9789174637090

Page 1


Rigvedic Formulas and Meter



Urban Lindqvist

Rigvedic Formulas and Meter


Š 2015 Urban Lindqvist Printing and Publishing: BoD ISBN: 9789174637090


CO TE TS

FOREWORD ............................................................................................. 7 1. RIGVEDIC METER ............................................................................... 1a. Dimeters ................................................................................... 1b. Trimeters ................................................................................. 1c. Stanzas consisting of verses of different types ....................... 1d. Strophes .................................................................................. 1e. Metrical irregularities ............................................................. 1f. Word-boundary patterns ..........................................................

8 9 10 13 14 14 16

2. FORMULAS AND FORMULAIC DICTION ............................................. 26 3. DIMETER VERSES WITH CADENCES OF THE TYPE \””\ ................. 3a. Introduction ............................................................................. 3b. hárivah° .................................................................................... 3c. hárayah° ................................................................................... 3d. duritá- ..................................................................................... 3e. gr°hápati-/vásupati- ................................................................. 3f. átithi- ....................................................................................... 3g. sanutár ....................................................................................

36 36 46 52 55 59 66 67

APPENDIX: WORD-BOUNDARY PATTERNS ............................................. Precaesural word-boundary patterns in triäTubh lines with an early caesura ................................................................ Precaesural word-boundary patterns in triäTubh lines with a late caesura .................................................................... Precaesural word-boundary patterns in jagatI lines with an early caesura ................................................................ Precaesural word-boundary patterns in jagatI lines with a late caesura ....................................................................

69 71 71 72 72


Precaesural word-boundary patterns in lyric 12-syllable lines with an early caesura .................................................................. 73 Precaesural word-boundary patterns in lyric 12-syllable lines with a late caesura ...................................................................... 73 Postcaesural word-boundary patterns in triäTubh lines with an early caesura .................................................................. 74 Postcaesural word-boundary patterns in triäTubh lines with a late caesura ...................................................................... 75 Postcaesural word-boundary patterns in jagatI lines with an early caesura .................................................................. 76 Postcaesural word-boundary patterns in jagatI lines with a late caesura ...................................................................... 77 Postcaesural word-boundary patterns in lyric 12-syllable lines with an early caesura .................................................................. 78 Postcaesural word-boundary patterns in lyric 12-syllable lines with a late caesura ...................................................................... 80 Word-boundary patterns in triäTubh lines with an early caesura ..... 81 Word-boundary patterns in triäTubh lines with a late caesura ......... 92 Word-boundary patterns in jagatI lines with an early caesura ...... 102 Word-boundary patterns in jagatI lines with a late caesura .......... 110 Word-boundary patterns in lyric 12-syllable lines with an early caesura ................................................................ 118 Word-boundary patterns in lyric 12-syllable lines with a late caesura .................................................................... 123 Word-boundary patterns in gAyatrI lines ....................................... 128 Word-boundary patterns in anuäTubh lines .................................... 133 Word-boundary patterns in lyric 8-syllable lines .......................... 138 BIBLIOGRAPHY....................................................................................... 142


FOREWORD This text is part of what should have become my doctoral dissertation. Having seen it lying around for about six years now without doing anything with it, I came to the conclusion that it was time to let it go. Before I do, I need to thank some people, in particular Gunilla GrenEklund, Uppsala, who has been by my side offering various kinds of support since my first stumbling steps into Classical Sanskrit, and Alexander Lubotsky, Leiden, who has not only read and commented on an earlier version of this text, but also offered help and advice in some other respects. I am also grateful to Christiane Schaefer, Uppsala, with whom I had the opportunity to discuss Vedic and Indo-European matters on numerous occasions, as well as several other people who in one way or another made my sojourn in the academic world a lot nicer and more interesting than it would have been without them and helped me find my place within – and without – Indo-European studies: Dimitrios Iordanoglou, Jenny Larsson, Rikard Karlsson, Viveka Velupillai, Peter Larsson, Anna-Pya Sjödin, Peter Jackson, Jonas Tiljander, Mats Eriksson, Klas Hagren and Barbro Eriksson (I am sure that those who consider themselves unduly unmentioned will realize that it is more a sign of absent-mindedness than ingratitude on my part). I hereby declare this thesis public – to be quoted, criticized or ignored like any other thesis. It’s all yours. Orsa, 2011 Urban Lindqvist

P.S. For those who prefer a traditional dead-tree edition, here it is. Ibid., 2015 U.L.

7


1. RIGVEDIC METER In this section the main features of Rigvedic meter will be outlined. The first four parts (1a-d) will to a large extent follow the standard works by Oldenberg (1888) and Arnold (1905).1 The basic unit of the Rigvedic meter is the verse or line (pAda-), the internal structure of which is regulated by three principles, viz. the number of syllables, alternations between long and short syllables and certain metrical boundaries.2 Verses are grouped together to form stanzas (Rc-). The most common kinds consist of three or four verses. A hymn (sUkta-) is made up of a varying number of stanzas (or strophes; see below, 1d), usually 3-15. Most Rigvedic verses consist of eight, eleven, or twelve syllables. The longer ones of eleven or twelve syllables are often grouped together as trimeters as opposed to dimeters (the eight-syllable verses). The quantity of the first and last syllables of all verses is metrically irrelevant. In the description of the various meters the following metrical symbols will be used: ’ ” \ 1

2

3

8

long syllable3 short syllable anceps (syllable with metrically irrelevant quantity)

The chronological relationship between the various meters and parts of the Rigveda as outlined by Arnold is, however, unacceptable. His relative chronology is “in vielen Punkten fraglich […], zumal ihm das Metrum Aufschluß über das Alter der Hymnen gibt, woraus wiederum auf das Alter der Metren geschlossen wird.” (Korn 1998:18). The Rigvedic accent is usually not considered metrically relevant in the sense that the placement of accents within a verse affects the degree of regularity of the verse, but a preliminary study by Lubotsky (1995) shows that the Vedic poets occasionally did make use of the possibility of creating certain patterns of interchange between high and low tones. Further studies along the same line will undoubtedly improve our understanding of Rigvedic poetics. Long syllables have as their syllabic peak a diphthong, a long vowel/syllabic ", or a short vowel/syllabic R or L followed by at least two consonants. All other syllables are short.


X |

syllable with descriptively irrelevant quantity caesura (4| = caesura after the fourth syllable, 5| = caesura after the fifth syllable)

1a. Dimeters A dimeter line consists of two parts, an opening and a cadence, and the most common type has the following structure: \’\’ ”’”\

devó devébhir Á gamat (1.1.5c) ‘May the god come with the gods’

Sometimes the second syllable is short; then the third is generally long: \”’’ ”’”\

sutám Á ganta tÚrNayaH (1.3.8b) ‘Come, quick, to the pressed (soma)’

Two of the four most common meters in the RV are made up of eightsyllable lines, viz. gAyatrI (stanzas of 3×8 verses) and anuäTubh (stanzas of 4×8 verses).4 4

Oldenberg (1888:12) notes that the cadences of the second and third verses (b and c) of the gAyatrI stanzas are generally more regular than the ones in the first (a). In the gAyatrI sample text of 6,683 verses (see 1f below) 92.65% of the a-verses have cadences of the type ”’”\, and the corresponding figures for the other verses are 96.41% (b-verses) and 95.97% (c-verses). If we look at the openings instead, a different pattern emerges, where c differs from a and b. But both patterns show increasing regularity within the stanza:

a b c

\’\’

\”’’

\’\”

\”’”

Other types

58.13% 55.99% 70.14%

16.40% 18.17% 13.29%

14.02% 16.33% 12.26%

7.89% 7.49% 2.91%

3.56% 2.02% 1.40%

9


Both these also have variants called trochaic gAyatrI and epic anuäTubh respectively. The trochaic gAyatrI has a cadence with the primary structure \”’\,5 and in the epic anuäTubh the cadences of the first and third verses (often) have the structure ”’’\, and the openings of the second and fourth verses more often have a long than a short third syllable. A paqkti is made up of 5 dimeter verses, the last one of which is often a refrain. There are also seven-syllable catalectic (\’\’ ”’\) and acephalic (\\’ ”’”\) variants of dimeter verses (see Vine 1977a).

1b. Trimeters The other two major meters of the RV are the triäTubh (4×11) and the jagatI (4×12).6 They are to a large extent identical in metrical structure except that the last syllable in the triäTubh verse corresponds to a short syllable followed by an anceps in the jagatI verse. The trimeters are divided into three parts, an opening, a break, and a cadence. The opening and the break consist of seven syllables (4 + 3 or 5 + 2). After the fourth or fifth syllable, between the opening and the break, most trimeter verses have a caesura, which usually corresponds to a word boundary, but sometimes it separates the two members of a devatAdvandva compound, such as mitrÁ|váruNA. According as the caesura comes after the fourth or fifth syllable it is called early or late caesura.7 The openings of the verses with an early caesura are generally of the same kinds as those in dimeter verses. Before a late caesura the first four 5

6

7

10

It is, however, subject to much variation. See Oldenberg (1888:23f.), Arnold (1905:168) and Vigorita (1979: 220, fn.1). The terms triäTubh and jagatI are sometimes used to refer also to individual verses of eleven and twelve syllables respectively. Arnold (1905:11f.) and van Nooten–Holland (1994:viii) divide the part before the cadence into an opening of four and a break of three syllables irrespective of whether the caesura is early or late (i.e. XXXX | XXX and XXXX X|XX), but see Korn 1996:14 with fn. 18 and literature.


syllables are identical to the four-syllable openings, and the fifth syllable is anceps, i.e. \’\’\ is the primary type, and \”’’\ the most common variant. But it is also generally acknowledged that the five-syllable openings are more strictly regulated than the four-syllable openings. A search through the trimeter sample texts (as described in 1f below) bears this out:

\’\’ | \’\’\ |

lyric 12-syllable lines8

triäTubh

jagatI

65.09% 80.73%

68.12% 83.11%

66.84% 81.91%

After the caesura comes the break, the most common structure of which is ””’ (after an early caesura) or ”” (after a late caesura).9 According to Gippert’s statistics (1999:110-111) both syllables after the caesura are short in 70.55% of the 11-syllable verses, 74.97% of the 12-syllable verses. However, if we look at the two syllables separately, it becomes clear that most of the deviations are in the first syllable:

8 9

11-syllable verses

12-syllable verses

| X” | X’ Total

15,587 = 93.64% 1,059 = 6.36% 16,646 = 100%

6,419 = 94.4% 381 = 5.6% 6,800 = 100%

| ”X | ’X Total

12,533 = 75.29% 4,113 = 24.71% 16,646 = 100%

5,366 = 78.91% 1,434 = 21.09% 6,800 = 100%

See 1c below. The ratio between |XX’ and |XX” in lines with an early caesura is about 3:1 in triäTubh lines and about 4:1 in jagatI lines.

11


Therefore we may say that the main tendency is that the second syllable after a caesura is short, which was noted already by Oldenberg (1888:5657), and regard the breaks ’”’ and ’” as less irregular than those that have a long second syllable. Furthermore, there is also a difference in strength of tendencies between the verses with an early caesura and those with a late caesura. Just as in the case of the openings the tendencies are stronger in the verses with a late caesura:10 triäTubh

jagatI

lyric 12-syllable lines

4| ” 5| ”

69.08% 84.01%

69.40% 90.09%

74.08% 77.59%

4| X ” 5| X ”

90.64% 96.18%

93.69% 97.45%

77.89% 96.47%

The trimeter cadences consist of four or five syllables and have the forms ’”’\ (triäTubh) or ’”’”\ (jagatI). Examples: 11-syllable line with an early caesura (2.9.3a): \’\’ | ””’ ’”’\ índrÁ gahi prathamó yajñíyAnAm ‘Indra, come here as the foremost among those worthy of sacrifice’ 11-syllable line with a late caesura (4.14.4a): \’\’\ | ”” ’”’\ Á vAM váhiäThA ihá té vahantu ‘May the best-carrying ones carry you two here’

10

12

These statistics are drawn from the same sample texts as those on the openings. Gippert’s statistics cannot be used in this case, since he does not distinguish between verses with different caesuras.


12-syllable line with an early caesura (1.114.6a): \’\’ | ””’ ’”’”\ idám pitré marútAm ucyate vácaH ‘This speech is delivered for the father of the Maruts’ 12-syllable line with a late caesura (6.47.27d): \’\’\ | ”” ’”’”\ índrasya vájraM havíäA ráthaM yaja ‘Honour Indra’s weapon, the chariot with a libation’ Besides the stanzas of four trimeter verses there are also those that have either more or fewer verses. Only one type, however, occurs more often than just sporadically, viz. the virAj, which consists of three eleven-syllable lines. Besides variation in the number of verses within a stanza, there is also variation in the number of syllables within a verse. For example, within a triäTubh stanza verses of ten or twelve syllables are sometimes found. One type that occurs often enough to be considered a separate meter is the dvipadA virAj. It consists of two verses of ten syllables each (\’\’\ | \’\’\) (or of four verses of five syllables each).

1c. Stanzas consisting of verses of different types Disregarding sporadic irregularities, all of the above-mentioned types of stanzas are characterized by the principle of isosyllabism, i.e. all verses within a stanza have the same number of syllables. In the RV there are also stanzas that consist of various combinations of eight- and twelve-syllable verses (termed “lyric metres” by Arnold 1905:8).11 The most frequently used ones are kakubh (8 12 8), bRhatI (8 8 12 8), satobRhatI (12 8 12 8), uänih (8 8 12 or 8 8 8 4), and atyaäTi (12 12 8 8 8 12 8). 11

A few examples of eleven-syllable verses in combination with eight-syllable verses are also met with.

13


1d. Strophes A strophe is a (non-obligatory) unit intermediate between stanza and hymn. A kakubh or a bRhatI combined with a satobRhatI makes up a pragAtha strophe, and three (gAyatrI) stanzas grouped together make up a tRca.

1e. Metrical irregularities Everything in connection with Rigvedic meter is best described in terms of tendencies. Some of them are strong enough to be called rules, but then it would be difficult to justify the arbitrary choice of a percentage that would be necessary in order to differentiate the rules from the (slightly weaker) tendencies. And no rule would be without exceptions.12 In accordance with this, “metrical irregularity” is not considered an absolute concept: a verse can be more or less irregular (statistically unusual), depending on the strength of the tendencies it deviates from. For instance, a verse with a long second syllable after the caesura is more irregular than a verse with a long first syllable after the caesura. In practice, however, a strict hierarchy is impossible, but we still need some general guidelines. There are three criteria which are generally used to decide whether a verse is regular or not: the number of syllables in the verse, the quantity of the syllables in certain portions of the verse and presence/absence of a caesura in trimeter verses. A fourth important criterion is the frequency of a statistically unusual feature within a hymn. It is not always obvious where to draw the line between metrical irregularities and unusual metres. A case in point is the trochaic gAyatrI. If the majority of lines of (a part of) a gAyatrI hymn show cadences of the type \”’\, then we cannot judge this hymn in the same way we judge a different gAyatrI hymn with only one such cadence besides 12

14

In cases where all possibilites are equal we may, of course, speak about rules without exceptions, e.g. “the quantity of the last syllable of a verse is metrically irrelevant”, but such cases are better regarded as complete absence of tendencies.


others of the most common kind (”’”\). In the first case we are dealing with a separate meter, in the second with a metrical irregularity. Between these two extremes there is a continuum, where all cases have to be judged individually. In the very middle of it we may speak about equal possibilities (\’”\ and \”’\), but often different tendencies apply to different parts of a hymn. Naturally, the frequency criterion applies not only to quantity patterns, but also to the other two main criteria, and not necessarily only to one at a time. Some hymns show various kinds of irregularities and may therefore be considered generally irregular within the Rigveda as a whole, whereas the individual lines must be considered regular (free from tendencies) within the hymns. The first criterion, the number of syllables in a verse, is fairly straightforward: if a line does not contain the expected number of syllables, then the verse is irregular. The main exception is seven-syllable verses, especially catalectic and acephalic ones, in contexts where eight-syllable verses are expected.13 As regards quantity, the two parts of a verse with the least variation are the breaks in trimeter verses and the cadences of all verses. But the fact that there is more variation in the openings does not mean that they are completely devoid of tendencies. The main difference between openings and cadences in this respect is that there are more different kinds of regular openings than there are kinds of regular cadences. But just as there are irregular cadences, there are also irregular openings, in particular those that have short both second and third syllables, and also those that have a short fourth syllable.14 The tendency for trimeters to have a caesura after the fourth or fifth syllable is about as strong as the tendency for the second syllable after the

13 14

See Vine (1977a:626ff.) for a discussion about the distribution of seven-syllable lines. See the statistics in van Nooten–Holland (1994:xvii) and Gippert (1999:109-110).

15


caesura to be short.15 Absence of a caesura after the fourth or fifth syllable in a trimeter line is therefore a strong indication of irregularity.16

1f. Word-boundary patterns If we apply the three main criteria discussed above to the verse 1.62.3c (= 10.68.11d) it seems to be a perfect triäTubh line. It has the right number of syllables and an early caesura, and the opening, the break and the cadence all show the most common metrical patterns: \’”’ | ””’ ’”’\

b1haspátir bhinád ádriM vidád gÁH ‘BRhaspati splits the rock, finds the cows’

However, in order to arrive at a proper evaluation of the degree of regularity of a verse we need to take as many tendencies as possible into account. If we look at the distribution of word boundaries in the RV, we notice that there is a statistical preponderance of certain word-boundary patterns as well, besides the generally accepted caesuras. The word-boundary pattern after the caesura in 1.62.3c/10.68.11d can be described as 5-6 # 7-8 # 9-10 # 11, where “5-6” refers to a word that occupies the syllables 5-6 (= bhinád), “#” to a word boundary, and so on. This particular pattern occurs only fifteen times in a sample of 8,461 triäTubh lines with an early caesura.17 In other words, from the point of 15

16

17

16

According to van Nooten–Holland (1994:ix) such a caesura is found in more than 94% of the trimeter verses. Vigorita (1979) has drawn attention to the fact that the trochaic gAyatrI, in contrast to other dimeter types, has a relatively fixed caesura. According to him, 94.2% of the trochaic gAyatrI verses show a caesura after the fourth or fifth syllable, as opposed to 71.4% of the ordinary gAyatrI verses. To the extent that the trochaic gAyatrI lines can be properly delimited, the caesura criterion can be used here as well (among the verses considered trochaic gAyatrI by Vigorita, there are some stanzas which may just as well be regarded as regular gAyatrI, e.g. 8.5.34). See below for the statistical details.


view of word-boundary patterns, 1.62.3c and 10.68.11d are highly irregular. Therefore we need to turn to the frequency criterion, and it appears that several other verses in 1.62 show statistically unusual features as well. Those that are connected with word boundaries are the following: six verses have a caesura after the third syllable (1b, 2d, 5a, 5b, 5c, 6b) and one has none (6a) (or, possibly, two have a caesura before the suffix -tama-, 6a and 6b; cf. Oldenberg 1888:45, fn. 1). Besides these there is also a number of other irregularities.18 Similarly, in 10.68 four verses show rare wordboundary patterns (4b, 5d, 9a, 12c), one verse has a caesura after the third syllable (2a) and one after the sixth syllable (or before an enclitic: 5b). Thus the hymn 1.62 may be considered generally irregular, and 10.68 irregular as regards word-boundary patterns. In the case of 1.95.8c the situation is somewhat different: \’’’ | ””’ ’”’\

kavír budhnám pári marmRjyate dhÍH ‘The poet – (his) thought – wipes the ground all round’

In this verse the postcaesural word-boundary pattern is 5-6-7-8-9-10 # 11, which is also rare (14 lines out of 8,461).19 There is only one more line in this hymn that shows a rare word-boundary pattern (10b: 6-7-8-9-10 # 11). Otherwise the hymn is generally quite regular, at least in regard to wordboundary patterns,20 and there is therefore good reason to regard 1.95.8c as an irregular line.

18

19

20

E.g., four verses have a long second syllable after the caesura (1d, 9d, 10d, 13d), and one has an irregular cadence (3d; regarding 5a, see Lubotsky 2000:317). There are also some verses that only have ten syllables, but the exact number of verses depends on the restoration: in the edition of van Nooten and Holland only one verse (8c) is left with ten syllables, and two (3a, 7d) are restored with so-called rests (see fn. 31 below). If we regard the preverb pári and the verb marmRjyate as two words, the word-boundary pattern would be different (5-6 # 7-8-9-10 # 11) but still rare (9 out of 8,461). There are seven verses with a long first syllable after the caesura (4d, 5a, 5c, 6a, 6c, 6d, 9d), and two openings that deviate from the most common patterns preceding a late caesura, \”’”’ (10b) and \’’”” (9a).

17


For statistical purposes various electronic sample texts have been used. They are all based on the electronic version of van Nooten’s and Holland’s edition of the Rigveda.21 It should be clear that no such statistics can be considered absolute or final, but they can be used as general guidelines, since some of the tendencies are too strong to be the result of infelicitous restorations. First of all, before we can identify word boundaries we need to identify clitics.22 Besides comparative evidence from other Indo-European languages, there are mainly two criteria commonly used to identify clitics in Vedic, absence of accent and “the second position” in one sense or another (Hale 1987; Delbrück 1888:22; Trask 1992, s.v. Wackernagel’s Law).23 Items which are both unaccented and occupy “the second position” can fairly safely be assigned the label clitics. To this group belong the pronouns and particles enumerated by Macdonell (1910:81f., §85b1α-β, §85b2δ) except ena- (see fn. 24 below). More problematic are those items that either occupy the second position (various accented particles; Delbrück 1888:22) or are unaccented (finite verbs in main clauses, vocatives and certain pronominal forms; Macdonell 1910:81f., §85b2α-γ).24 Neither of these two criteria is sufficient by itself, since the fact that some particles are postpositive may simply be due to a word-order rule, and absence of accent is in itself no proof that an item is a 21 22

23

24

18

Some obvious errors were first corrected, primarily those noted by Gippert (1996:66f.). I use the term ”clitic” to denote items intermediate between affixes and independent words (Trask 1992, s.v. clitic; Zwicky 1985:283) without any inherent implications about their position in relation to other items in the sentence (cf., e.g., Vine 1977b:662f.). Such relations are better expressed by the terms “proclitic” and “enclitic”. Internal sandhi is less useful in the case of the RV (cf. Zwicky 1985:286), partly because it is inconsistent, e.g. prá No naya (8.71.6c) vs. prá no muñcataM (6.74.4c), partly because “close” sandhi can also be conditioned by syntactic relations like noun + postposition and genitive + headnoun (Hale 1990:83f.). In accordance with Kupfer (2002:49-79), ena- and the unaccented forms of the pronominal stem a- are regarded as suppletive forms of the same paradigm. A consequence of this is that ena- is considered capable of occuring in non-second position just like the forms ásya, ásmai and ÁbhiH with sentence-/pAda-initial accent (on which see Klein 1992:2; see Kupfer 2002:58-60 on ásya and ásmai, and pp. 73-74 on the problematic enÁm in 8.6.19c).



Issuu converts static files into: digital portfolios, online yearbooks, online catalogs, digital photo albums and more. Sign up and create your flipbook.