam (if we could bring him into the world fully rational but without any experience, as God and the Early Moderns are wont to do) would not be warranted in believing anything, except perhaps truths of appear足 ances and some obvious (to them) necessary truths. Experience is necessary for warrant, since it is needed to provide assurance of the reliability of our cognitive means. But this seems to work for the present theory rather than against it. Babies probably do not know anything; even small children, sometimes, close their eyes in the hope that something frightening will therefore go away, seemingly not fully aware yet of the endurance of external objects. It seems plausible that Adam would be like an infant (as far as knowledge is concerned) who, lacking only the requisite experience, would grow up very quickly in this respect. Obviously more work will have to be done to thoroughly flesh this theory out. 1have tried to give a rudimentary approximation of warrant and to show why it might be initially plausible. It seems that this theory, in combining reliabilism with elements of coherentism and foundationalism, as well as externalism and internalism, might avoid some of the pitfalls of its component theories ..
Hurne. David. A Treatise ofHuman Nature. P. H. Nidditch, ed. Oxford University Press, 1978.
[WCD] Plantinga, Alvin. Warrant: the Current Debate. Unpublished. [WPF] Plantinga, Alvin. Warrant and Proper Function. Unpublished.