sahasram ati srajas - lmu
TRANSCRIPT
ii
“Prelims” — // — : — page — # ii
ii
ii
Sahasram Ati Srajas
Indo-Iranian and Indo-European Studiesin Honor of
Stephanie W. Jamison
edited by
Dieter GunkelJoshua T. KatzBrent Vine
Michael Weiss
Beech Stave PressAnn Arbor •New York
ii
“Prelims” — // — : — page — # ii
ii
ii
© Beech Stave Press, Inc. All rights reserved.
No part of this publication may be reproduced, translated, stored in a retrieval system,or transmitted in any form or by any means, electronic, mechanical, photocopying,recording or otherwise, without prior written permission from the publisher.
Typeset with LATEX using the Galliard typeface designed by Matthew Carter andGreek Old Face by Ralph Hancock. The typeface on the cover is Garland by StevePeter.
Library of Congress Cataloging-in-Publication Data
ISBN ---- (alk. paper)
Printed in the United States of America
ii
“TOC” — // — : — page v — # ii
ii
ii
Table of ContentsSJSJSJSJSJSJSJSJSJSJSJSJSJSJSJSJSJSJSJSJSJSJ
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viiBibliography of Stephanie W. Jamison . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ixList of Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxi
Gary Beckman, The Role of Vassal Treaties in the Maintenance of theHittite Empire . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Joel P. Brereton, The Births of the Gods and the Kindling of Firein ˚Rgveda . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Andrew Miles Byrd, Schwa Indogermanicum and Compensatory Lengthening . .
George Cardona, A Note on TS ...– . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
George Dunkel, Proto-Indo-Iranian *stríH- and PIE *sór- ‘female, woman’ . . . . . .
James L. Fitzgerald, The Blood of Vr.tra May Be All Around Us. . . . . . . . . . . . . . . .
Bernhard Forssman, Homerisch πρÒκλυτος, avestisch frasruta- . . . . . . . . . . . . . . . . . . .
José Luis García Ramón, Vedic indrotá- in the Ancient Near East and the Shiftof PIE *h2eu˘h1- ‘run’→ Core IE ‘help, favor’ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Dieter Gunkel, The Sanskrit Source of the Tocharian ×-Syllable Meter. . . . . . .
Olav Hackstein, Rhetorical Questions and Negationin Ancient Indo-European Languages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Mark HaleThe Pahlavi and Sanskrit Versions of the Gathas:What Can They Teach Us? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Heinrich Hettrich, Zur Verbalbetonung im ˚Rgveda. . . . . . . . . . . . . . . . . . . . . . . . . . .Hans Henrich Hock, Narrative Linkage in Sanskrit . . . . . . . . . . . . . . . . . . . . . . . . . . .
Jay H. Jasanoff, Vedic stus.é ‘I praise’ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Brian D. Joseph, Gothic Verbal Mood Neutralization Viewed from Sanskrit . . .
Jean Kellens, Observations sur l’intercalation du Hadoxt Nask dans le Yasna . . . .
Sara Kimball, Hittite dapi- ‘all, whole, each’ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Paul Kiparsky, The Agent Suffixes as a Window into Vedic Grammar . . . . . . . . . .
Jared S. Klein, Rigvedic u and Related Forms Elsewhere:A Reassessment Forty Years Later . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Masato Kobayashi, The Attributive Locative in the ˚Rgveda . . . . . . . . . . . . . . . . . . .
v
ii
“TOC” — // — : — page vi — # ii
ii
ii
Table of Contents
Martin Joachim Kümmel, Zur „Vokalisierung“ der Laryngaleim Indoiranischen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Melanie Malzahn, Tudati-presents and the tezzi Principle . . . . . . . . . . . . . . . . . . . . . .
H. Craig Melchert, The Case of the Agent in Anatolianand Proto-Indo-European . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Angelo Mercado, Šahs at the Pass of Thermopylae . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Kanehiro Nishimura, Elision and Prosodic Hiatus between MonosyllabicWords in Plautus and Terence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Alan J. Nussbaum, Replacing locus ‘place’ in Latin locuples . . . . . . . . . . . . . . . . . . . . .
Thomas Oberlies, „Und von ferne sah ich den Rauch des Pferdedungs“:Zum „Rätsellied“ RV . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Patrick Olivelle, Judges and Courts in Ancient India:On dharmastha and prad. vivaka . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Lisi Oliver†, Old English Riddles, Comparative Poetics,and the Authorship of Beowulf . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Asko Parpola, Rudra: ‘Red’ and ‘Cry’ in the Nameof the Young God of Fire, Rising Sun, and War . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Martin Peters, Rebels without a Causative . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Theodore N. Proferes, The Mımam. sa Influence on the Formationof the Bhagavadgıta . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Jeremy Rau, Ancient Greek φε�δοµαι . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Elisabeth Rieken, Hittite ukturi: A “Thorny” Problem in Anatolian . . . . . . . . . . . .
Don Ringe, Phonological Rules and Dialect Geography in Ancient Greek. . . . . . .
Gregory Schopen, A Tough-talking Nun and Women’s Languagein a Buddhist Monastic Code . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Nicholas Sims-Williams, Iranian Cognates of Vedic sásvant- and -sás . . . . . . . . . . .
Prods Oktor Skjærvø, Justice in Khotan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Elizabeth Tucker, Avestan fraspaiiaox eδra- and an Indo-Iranian Termfor a Ritual Girdle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Ana Vegas Sansalvador, Iranian Anahita- and Greek Artemis:Three Significant Coincidences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Aurelijus Vijunas, Vedic ketú- ‘brightness’ Revisited: Some AdditionalConsiderations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Brent Vine, On the Vedic Denominative Type putrıyánt- . . . . . . . . . . . . . . . . . . . . . . .
Michael Weiss, “Sleep” in Latin and Indo-European:On the Non-verbal Origin of Latin sopio . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Martin L. West†, So What Is It to Be? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Kazuhiko Yoshida, Hittite Mediopassives in -atta . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Index Verborum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
vi
ii
“Gunkel” — // — : — page — # ii
ii
ii
The Sanskrit Sourceof the Tocharian ×-Syllable MeterSJSJSJSJSJSJSJSJSJSJSJSJSJSJSJSJSJSJSJSJSJSJ
Tocharian meter as an indigenous traditionWere Tocharian meters influenced by Indic meters, and if so, to what extent? Since theoutset of Tocharian studies, the prevalent opinion has been that the Tocharian met-rical tradition, which is shared by both Tocharian languages, is independent of theIndian tradition. In Sieg and Siegling’s original formulation, “Die tocharische Metrikscheint selbständig dazustehen und nicht der indischen entlehnt zu sein” (:x).
The supposed independence of the metrical form of Tocharian poetry may seem sur-prising given that the poetic texts are translations and adaptations of Buddhist San-skrit originals. Furthermore, the Tocharian Buddhists did adopt the form of narrationknown as campu, in which prose and verse alternate. However, Tibetan shows thatit is possible to retain and modify indigenous meters for the translation/adaptation ofSanskrit texts.
Two distinctive differences between Sanskrit and Tocharian meter are taken toadvocate the latter’s independence. First, Tocharian meter does not seem to regulatesyllable weight. Second, Tocharian verse-internal cola, i.e. the metrical units delimitedby caesurae, are only – syllables long, which is shorter on average than Sanskritcola, to judge from the traditional metrical treatises. For example, four of the fivemost common Tocharian meters are matched with respect to verse length in syllablesby eight relatively common Sanskrit meters. The average Tocharian colon length is
See also Watkins : (“There is no obvious external contact source for Tocharian meter, the systemof isosyllabic verse lines rigorously divided into even or uneven cola, and organized into four-line stanzas”)and Pinault : (“Le système de versification, en dépit des termes d’origine indienne, est totalementétranger à celui de la métrique du sanskrit”).
On the Tocharian adoption of campu, see Pinault :.“Relatively common” is here defined as belonging both to Velankar’s (b) “list of metres used for
continued narration” and Hahn’s () list of “the most frequently used Sanskrit meters.” For the cola ofthe Tocharian meters, which have strophes of verses of syllables (“×”), ×, ×, and ×, seePinault :; Bross, Gunkel, and Ryan ; and Peyrot forthcoming. Velankar (a) gives the colaof the Sanskrit comparanda, i.e. the ×-syllable Drutavilambita, Pramitaks.ara, and Vam. sastha, the ×
Vasantatilaka, the × Malinı, and the × Narkut.aka, Mandakranta, and Harin. ı.
ii
“Gunkel” — // — : — page — # ii
ii
ii
The Sanskrit Source of the Tocharian ×-Syllable Meter
. syllables versus . in Sanskrit. The second difference has emerged more clearlyas further caesurae have been identified in various Tocharian meters. Let me brieflyillustrate this and introduce the Tocharian ×.
The Tocharian ×
Sieg and Siegling (:x–xi) first described the Tocharian meter whose stanza consistsof four metrically identical -syllable verses/padas, which they dubbed the “×.” Itis one of the best-attested meters in both Tocharian A and B. In a sample of ,
padas of Tocharian B poetry drawn from CEToM, it is the best-attested meter bysyllable, making up .% of the sample. Sieg and Siegling identified caesurae (|)after the th, th, and th syllables (σ) of the verse/pada:
σ σ σ σ σ | σ σ σ σ σ | σ σ σ σ σ σ σ σ | σ σ σ σ σ σ σ
According to their description, the pada consisted of four cola of various lengths,which we can represent shorthand as |||. Stumpf (:–) identified two fur-ther caesurae after syllables and . These are sometimes considered to be “minor”or “secondary” caesurae (|) that divide cola into subcola (e.g. Pinault :–),though the diagnostics for their minor status are partly problematic (see below):
σ σ σ σ σ | σ σ σ σ σ | σ σ σ σ | σ σ σ σ | σ σ σ σ | σ σ σ = |||||
Stanza of the story of the nun Sundarı (THT a– + THT a–b), fromthe eighth book (the Vacavarga) of the Tocharian B Udanalankara, recently editedwith translation and commentary in Hackstein, Habata, and Bross , exemplifiesthe meter. Padas are printed on two lines for typographical reasons, caesurae andjunctures involving clitics (-, +) are marked, and restorations and emendations arein parentheses and square brackets, respectively. The pada-final punctuation and thestanza-final numeral are original, in accordance with the Tocharian scribes’ usual prac-tice, which facilitates the identification of the meter in more fragmentary contexts.
su temeñ sraukam. | nraine tänmastär |
maka lykwarwa | maka cmela | maka lkas.s.äm. | läklenta :maka pudñäkti | tsankam. sais.s.ene |
s.ärpsentär-ne | alyauce+ka | nrais.s.e wnolme | tallantä :spelke sompastär | krentats su wnolmets |
naki welñe | preresa ceu | aunas.s.än-me | arañcne :
If we disregard “minor” caesurae (§), the Tocharian average is ..The ten best attested meters by syllable in the TB sample are × (.%), × (.%), × with
++-syllable cola (.%), × (.%), × (.%), × (.%), /// (.%), ////
(.%), × with ++-syllable cola (.%), and /// (.%).
ii
“Gunkel” — // — : — page — # ii
ii
ii
Dieter Gunkel
aisamñe spakta(m. ) | slek+ompalskoññe |
cowai ram+no | tärkana[m. -m]e | pälskos.s.ana | krentauna
[Richtet jemand gegen einen solchen Tadel und Verleumdung, die gegen-standslos ist,] und stirbt dieser deswegen, so wird er in der Hölle wiederge-boren, viele Male, und erlebt viele Wiedergeburten und Leiden. Erheben sichviele Buddhas auf der Welt, so weisen sie einander auf dieses unglückliche Höl-lenwesen hin. Den Eifer raubt ein solcher den guten Wesen. Mit diesem Pfeildes Tadelaussprechens trifft er sie (die Guten) ins Herz. Wissen, religiösenDienst sowie die Fähigkeit zur geistigen Versenkung (Meditation) raubt er ih-nen gleichsam, und somit die geistigen Güter. (Hackstein, Habata, and Bross:)
The caesurae are quite strict, as can be seen from Figure , which plots the incidenceof verse-internal word boundaries in the ×. The TB data are based on a corpus of partly fragmentary padas drawn from seventeen texts. The TA data are based ona smaller corpus of forty non-fragmentary padas drawn from fifteen texts. Juncturespreceding enclitics and following proclitics are not counted as boundaries. The inci-dence of boundaries is predictably similar: a Spearman’s rank-order correlation yieldsa strong, positive correlation (ρ = .), which is statistically significant (p < .).
The overlapping error bars also give a sense of how insignificant the differences be-tween the two data sets are. The high incidence of word boundary after syllables ,, , , and reflect the caesurae.
PK AS E ( padas), D (), E (), F (), G (), H (), THT (), (), (), (),
(), (), (), (), (), (), ().A (), (), (), (), (), (), (), (), (), (), (), (), ,
+ ().Following Bross, Gunkel, and Ryan : n. and Koller , I treat the following forms as clitic.
Enclitic in TA: aci ‘(starting) from’; the modal particle assi; the emphatic particles ats, atsam. , atsek; thecomparative particle oki ‘like’; the negative polarity item ontam. ; the ablative and allative pronouns anäs. andanac; the conjunctions skam. ‘and, also’ and nu ‘and, but’; the disjunction pat ‘or’; the focus particles pe andpenu; and the relative particle ne. Proclitic in TA: the prepositions sla ‘with’ and sne ‘without’. Enclitic inTB: the emphatic particles ka, nai, nta, pi, ra, tsa; the comparative particle ram(t); the indefinite pronounskca, ksa; the conjunctions no ‘and, but’, s.pä/s.äp/s.p/s. ‘and, also’, wat ‘or’, and wa ‘therefore’; ñke ‘now’, tne‘here(upon)’; and the forms of ‘to be’ s.ai/s.ey and ste. Proclitic in TB: the prepositions sle ‘with’ and snai‘without’.
Running this on the number of boundaries as opposed to the percentages is slightly problematic: theTB data contains fragmentary verses, so the total number of boundaries per metrical position ranges from (most lacunae) to (least lacunae). Nevertheless, this yields similar values (ρ = ., p < .).
These are %-confidence Clopper-Pearson intervals for proportions (Clopper and Pearson ).
ii
“Gunkel” — // — : — page — # ii
ii
ii
The Sanskrit Source of the Tocharian ×-Syllable Meter
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24TB 28.6 51.3 44.7 10.3 100 15.1 59.9 38.1 4.4 99.3 18.9 40 6.8 96.8 21.9 32.1 11.7 100 20.6 47.5 7.9 96.3 6.8 1.2TA 35 42.5 32.5 20 100 27.5 52.5 45 17.5 100 20 55 17.5 100 22.5 47.5 10 97.5 15 55 17.5 97.5 27.5 15
0
20
40
60
80
100Bo
unda
ry in
cide
nce (
%)
Figure . Boundary incidence in the Tocharian B and A ×.
A case can be made for the minor status of the caesurae after syllables and .As discussed by Bross, Gunkel, and Ryan (:–), the violability of caesurae iscurrently the only reliable way to distinguish between major and minor caesurae.
While the poets realize the caesurae after syllables and at least % of the timein both TA and TB ×, they are nevertheless violated × as frequently as the other,major caesurae, and the difference is statistically significant (Fisher’s Exact Test p =.).
Respected Violated Nprimary (after σ , , ) (.%) (.%)
secondary (after σ , ) (%) (%)
total caesurae =
Caesura violability thus supports their minor status and points to some hierarchicalstructure ()()(|)(|). In the four Tocharian meters that have been closely studied,all and only the caesurae between - and -syllable cola are minor as diagnosed byviolability (Bross, Gunkel, and Ryan :–). The × provides a further case of(|) and supports the existence of the (|) cola described, for example, in Pinault and Peyrot forthcoming:
×: ||
×: |||
×: |||
×: |||
|
×: |||||
Studies of the alignment of syntax and meter are an obvious desideratum.
ii
“Gunkel” — // — : — page — # ii
ii
ii
Dieter Gunkel
Comparing prose constituents with the verse cola of the first four meters listedabove, Bross, Gunkel, and Ryan () find that there are fewer boundaries thanexpected in colon- and some verse-penultimate positions, apparently reflecting theavoidance of constituent-final monosyllables. In the more robust TB data plotted inFigure , the two points with the lowest boundary incidence are after the th (.%)and th (.%) syllables, which may indicate half-verse (h) and verse (v). If so, theconstituency of the × would be
( ( ( σσσσσ ) ( σσσσσ ) )h ( ( ( σσσσ ) ( σσσσ ) ) ( ( σσσσ ) ( σσσ ) ) )h )v .
The motivation for the organization of the cola is not immediately clear. I will arguebelow that the caesurae are carried over from the Krauñcapada, where their locationis clearly motivated.
Tocharian meter as influenced by SanskritHaving provided the × with a fuller description, let us return to the scholarshipregarding the relationship between Sanskrit and Tocharian meter. Widmer ()challenged the independent status of Tocharian meter, pointing out that aside fromthe non-regulation of weight, most Tocharian meters are structurally like the Sanskritsamav ˚rttas, whose stanzas consist of four metrically identical, isosyllabic verses/padas(catus.padı) that are often further articulated by caesurae. Widmer compared the struc-ture of four Sanskrit samav ˚rttas with four Tocharian meters, claiming that they corre-spond both in syllable count and in the location of the caesurae, which I refer to as the“colometry.” Since then, it has become clear that two of the comparisons cannot beupheld, since they are based on a frequently cited but empirically unfounded analysisof the Tocharian × (Bross, Gunkel, and Ryan ). The remaining comparisonsare between the Tocharian × and the Sanskrit Candravarta, which I cannot discusshere, and the Tocharian × and the Sanskrit Krauñcapada; note that Widmer citesthe colometry of the latter pair as “|||,” i.e. without reference to Stumpf’s (minor)caesurae.
While Widmer’s study has had some resonance (cf. Pinault :–), the com-parisons above have apparently not convinced scholars that the Tocharian meters inquestion were borrowed/adopted from Sanskrit. I suspect this is because the follow-ing two questions have not yet been addressed. First, were the Tocharians familiarwith meters like the Krauñcapada, which are described in metrical treatises but rarelyor never attested in Sanskrit poetry? Velankar (b) gives a sense of just how rarethe Krauñcapada is. The author compared the meters listed in ten metrical treatises
It is possible that the shorter -syllable subcolon in the (|) units may have a clausular function, markingverse-end in the × and ×, and both half-verse- and verse-end in the × (cf. the concept of “saliency”in Hayes and MacEachern ; Kiparsky ; Ollett ), but it is not clear whether an analysis alongthese lines could be extended to the × and ×, let alone the other twenty-five or so Tocharian meters.
ii
“Gunkel” — // — : — page — # ii
ii
ii
The Sanskrit Source of the Tocharian ×-Syllable Meter
with the actual poetic practice in twenty-eight Mahakavis of the ancient and medievalperiods and found that of the samav ˚rttas of the catus.padı (×n) type listed inthe treatises, the poets only used only frequently, and another “for a changeand ornamentation” (). The Krauñcapada is not among those . And second, whydoes the × have two more caesurae than the Krauñcapada? Let us take a closerlook at the attestation and structure of the Krauñcapada.
The Krauñcapada
The Sanskrit metrical treatises describe the Krauñcapada as a ×-syllable meter witha ||| colometry and the following rhythm, beginning with Pingala’s Chandah. sastra(.):
– ™™– – | – ™™– – | ™™™™™™™™| ™™™™™™–
Outside the metrical treatises, the Krauñcapada is to my knowledge only attestedonce, namely in five surviving stanzas of a buddhastotra discovered in the “Rotkuppel-raum” of the Kizil caves (CEToM’s “Qizil Miµ-Öy”) during the third German Turfanexpedition (–). The stanzas, numbered – in the manuscript, were editedand translated in by Schlingloff, who dubbed the stotra Preis der BekehrungenBuddhas.
After the Buddha achieves release (vimoks.a) and brings tranquility (upasama) to thegood (stanza ), the poet dedicates a stanza to each of four miraculous conversions,that of the demon Ad. avaka (), the snake king Apalala (), the elephant lord Nalagiri(), and the finger-collecting serial killer Angulimala (). The conversion of Nalagiri,depicted here as a rampaging bull elephant in musth, should give a sense of the me-ter and the quality of the poem. Schlingloff’s conjectures and emendations are givenin parentheses and square brackets, respectively. I have marked caesurae, clitic-hostjunctures (+), and compound boundaries (-), which the poet treats as (or much like)word boundaries with respect to caesurae (§):
ros.a-viv ˚rtta-|vyakula-d ˚rs.t.ih.|
pravis ˚rta-mada-kat.a-|vilud. ita-vadanah.sonita-digdh[o] | bhranta-karagro |
hata-nara-sirasi-ja-|parigata-dasanah. |yena vinıto | raja-g ˚rhe ’sau |
mimathis.ur arir+iva | gaja-patir asivastasya mahars.eh. | sasana[m a]gry[am. ] |
vina(yatu) mama mati(m | iha) mati-manasah. || ||
krauñcapada bhmau sbhau nau nau ga bhutendriyavasv ˚rs.ayah. .On the expeditions, see Fellner .
ii
“Gunkel” — // — : — page — # ii
ii
ii
Dieter Gunkel
His eyes were confused and rolled with rage, his face was besmirched by musth-secretion that flowed down from the temples, he was blood-smeared, the tipof his trunk swayed unpredictably, his tusks were garlanded by the hair of deadmen—the one who tamed that baleful lord of elephants in Rajag ˚rha as he wasseeking to crush (everything) like an enemy, let the prime teaching of thatgreat sage, whose mind is thoughtful, tame my thinking here.
Regarding the style of the poem, Schlingloff wrote (:):
Obwohl die chronologische Bestimmung unbekannter Werke allein nachstilistischen Gesichtspunkten sehr zweifelhaft ist, wird man doch mit eini-gem Vorbehalt sagen können, daß die Dichter unserer Hymnen von demDichterkreis um Asvaghos.a [nd c. ] und Mat ˚rcet.a [pre-th c. ]zeitlich nicht sehr entfernt sind.
Since there is no other evidence for the existence of this text, it is impossible to sayanything certain about its provenance. It may have been composed in India as earlyas the nd c. , transmitted along the Silk Road to Central Asia, and preserved inKizil. It is not out of the question, however, that a Central Asian Buddhist who stud-ied Sanskrit grammar, meter, and early Buddhist kavya poetry composed the text.
The birchbark fragments known as the Turfan Chandoviciti (Schlingloff ), whichcontain a collection of Sanskrit verses exemplifying various meters, demonstrate thatBuddhists in East Turkestan were studying Sanskrit meter as early as the th/th c.. Given the find spot and the fact that the manuscript exhibits the occasional con-fusion of vowel length and voicing among stops that is typical of Sanskrit texts fromthe region, e.g. ad.avakam. for ad.avakam. ‘Al.avaka’ (c) and pannakarajam. for panna-garajam. ‘king of snakes’ (c), it is reasonably likely that the copyist spoke Tocharian,which had neither phonological contrast.
The location of the caesurae in the Krauñcapada is clearly motivated. The caesuraafter the tenth syllable, which marks the palpable rhythmic transition from the twoadonics to the extended stretch of light syllables, divides the pada into balancedsixteen-mora half-lines. The other two caesurae divide the half-lines into eight-moraquarter-lines. The symmetry suggests the following hierarchical organization:
( ( (– ™™– –)µ (– ™™– –)µ )µ ( (™™™™™™™™)µ (™™™™™™–)µ )µ )µ
The identity of the first half-line with the Rukmavatı meter and the second with the
On the question of the provenance of stotras attested only in Central Asian manuscripts, see Hartmann: n. .
For a recent discussion of the nature and dating of the fragments, see Chen .Wolfgang Krause already suggested to Schlingloff that the confusion among stops could be attributed
to a Tocharian scribe (Schlingloff : n. ), and he surely would have suggested the same for the vowelshad he thought, as we do now, that Tocharian lacked phonemic vowel length.
ii
“Gunkel” — // — : — page — # ii
ii
ii
The Sanskrit Source of the Tocharian ×-Syllable Meter
Man. igun. anikara invites further analysis of the Krauñcapada, along the lines of Deo, as a syzygy of two trochaic tetrameters. The meter is binary and trochaic (i.e.rhythmically Strong-Weak) at every level: each position is bimoraic, each foot con-tains two such positions, each dipody (=quarter-line) contains two feet, etc.:
W
W
W
WS
S
WS
S
W
WS
S
WS
S
W
W
WS
S
WS
S
W
WS
S
WS
– ™™ – – | – ™™– – | ™™ ™™™™ ™™|™™ ™™™™ –
If the Krauñcapada, like the ×, had caesurae after the th and nd syllables, theywould divide the last two dipodies (quarter-lines) into their constituent feet (eighth-lines).
Boundaries in the Preis der Bekehrungen Buddhas
Since the metrical treatises do not necessarily reflect the poets’ treatment of caesuraeexactly, and since the Tocharians may well have learned the Krauñcapada from actualpoetic texts such as the Preis der Bekehrungen Buddhas (PdBB), we should examine theword boundary distribution in that stotra. As noted above, a number of the caesuraecoincide with compound boundaries. I have counted compound boundaries betweeninflectable stems as word boundaries. This conforms to poetic practice and to theprescriptions of the metricians. Note that the use of Schlingloff’s emendations andconjectures for the word-boundary distribution is less problematic than it may seemat first glance. For example, in b
sonitadigdh[o] | bhranta-karagro
the manuscript reads °digdhoh. . Schlingloff entertains the emendation printed as wellas a single compound sonitadigdha-|bhranta-karagro. With respect to boundaries, the
Cf. Steiner : for regular caesurae in Anus.t.ubh Vipulas that are nowhere noted in the treatises.Steiner (:–) provides a clear, concise discussion of Halayudha’s definition of caesura (yati)
in the Yatyupadesopanis.ad as well as a study of what appear to be regular exceptions to that definition inHars.adeva’s poetic practice, e.g. caesura between prefix and stem, explicitly forbidden by Halayudha. Sincethe exceptions appear to be less metrically felicitous and may require particular pragmatic motivation insome cases (), I have not counted them here. I also consider the realization of caesurae before and af-ter vowels fused across compound boundaries to be less felicitious, so I have treated bhrantakaragro asbhranta-karagro, as opposed to bhranta-kara-gro, bhranta-kar-agro, or bhranta-kar-a-gro. The same holds forprasphuritaus.t.ham. (spanning syllables – in a) and °mukhagnim (– in a).
ii
“Gunkel” — // — : — page — # ii
ii
ii
Dieter Gunkel
two options are equivalent. In d, one may take issue with Schlingloff’s conjecture ofthe aks.aras miha and the supposed compound mati-manasah. ‘einsichtig’:
vina(yatu) mama mati(m | iha) mati-manasah.
In my view, the parsing of matiCVCVmatimanasah. as mati-CVCVmati-manasah. ormatiC VCVmati-manasah. is extremely likely, and the more difficult choice betweenthe printed text and, for example, mati-CVCVm ati-manasah. makes no difference forthe tally. Figure plots the boundary incidence in the TB × and the Krauñcapada.As above, host-clitic junctures do not count as boundaries.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24TB 28.6 51.3 44.7 10.3 100 15.1 59.9 38.1 4.4 99.3 18.9 40 6.8 96.8 21.9 32.1 11.7 100 20.6 47.5 7.9 96.3 6.8 1.2K 15.8 52.6 26.3 5.3 100 5.6 55.6 44.4 5.6 100 0 35.3 0 100 5.9 35.3 5.6 100 0 33.3 29.4 70.6 26.3 0
0
20
40
60
80
100
Boun
dary
inci
denc
e (%
)
Figure . Boundary incidence in the TB × and the Krauñcapada.
Strikingly, the correlation between the two is virtually as strong as the correlationbetween the TB and TA data: Spearman’s ρ = ., p < .. The most importantfact reflected in the plot is the high boundary incidence after the th (%) andnd (.%) syllables, which correspond to Stumpf’s (minor) caesurae. Thus whilethe metrical treatises do not prescribe caesurae in those positions, the actual poeticpractice of the PdBB provides the basis for all five Tocharian caesurae and points toborrowing.
The high boundary incidence after syllables and does not necessarily reflectcaesurae in the Krauñcapada. The peaks in the plot there could be “caesurals,” bywhich I mean byproducts of the other caesurae, the rhythm required in that stretchof the meter, the shape of Sanskrit lexical items, and other aspects of the grammar. Inorder to address the question and by extension the accuracy of the treatises, we wouldrequire more poetry composed in Krauñcapada as well as Sanskrit prose passages that
In practice, these are aham™iti b, imam. ™hi c, prapya™ca, bhis.ag™iva d, aham™iha d, (gha)na[m]™iva a,tarum™iva b, yo™hy c, arir™iva c, °pa(tir™iva) b, gaja™iva c, and nabhasi™ca d.
Running this on the number of boundaries (see the caveat above, n. ) again yields similar values (ρ =., p < .).
ii
“Gunkel” — // — : — page — # ii
ii
ii
The Sanskrit Source of the Tocharian ×-Syllable Meter
happen to have comparable stretches of light syllables. The issue has no consequencesfor the borrowing scenario: if the Tocharians learned the Krauñcapada from the PdBBand/or comparable texts, they had access to the surface form of the meter (boundarydistribution), not to its grammar (caesurae).
Tune or meter names (kenes)Tocharian metrical passages are preceded by a term in the locative or perlative sin-gular, which is usually set off by double dan. d. as. For example, on the wooden tabletfrom the Kizil caves recently published by Ogihara (), || apratitulyem. ne || precedesTB verses in the × praising a local monastery. Sieg and Siegling () referred tothe terms as meter names. The majority are of Sanskrit origin, but so far, only oneof the ca. terms (B harin. aplutne, A harin. aplutam. ) has been identified with a San-skrit meter name (Malzahn ). Since Winter , the prevalent opinion holds thatthey do not refer (only) to the meter, but (also) to an aspect of the performance ofthe poetry, perhaps the tune. Winter’s reasons for this were that () the same me-ter is often associated with a number of different terms and () occasionally the sameterm is associated with more than one meter. Furthermore, in archaic texts, the termis occasionally followed by kenene, the locative singular of TB kene (: A kam. ), whichmay be cognate with Latin cano ‘sing’, Old Irish canu, the Hesychius gloss ºικανÒς:
Ð ¢λεκτρυèν (‘dawn-singer’, i.e. the cock), etc. and mean ‘tune, melody’ (Winter ,Watkins ). To borrow a TB example from Watkins’ admirably clear discussionof the kenes (:–), THT b preserves nis.kramam. kenene, which Watkinstranslates “in nis.kramant- [Skt. ‘departing’] melody” ().
Several experts have recently been working on the kenes (Malzahn , Peyrotforthcoming), and the results will no doubt elucidate the phenomenon. Here I wishto make one observation. The × is associated with at least three kenes that are com-mon to both TB and TA: TB apratitulyem. ne (: A apratitulyenam. ), bahudantakne (: Abahudantakam. ), and bahupraharne (: A bahuprakaram. ), as well as three further kenesattested only in TA, arsi-lañcinam. , watañi-lantam. , and s.erasi-nis.kramantam. . I find itstriking that the kene derived from Sanskrit apratitulya-, which apparently means ‘notto be compared, incomparable’ (cf. SWTF s.v.), scans – ™™– × and thus fits the be-ginning of the Krauñcapada pada, as do many Sanskrit meter names including Krauñ-capada itself. (The beginning of a popular verse was used as a mnemonic and eventu-ally as the name of the meter.) While the scansion of apratitulya- may be accidental,I cannot help but wonder whether a popular Krauñcapada verse that began aprati-tuly° was the source of the kene. To be sure, this is not the only source of kenes. Forexample, arsi-lañcinam. means either ‘in the (tune) of Tocharian A kings’ or ‘in theTocharian A (tune) of kings’ (Watkins , Peyrot forthcoming). Whether it is oneof the sources should be easier to gauge with the results of the studies noted above.
For a different view, see Widmer : n. .
ii
“Gunkel” — // — : — page — # ii
ii
ii
Dieter Gunkel
The borrowingI would like to suggest the following borrowing scenario with all due caution. Anerudite Tocharian Buddhist who studied Sanskrit grammar and meter and had a tastefor Buddhist kavya poetry adopted the Krauñcapada for his or her Tocharian compo-sitions. Given the general popularity of buddhastotras in Central Asia, the fact thatparticular stotras enjoyed local popularity (Hartmann :–), and the rarity ofthe Krauñcapada, the PdBB may have been one of the source texts. While the ×
is used for buddhastotras, e.g. to translate and adapt Mat ˚rceta’s Varn. arhavarn. astotrainto TA, the meter is by no means restricted to the genre, nor is it rare. I mustassume that () the Sanskrit source texts belonged to multiple genres and/or () thegenre-specificity of the meter was lost as it gained popularity in the Tocharian tradi-tion.
The boundary incidence in the ×, specifically the greater violability of the mi-nor caesurae and the apparent bridges after the th and th syllables, suggest that(s)he not only borrowed the caesurae but also carried over the constituency of themeter, which was perfectly transparent in the weight-regulating Krauñcapada, withits moraically balanced half- and quarter-verses. The fact that the constituency contin-ued to be transmitted is somewhat surprising, since it became opaque when syllable-weight regulation was given up.
( ( (–™™– –)µ (–™™– –)µ)µ ( (™™™™|™™™™)µ (™™™™|™™–)µ )µ )µ
( ( (σ σ σ σ σ)σ (σ σ σ σ σ)σ )σ ( (σσσσ|σσσσ)σ (σσσσ| σσσ)σ )σ )σ
It is possible that the manner of recitation of the Tocharian ×, which is presumablyreferred to by the kenes, facilitated the retention of the constituency.
The reason for giving up weight regulation should probably be sought in Tochar-ian phonology. While counterexamples exist, Gordon (:) cautiously ob-serves that of the seventeen languages with weight-sensitive metrical traditions in hissurvey, sixteen have a phonemic distinction in vowel length, and all seventeen treatCVV(C) and CVC syllables as heavy in meter, as in Classical Sanskrit. Furthermore,stress tends to agree with the meter in treating those syllable types as heavy. In contrastto Sanskrit, neither Tocharian language has phonemic vowel length. The stress sys-tem of Tocharian A treats non-high vowels as heavy and high vowels as light (Nevinsand Plaster ). Tocharian B stress is weight-insensitive, but stress assignment in
This Tocharian may also have considered rare, long, ornate meters to be particularly well-suited forpraising the Buddha (cf. Hahn :). With the exception of the PdBB, Hahn’s examples are later stotras.
Cf. Hartmann and Pinault :–, both with further references. The latter, together withthe following chapter (–), provides an insightful exemplification and discussion of issues of Tochariantranslation and adaptation of Sanskrit buddhastotras.
The counterexample in Gordon’s study is Berber. Paul Kiparsky informs me that Ottoman Turkish,which did not have contrastive vowel length, borrowed the Persian/Urdu quantitative tradition.
ii
“Gunkel” — // — : — page — # ii
ii
ii
The Sanskrit Source of the Tocharian ×-Syllable Meter
certain morphological categories (e.g. class I subjunctives of the type tékäm. -me ‘willtouch’) and individual lexical items (e.g. patär ‘father’) point to a similar prehistoricdistinction between non-high and high vowels (Malzahn :–, –; Jasanoff
). I suggest that Tocharian phonology did not make the kinds of distinctions thatfacilitate the development, borrowing, or retention of quantitative meters.
Concluding remarksAccepting that the Krauñcapada is the source of the ×, we can draw several conclu-sions. First, it speaks for the utility and accuracy of the quantitative corpus-linguisticmethods developed for the analysis of Tocharian meter in Bross, Gunkel, and Ryan and and applied to the × here. Second, it shows that even rare San-skrit meters were candidates for borrowing into Tocharian. Third, the similarity be-tween the × and other Tocharian meters suggests that the puzzling colometriesof Tocharian meter, i.e. the seemingly unmotivated combinations of even and un-even cola, may have arisen via borrowing and the loss of syllable weight regulation.
Regarding the last point, however, I wish to stress that it remains possible that an in-digenous system with those characteristics was already in place when the Tochariansborrowed the Krauñcapada.
AbbreviationsCEToM = A Comprehensive Edition of Tocharian Manuscripts. Accessed –. http://
www.univie.ac.at/tocharian/SWTF = Bechert, Heinz (ed.). . Sanskritwörterbuch der buddhistischen Texte aus
den Turfanfunden. Vol. : Vokale. Göttingen: Vandenhoeck & Ruprecht.
ReferencesBross, Christoph, Dieter Gunkel, and Kevin M. Ryan. . “Caesurae, bridges, and
the colometry of four Tocharian B meters.” Indo-European Linguistics :–.———. . “The colometry of Tocharian ×-syllable verse.” In Malzahn et al.,
–. Bremen: Hempen.
I say “facilitate” because Old Javanese kakawin (– c. CE) shows that poets of a language that probablylacked phonemic vowel length could borrow Sanskrit syllabo-quantitative and mora-counting meters as wellas the Sanskrit-type distinction between light and heavy syllables. It is not clear to me whether Old Javanesehad a contrast between heavy and light syllables prior to the borrowing, and if so, whether it was differentfrom the Sanskrit-type distinction. I am grateful to Andrew Ollett for pointing this out to me and to ArloGriffiths for answering a number of questions about Old Javanese. For some basic information about thetradition and its relationship to Sanskrit, see Pollock :–.
Mordvin meter, if cognate with Kalevala meter (Kiparsky ), provides the closest parallel I am awareof.
ii
“Gunkel” — // — : — page — # ii
ii
ii
Dieter Gunkel
Chen, Ruixuan. . “Bemerkungen zur Turfan-Chandoviciti.” Indo-Iranian Journal:–.
Clopper, C. J., and Egon S. Pearson. . “The use of confidence or fiducial limitsillustrated in the case of the binomial.” Biometrika :–.
Deo, Ashwini S. . “The metrical organization of Classical Sanskrit verse.” Journalof Linguistics :–.
Fellner, Hannes A. . “The expeditions to Tocharistan.” In Instrumenta Tocharica,ed. Melanie Malzahn, –. Heidelberg: Winter.
Gordon, Matthew Kelley. . Syllable Weight: Phonetics, Phonology, Typology. NewYork: Routledge.
Hackstein, Olav, Hiromi Habata, and Christoph Bross. . “Tocharische Texte zurBuddhalegende I: Die Geschichte von der Nonne Sundarı. B–.” MünchenerStudien zur Sprachwissenschaft :–.
Hahn, Michael. . “Sanskrittexte aus dem tibetischen Tanjur (I): Das Nagarjunazugeschriebene Dan. d. akav ˚rttastotra.” Berliner Indologische Studien :–.[Corrected version, accessed January , . https://uni-marburg.academia.edu/MichaelHahn]
———. . “A Brief Introduction into the Indian Metrical System for Students.”Accessed November , . https://uni-marburg.academia.edu/MichaelHahn
Hartmann, Jens-Uwe. . Das Varn. arhavarn. astotra des Mat ˚rcet.a. Göttingen: Van-denhoeck & Ruprecht.
———. . “Der Pran. amastava. Ein Buddhastotra aus den Gilgit-Funden.” InBauddhavidyasudhakarah. : Studies in Honour of Heinz Bechert on the Occasion of Histh Birthday, ed. Petra Kieffer-Pülz and Jens-Uwe Hartmann, –. Swisttal-Odendorf: Indica et Tibetica.
Hayes, Bruce, and Margaret MacEachern. . “Quatrain form in English folk verse.”Language :–.
Jasanoff, Jay H. . “The Tocharian B accent.” In Malzahn et al., –.Kiparsky, Paul. . “A modular metrics for folk verse.” In Formal Approaches to
Poetry: Recent Developments in Metrics, ed. B. Elan Dresher and Nila Friedberg,–. Berlin: de Gruyter.
———. . “Catastrophic metrical change: Kalevala and Mordvin meter.” Paper pre-sented at “Frontiers in Comparative Metrics ,” Tallinn University.
Koller, Bernhard. . “Studies in Tocharian phonology above the word-level.” Ph.D.diss., University of California, Los Angeles.
Malzahn, Melanie. . The Tocharian Verbal System. Leiden: Brill.———. . “Written and oral culture in Tocharian—the case of poetry.” Paper pre-
sented at “Transmission of the Buddhist Texts in Central Asia: Tocharian Bud-dhism and Its Role,” Ludwig-Maximilians-Universität Munich.
Malzahn, Melanie, Michaël Peyrot, Hannes Fellner, and Theresa-Susanna Illés (eds.).
ii
“Gunkel” — // — : — page — # ii
ii
ii
The Sanskrit Source of the Tocharian ×-Syllable Meter
. Tocharian Texts in Context: International Conference on Tocharian Manuscriptsand Silk Road Culture, June –th, . Bremen: Hempen.
Nevins, Andrew, and Keith Plaster. . Review of Paul de Lacy, Markedness: Reduc-tion and Preservation in Phonology (Cambridge, UK: Cambridge University Press,). Journal of Linguistics :–.
Ogihara, Hirotoshi. . “Kuchean verses written on a wooden tablet kept at Xin-jiang Kucha Academy.” In Malzahn et al., –.
Ollett, Andrew. . “Moraic feet in Prakrit metrics: A constraint-based approach.”Transactions of the Philological Society :–.
Peyrot, Michaël. Forthcoming. “A comparison of the Tocharian A and B metrical tra-ditions.” In Language and Meter, ed. Dieter Gunkel and Olav Hackstein. Leiden:Brill.
Pinault, Georges-Jean. . “Narration dramatisée et narration en peinture dans larégion de Kucha.” In La Serinde, terre d’echanges: Art, religion, commerce du Ier auXe siècle, ed. Jean-Pierre Drege, –. Paris: La Documentation française.
———. . Chrestomathie tokharienne: Textes et grammaire. Leuven: Peeters.Schlingloff, Dieter. . Buddhistische Stotras aus ostturkistanischen Sanskrittexten. Ber-
lin: Akademie-Verlag.———. . Chandoviciti; Texte zur Sanskritmetrik. Berlin: Akademie-Verlag.Sieg, Emil, and Wilhelm Siegling. . Tocharische Sprachreste. Vol. : Die Texte. A.
Transcription. Berlin: de Gruyter.Steiner, Roland. . Untersuchungen zu Hars.adevas Nagananda und zum indischen
Schauspiel. Swisttal-Odendorf: Indica et Tibetica.Stumpf, Peter. . Der Gebrauch der Demonstrativ-Pronomina im Tocharischen. Wies-
baden: Harrassowitz.Velankar, H. D. (ed.). a. Jayadaman: A Collection of Ancient Texts on Sanskrit
Prosody and a Classified List of Sanskrit Metres with an Alphabetic Index. Bombay:Haritosha Samiti.
———. b. “Prosodial practice of Sanskrit poets.” Journal of the Bombay Branch ofthe Royal Asiatic Society –:–.
Watkins, Calvert. . “Questions of syntax and meter in Tocharian.” In Composi-tiones Indogermanicae in memoriam Jochem Schindler, ed. Heiner Eichner and HansChristian Luschützky, –. Prague: Enigma.
Widmer, Paul. . “La métrique tokharienne: L’influence indienne et quelquesdéveloppements tokhariens.” In La langue poétique indo-européenne: Actes du Col-loque de travail de la Société des Études Indo-Européennes. Paris, – octobre , ed.Georges-Jean Pinault and Daniel Petit, –. Leuven: Peeters.
Winter, Werner. . “Zur ‘tocharischen’ Metrik.” In Akten des . InternationalenOrientalistenkongresses München , ed. Herbert Franke, –. Wiesbaden: Steiner.