A Critical Edition of S barabh s
4ya ad 1.1.6–
23: S ´ abdanityatv dhikaran
4a
Kei K
ATAOKAIntroduction
The portion I edit in the following is S´ barabh s4ya ad 1.1.6–23, i.e. the so- called s´abdanityatv dhikaran4a (or s´abdanityat dhikaran4a)（1） or s´abd dhikaran4a,（2）
which establishes the eternality of s´abda (sound or speech), in particular of phonemes (varn4a).（3）
Frauwallner 1968 critically reedited S´ barabh s4ya ad 1.1.1–5, including the so-called Vr4ttik ragrantha, which contains the most important philosophical discussions in the S´ barabh s4ya, e.g. with regard to pram n4as. Frauwallnerʼs contribution to our field is enormous. Many ambiguous lines that were a characteristic of the previous, uncritical editions, were for the first time
１ This is a title postulated on the basis of Kumārilaʼs words: S´lokav rttika s´abdanityat- va, v. 283d: s´abdanityatvakalpan ; 356b: iyam4 s´abdanityat ; 362c: s´abdanityatvam tre
pi.
２ This is a title used by Kumārila. S´lokav r ttika s´abdanityatva, v. 355ab:
sam4bandhanityat y h4 kim4 s´abd dhikaran4e bhidh /. Although s´abd dhikaran4a seems to be historically a more authentic title for the present section, I prefer s´abdanitya- tv dhikaran4a. If I would designate the present section as s´abd dhikaran4a and use s´abda for the abbreviation in such a way as “S´lokav rttika s´abda”, some readers might confuse it with the s´abdapariccheda, i.e. the section on s´abdapram n4a that is included under Jaiminis tra 1.1.3–5 (or more precisely 1.1.4a).
３ For modern studies of this section, see, e.g. Frauwallner 1961, Biardeau 1964, DʼSa 1980 and Houben 1995. In particular Biardeau 1964 expounds the present section of the S´ barabh s4ya in detail.
textcritically clarified by Frauwallner. He solved many textual problems by consulting south Indian manuscripts, mainly from Tanjore. He worked on three Tanjore manuscripts (TA, TB and TC) on the basis of a collation-note, dated 1-9-58, prepared by the eminent Mīmām4sā pandit of the Tanjore Maharaja Serfoji Sarasvati Mahal Library, N.S. Devanāthācārya.（4）
The present work is in a sense a continuation of Frauwallnerʼs work. I start from the portion where Frauwallner stopped. But I consult more and different manuscripts than Frauwallner, including manuscripts from Kerala written in Malayalam script.
Editions and manuscripts consulted
The following is a list of the consulted editions and manuscripts.
A S´r majjaiminipran4 tam4 M m m4s dars´anam. Vol. 1. Ed. Subbās´āstrī.
Ānandās´ramasam4skr4tagranthāvalih4, No. 97. Poona: Ānandā- s´ramamudran4ālaya, 1929. pp. 72.4–91.8.
M Br4hat of Prabh kara Mis´ra. Part I. Ed. S.K. Rāmanātha S´āstrī.
Madras: University of Madras, 1934. pp. 257.1–347.5.
I1 A manuscript kept in the British Library, London, No. SAN MS IO 1808 (2139), OIC.NEG.5021. Paper. Devanāgarī. ff. 15v5–19v5.
K1 A manuscript kept in the Oriental Research Institute & Manuscripts Library, Kerala University, Trivandrum, No. 13628 (No. 932). Palm leaf. Malayalam. ff. 11v3–14v3.
K2 A manuscript kept in the Oriental Research Institute & Manuscripts Library, Kerala University, Trivandrum, No. 13619 (No. 3). Palm leaf.
Malayalam. ff. 6r12–7v11.
S1 A manuscript kept in the Sarasvati Bhavan Library of Sampurnananda
4 A copy of the note is kept in the Indological Research Library, University of Wien. So it is clear that Frauwallner checked the Tanjor manuscripts only indirectly, though he does not mention this fact.
Sanskrit University, No. 28947. Paper. Devanāgarī. ff. 18v3–23v7.
S2 A manuscript kept in the Sarasvati Bhavan Library of Sampurnananda Sanskrit University, No. 29015. Paper. Devanāgarī. ff. 10v11–13v5.
T1 A manuscript kept in the Tanjore Maharaja Serfoji Sarasvati Mahal Library, No. 2083 (D 6757). Paper. Devanāgarī. ff. 13v6–17r9. [This is designated as TB in Frauwallner 1968.]
U1 A manuscript kept in the Library of the University of Pennsylvania, Poleman number 3550 (UP number 1954). Paper. Devanāgarī. ff. 13v4 –16v3.
V1 A manuscript kept in the Bharat Kala Bhavan Library, Banaras Hindu University, No. B 1911 (No. 2771). Paper. Devanāgarī. ff. 13v7–16v11.
Previous editions not consulted
Besides A and M that I consulted in this edition, there are a number of other S´ barabh s4ya editions that contain the s´abdanityatv dhikaran4a. Editions that I noticed are the following.（５）
A1 The Aphorisms of the M m m4s by Jaimini with the Commentary of S´abara-sv min. Ed. Mahes´acandra Nyāyaratna. 2 vols. Bibliotheca Indica: A Collection of Oriental Works, New Series, Nos. 44, 85, 95, 101, 115, 142, 154, 174 & 208. Calcutta: The Asiatic Society of Bengal, 1873; Reprint. Osnabrück: Biblio Verlag, 1983.
５ I mainly follow the conventions established in Kataoka 2004 for the following abbreviations of the sources listed there. A10 does not contain the Tarkap da. This is:
The M m m4s Dars´ana of Mahars4i Jaimini with S´ barabh s4ya of S´abaramuni with the Commentaries Tantrav rttika of Kum rila Bhatt44a and his Commentary Ny yasudh of Somes´vara Bhatt44a, Bh s4yavivaran4a of Govind mr4tamuni and Bh vaprak s´ik , the Hind Translation by Mah prabhul la Gosv m. Ed. Mahāprabhulāla Gosvāmī.
Prachyabharati Series, No. 23. Varanasi: Tara Printing Works, 1984, 87, 87, 84(?).
A2 M m m4s dars´anam, Jaiminipran4 tam4 S´r s´abarasv mikr4tena Bh s4yen4a Sahitam. Ed. Jīvānanda Vidyāsāgara. 2 vols. Calcutta, 1883–84.
A3 M m m4s dars´anam. Ed. Ratna Gopāla Bhatt44a. 2 vols. Kās´ī Sanskrit Series, No. 42. Benares, 1910.
A6 M m m4s kos4ah4. Ed. Kevalānanda Sarasvatī. 7 parts. Wai: Prājña Pāt4has´āl4ā Man4d4al4a, 1952–66.
A7 S´r majjaiminipran4 tam4 M m m4s dars´anam. Ed. Kās´īnātha Vāsudeva Abhyam4kara & Gan4es´as´āstrī Am4bādāsa Jos´ī. 7 parts. Ānandā- s´ramamudran4ālaya, 41976, 21970, 21971, 21972, 21973, 21974,
21974.
A8 c r ya-s´abarasv mi-viracitam Jaimin ya-m m m4s -bh s4yam rs4amata-vimars´iny Hind -vy khyay sahitam. Ed. Yudhis4t4hir Mīmām4sak. 7 bhāgas. Bahālgar4h: Rāmlāl Kapūr T4rast4, 21987, 1978, 1980, 1984, 1986, 1990, 1993.
A9 M m m4s -s´ bara-bh s4yam. Ed. Yudhis4t4hir Mīmām4sak. Prathamo Bhāgah4. Bahālgar4h: Rāmlāl Kapūr T4rast4, 1987. [Only up to the third adhyāya.]
A11 M m m4s bh s4yabh s4an4am c ryas´abarasv mibh s4ya Pratham dhy ya Prathamap da T4ippan4am. Laks4mīpuram4 S´rīnivāsācārya. Mysore:
Rājakīya S´ākhā Mudrālaya, 1928. [This is a modern commentary composed by Laks4mīpuram4 S´rīnivāsācār ya that contains the Tarkap da text of the S´ barabh s4ya.]
A12 M m m4s -Dars´anam with S´ bara-Bh s4ya, P da I (Tarkap da) with an Elaborate Hindi Commentar y. Ed. Uma Shankar Sharma.
Chaukhamba Surbharati Granthamala, No. 51. Varanasi: Chaukhamba Surbharati Prakashan, 1980.
I did not attempt to consult all the published editions of the S´ barabh s4ya.
Instead I took two representative editions, i.e. the older edition from Ānandās´rama (that I designated as A4 in Kataoka 2004) and the Madras edition (that I designated as A5 in Kataoka 2004), because the published S´ barabh s4ya
editions, except for the Madras edition, which often contains unique readings for the Tarkap da, tend to be copies of previous editions. They are directly or indirectly based on the first edition A1 published from Calcutta in the series Bibliotheca Indica.（6） Whichever edition one uses, there is no substantial difference except for minor corrections and (typographical) mistakes. See, for example, the following cases:
S´ barabh s4ya ad 1.1.13: tasya *p t4aliputre py anupalambho
*p t4aliputre py] I1K1K2S1S2T1U1V1; p t4aliputres4v A1A2A3A(=A4)M(=A5)A6A7A8 A9A11A12
All editions read p t4aliputres4v, but the alternative reading is supported by S´arabaʼs usage elsewhere (a few lines above): p t4aliputre py upalabhyeta. The reading p t4aliputres4v adopted by A1 is clearly a mistake for p t4aliputre py. And it is most likely that this mistake was simply inherited by later editions.
S´ barabh s4ya ad 1.1.13: sam4yogavibh g n *utp dayanto y vadvegam abhipratis4t4hante
*utp dayanto] M(=A5)I1K1K2S1S2T1U1; utp dayanti A1A2A3A(=A4)A6A7A8A9A11 A12; utp dayante V1
All manuscripts except V1 read utp dayanto, whereas most of the previous editions read utp dayanti. This seems to be a simple mistake that occurred in the first published edition A1 and that has been inherited by the following editions.
S´ barabh s4ya ad 1.1.14: yath gomay ni kuru iti *sam h re
*sam h re] K1K2T1U1pc; sam4h re A1A2M(=A5)A11S1U1acV1; sam4v he A1fnA3A(=A4)A6A7A8A9A12; sam4vyavah re I1; s h re S2
6 See Kataoka 2004 for a more detailed discussion of the textual problems concerning the previous editions of the S´ barabh s4ya.
A1 reads sam4h re and reports a variant reading in a footnote: ʻsam4v heʼ iti k - kr -. The abbreviation k - kr represents k s´y m4 kr tam, i.e. “a manuscript procured by the editor from a Tailangi Pan4d4it at Benares” according to the preface (p. i). Some of the following editions read sam4h re and other editions adopt the variant reading sam4v he. The adopted reading sam h re that is found in the south Indian manuscripts is also suppor ted by the testimony of Pārthasārathimis´ra in the Ny yaratn kara. The adoption of sam h re as the original reading explains well the change from sam h re to sam4h re and from sam4h re to sam4v he. Both sam4h re and sam4v he in the later editions seem to have been inherited from A1 either directly or indirectly without consulting extra manuscripts.
S´ barabh s4ya ad 1.1.18: uccaritam tre hi vinas4t4e s´abde *navo nyo n rtham4
praty yayitum4 s´aknuy t
*navo nyo n rtham4] K2S2T1V1; na c nyo ny n artham4 A1A2A3A(=A4)M(=A5)A6 A7A8A9A11A12; tatv nyo n rtham4 I1; nav nyo n rtham4 K1; na c nyortham4 S1;
naivo nyo hy artham4 U1
The adopted reading is supported by the Adyar manuscript of Sucarita Mis´raʼs K s´ik . The previous editions unanimously assume a double accusative that does not make good sense. Especially the presence of ca is inappropriate. It is most likely that all the following editions simply inherited the wrong reading of A1.（7）
S´ barabh s4ya ad 1.1.18: bahus´a *upalabdhap rv d ev rtho vagamyata iti
*upalabdhap rv d ev rtho vagamyata] K1K2T1; upalabdhatv d arth vagama A1A2A3A(=A4)A6A7A8A9A11A12; upalabdhap rvatv d ev rth vagamah4 M(=A5);
upalabdhap rv d arth vagama I1S1S2V1; upalam4bhap rv d arth vagama U1ac; 7 Biardeau 1964:186 follows the Ānandās´rama edition A (=A4) and quotes it in a
footnote as given.
upalam4bhap rv d ev rth vagama U1ac
All manuscripts presuppose the existence of p rva that is absent in the previous editions except for M(=A5). The adopted reading upalabdhap rv d makes perfect sense, whereas upalabdhatv d found in most editions seems to be secondar y and in fact does not make good sense. The reading of M (upalabdhap rvatv d) that is a mixture of upalabdhap rv d and upalabdhatv d is probably a conjectural correction by the editor. It is likely that all the editions, including M in a different form, simply inherited a bad reading of A1（8）.
S´ barabh s4ya ad 1.1.18: na *c traikenocc ran4ayatnena
*c traikenocc ran4ayatnena] K1K2T1; caikenocc ran4 yatnena A1A2A3A(=A4);
caikenocc ran4ayatnena M(=A5)A6A7A8A9A11A12I1S1S2U1V1
The adopted reading c traikenocc ran4ayatnena is supported by the testimony of the M m m4s bh s4yaparis´is4t4a and a manuscript of Sucaritamis´raʼs K s´ik . The long at the end of ucc ran44 found in A1, A2, A3 and A(=A4) is clearly wrong. It is clear that A2, A3 and A(=A4) ignorantly followed A1. A8 and A9 in footnotes corrected the wrong reading of the previous editions, saying: ʻna caikocc ran4 yatnenaʼ iti mudritapustakes4v apap t4hah4. In fact the editor of A8 and A9, Yudhis4t4hir Mīmām4sak, should have written caikeno- instead of caiko-.
S´ barabh s4ya ad 1.1.20: na katham4cid apy abh vah4
katham4-] M(=A5)K1K2S1T1U1; kva- A1A2A3A(=A4)A6A7A8A9A11A12; kas´- I1S1; V1
omits the whole line
All editions except M(=A5) read kva- in the place of katham4- which makes better sense. Most probably the later editions blindly followed A1 either directly 8 Biardeau 1964:186 follows the Ānandās´rama edition A (=A4) and quotes it in a
footnote as given.
or indirectly.
These examples suffice to demonstrate what S´abara calls j tyandha- param4par (succession of the blinds by birth), or the innocent acceptance without question of an idea current among previous generations that is also baseless.（9）
The importance of the manuscripts from Kerala and testimony of the M m m4s bh s4yaparis´is4t4a
The eminence of the two manuscripts from Kerala is clear in contrast to the north Indian manuscripts. The Tanjore manuscript T1, though written in Devanāgarī, is also related more closely to the manuscripts from Kerala.
S´ālikanāthaʼs glosses in the M m m4s bh s4yaparis´is4t4a often support the readings of the south Indian group.
The critical edition of the M m m4s bh s4yaparis´is4t4a is of enormous help in deciding the original readings, for the editor gives not only the edited text but also manuscript readings. Namely, he gives a critical edition in the upper layer and a diplomatic transcript in the lower layer. It is not uncommon in Indian editions, e.g. S´lokav rttika commentaries such as the Ny yaratn kara, that editors change (or correct) manuscript readings without notification. And when they find prat kas or quotations from m la texts, e.g. S´lokav rttika, they often consult published editions of those m la texts and change their readings in opposition to the manuscripts. As a result, subsequent editors of those m la texts cannot be sure about the reliability of those later testimonies found, e.g. in an 9 The maxim j tyandhaparam4par ny ya is representatively explained in S´ barabh s4ya ad 1.3.1. A blind by birth insists that he knows a particular color. When one asks him about the source, he points out another blind. Another blind, too, points out yet another blind as a source of information. In this way they have no ultimate source for the information of a particular color. In the S´ barabh s4ya thereon the opponent uses this maxim in order to insist that Smr4tis which teach Ast4 4ak , etc. are based on erroneous cognition.
edition of the Ny yaratn kara. One may be in danger of “confirming” wrong reading of published m la editions by indirectly consulting the same editions.
This kind of danger is avoided in the case of the M m m4s bh s4yaparis´is4t4a thanks to the conscientiousness of the editor.
Improvement of the text of the Jaiminis tra
The present investigation of original manuscripts reveals various facts. Even the readings of some sūtras have been improved.
Jaiminis tra 1.1.11: vr4ddhis´ ca *kartr4bh mn sy t
kartr4bh mn syāt] I1pcK1K2T1; kartr4bh mn sya A(=A4)M(=A5)I1acS2S1U1V1
The previous editions as well as some manuscripts read asya instead of sy t.
Furthermore, the Subodhin (Jaiminis travr4ttih4) written by Rāmes´varasūri in 1839 AD (S´aka 1761) assumes the reading asya, because this commentary glosses it as asya s´abdasya vr4ddhih4. Therefore, this is neither a hypothetical reading nor a mere mistake newly introduced by A1. Rather it is a historically inherited reading that was found in the manuscripts used to prepare A1, too. Kevalānandasarasvatīʼs edition of Jaiminim m m4s s trap t4hah4 does not report the existence of any variants.
The three south Indian manuscripts unanimously read sy t. The reading sy t is supported by the M m m4s bh s4yaparis´is4t4a. This is the strongest evidence of all. Ending with sy t is very common in the Jaiminis tra and also makes good sense. The other reading asya which refers to s´abdasya by force of the context is a bit strange from the perspective of word order.（10） This must be a secondary reading derived from the original reading sy t.
10 There is only one case elsewhere in the entire Jaiminis tra which has asya in the end. Jaiminis tra 9.3.10: any yas tv avik ren4 drst4 4 4apratigh titv d avis´es4 c ca ten sya.
On the other hand, there are plenty of cases in which the Jaiminis tra ends with sy t.
Jaiminis tra 1.1.16: *s´abd ntaram4 vik rah4
*s´abd ntaram4 vik rah4] K2; varn4 ntaram avik rah4 A(=A4)M(=A5);
s´abd ntarah4 vik rah4 K1; varn4 ntaram4 vik rah4 I1S1S2U1acV1; s´abd ntaram avik rah4 T1U1pc
There are variant readings found in the editions and manuscripts for Jaiminis tra 1.1.16. Besides the minor problem of the switch between s´abda and varn4a, the most significant difference concerns the presence or absence of the negative a before vik ra. Rāmes´varasūriʼs Subodhin and Kevalānandasarasvatīʼs Jaiminim m m4s s trap t4hah4 both presuppose avik rah4 and did not notice at all the existence of this important variant vik rah4.
Although the situation of the manuscripts is too complicated for a definite judgment, here again the testimony of quotations helps us establish the original reading. Both the diplomatic edition of the M m m4s bh s4yaparis´is4t4a and a manuscript of the K s´ik read s´abd ntaram4 vik rah4. This must be the original.
The variant varn4a is probably a secondary reading introduced later in accordance with its referent, i.e. the sound i that changes into y in sandhi.
Presupposing avik rah4, G. Jhā translated the sūtra as follows (Jhā 1973:36):
“It is a different letter, not a modification.” But the original reading vik rah4
suggests the following interpretation: “The change [from i to y] is [not a change of the same sound, but is] a different sound.”
Jaiminis tra 1.1.17: *n davr4ddhih4 par
*n davr4ddhih4 par ] M(=A5)K1K2S1S2; n davr4ddhipar A(=A4)U1; n davr4ddhih4
par t I1; n radavr4ddhipar T1; n davr4ddhih4 V1
Both the M m m4s bh s4yaparis´is4t4a and the K s´ik read n davr4ddhih4 par . The diplomatic edition of the Br4hat also reads the same. Accordingly the Madras edition M, which contains the S´ barabh s4ya and the Br4hat, reads the same n davr4ddhih4 par . The editor of the Madras edition must have adopted this reading on the basis of the testimony of the Br4hat and the M m m4s -
bh s4yaparis´is4t4a. Kevalānandasarasvatī corrects n davr4ddhipar to n davr4ddhih4
par and notes his correction in a footnote.
One should take into consideration the usage of para in the sūtra, e.g. in 1.1.13 (satah4 param ...) and 1.1.14 (prayogasya param), which S´abara glosses (according to my edition) as yat param4 k ran4am uktam.
G. Jhā translates the sūtra as follows (Jhā 1973:36): “The ʻaugmentationʼ spoken of is the augmentation of the noise (not of the word).” He most probably presupposes n davr4ddhipar . Taking the separate par into consideration, one can translate it rather as follows: “The next one [that you have questioned] is the augmentation of n da [and not of s´abda].”
These examples suffice to demonstrate the importance of checking manuscripts and the testimony of quotations for establishing the original readings of the Jaiminis tra.
Improvement of the text of the S´ barabh s4ya
There are a number of improvements of the text of the S´ barabh s4ya in the present edition in comparison to the previous ones. Since the necessar y information is given in the critical apparatus, I do not have to mention all of them in this introduction. In the following I highlight some examples that reveal the unreliability of the previous editions.
S´ barabh s4ya ad 1.1.6: vinas4t4ah4 s´abdah4, punar *anyasya kriyam n4asy rthen - kr4takah4 sam4bandho nopapadyate
*anyasya] I1K1K2S1T1; asya A(=A4)M(=A5)S2U1
Presupposing asya, G. Jhā translates the line as follows (Jhā 1973:32):
Word is often found to be destroyed, so that when it comes to be produced (uttered) again, its relation (to its meaning) cannot but be artificial (newly made).
In fact asya, referring to “the same word”, does not fit the context, in which the opponents emphasize the transience of non-eternal words that are newly produced for each utterance. The adopted reading anyasya solves the problem.
And this reading is found as such in the M m m4s bh s4yaparis´ist4a. The whole line can be translated as follows: “A word [once uttered] has perished. It is impossible for another [word] being [newly] produced to have an unproduced [eternal]
connection with [its] meaning.” It is not “the same word” (asya) but rather
“another word” (anyasya) that is newly produced (kriyam n4asya) after the
disappearance of a previous word.
S´ barabh s4ya ad 1.1.8: tena n nam avagacchanti “*asann ev yam4 s´abdah4” iti.
*asann ev yam4] K1K2; sa ev yam4 A(=A4)M(=A5)I1S1S2T1U1V1
According to the opponents, everyday usage such as “Make a sound! (s´abdam4 kuru)” demonstrates that a sound indeed is produced and therefore not eternal.
Those people who use this expression assume that a sound is not yet in existence (asann eva) and therefore has to be produced anew.
The other reading sa ev yam4 that has been unanimously adopted in the previous editions does not make good sense. Therefore they read na te instead of tena. But the entire sentence na te n nam avagacchanti “sa ev yam4 s´abdah4” iti is not straightforward. Logically, it is not a direct continuation of what precedes.
The translation of G. Jhā reveals the difficulty of the connection (Jhā 1973:32):
Further, in practice people make use of such expressions as ʻmake (kuru, utter) the wordʼ, ʻdo not make the wordʼ: and yet they do not feel that they are referring to the same word (that has been in existence). (My emphasis)
The adopted reading asann ev yam4 s´abdah4 solves the problem. The entire paragraph of S´ barabh s4ya ad 1.1.8 can be translated as follows.
People use expressions such as “Make a sound!”, “Do not make a sound!”
and “This pupil makes a sound.” Therefore they indeed think that this sound is not yet in existence.
The connection of the latter sentence with the former is smooth and logically straightforward. Furthermore the sequence tena n nam is supported by other instances of its usage in the S´ barabh s4ya, e.g. tena n nam4 kart s´abd rtha iti gamyate (S´ barabh s4ya ad 3.4.13, 930.2) and tena n nam abhigh ran4am4
pray jas´es4en4 st ti (S´ barabh s4ya ad 4.1.37, 1217.23), whereas there is none that supports the strange word order na te n nam.
S´ barabh s4ya ad 1.1.13: v yav y h4 sam4yogavibh g v yv s´ritatv d v yus4v eva karis4yanti. yath *t ntav s tantus4v eva pat4am.
t ntav s] I1K2T1U1V1; tantavas AMS2; tav n K1; t m4tav sam4yogos S1
The previous editions read tantavas instead of t ntav s. Accordingly, G. Jhā translates the entire line as follows (Jhā 1973:34):.
[T]he Conjunctions and Disjunctions, being aerial and subsisting in the air, would produce the word in the air itself; just as the yarns produce the cloth in the yarns themselves. (My emphasis.)
But, as S1 or its ancestor probably intended, the agents of production of a cloth (pat4a) are the same sam4yogavibh g h4 that are mentioned in a line before, and not the threads (tantavah4). Otherwise the parallel in this example of a cloth would be lost. Threads are the samav yik ran4a of a cloth, and their conjunctions and disjunctions (sam4yogavibh g h4) are the asamav yik ran4a. Similarly, the opponents think that the conjunctions and disjunctions of wind atoms produce a sound only in the same wind atoms.
sam4yogavibh ga → ´abda sam4yogavibh ga → pat4a
v yu tantu
Therefore, the agent should be “conjunctions and disjunctions belonging to threads” (t ntav h4 sam4yogavibh g h4) and not the threads themselves (tantavah4).
On Jaiminis tra 1.1.13 (satah4 adars´anam4 vis4ay n gam t), S´abara commences his commentary as follows.
S´ barabh s4ya ad 1.1.13: *yat param4 k ran4am uktam
*yat param4] K1K2; yad aparam4 AMI1S1S2T1U1V1
Again on Jaiminis tra 1.1.14 (prayogasya param), S´abara begins his commentary with the same phrase.
S´ barabh s4ya ad 1.1.14: *yat param4 k ran4am uktam
*yat param4] K1K2T1; yad aparam4 AMI1S1S2U1V1
The previous editions read aparam4 instead of param4. The meaning of para and apara is the same, i.e. “another”. Both are direct glosses of para in the sūtras.
However, the M m m4s bh s4yaparis´is4t4a (ad 1.1.13) supports the reading yat param4. The variant reading aparam4 must be a secondary reading that arose later in accordance with its required meaning “another”. The old Jaiminis tra usage of para in the sense of “another” must have been forgotten, and instead apara, which is more common in this meaning, must have been newly introduced in the S´ barabh s4ya. This process is probably connected with the wrong transmission of Jaiminis tra 1.1.17 as n davr4ddhipar , for which n davr4ddhih4 par must have been the original.
S´ barabh s4ya ad 1.1.17: bahubhir bher m * dhmadbhir gos´abdam4
cocc rayadbhir mah chabda upalabhyate
* dhmadbhir gos´abdam4 cocc ra-] K1K2T1; dhamadbhih4 s´abdam ucc ra- A(=A4)I1; dhmadbhir gos´abdam ucc ra- M(=A5); dmabhih4 s´abdam ucc ra- S1; dmadbhir n4as´abdam4yoc ra- S2; dhmadbhih4 s´abdam4 samucc ra- U1;
dhmadbhih4 s´abdam ucc ra- V1
The older Ānandās´rama edition A (=A4) reads the first half as bahubhir bher m dhamadbhih4 s´abdam ucc rayadbhir. Apart from the difference of the form dhamadbhih4, it lacks the elements go and ca that are necessar y in interpreting the whole line correctly and smoothly. There are clearly two distinct examples that require the connective ca. Accordingly, G. Jhā translated it as follows, although he must have had before him a reading without ca (Jhā 1973:36):
[W]hen several people are beating a drum or pronouncing a word, the sound heard is very loud.（11） (My emphasis.)
Furthermore, the two examples must be concrete and not something very general. The word gos´abda is better than s´abda in general in juxtaposition with the concrete example of the sound of a drum. Thus it is better to adopt the reading with go and ca transmitted unanimously in the south Indian manuscripts.
Furthermore, it is not uncommon in the S´ barabh s4ya to give gos´abda as an example (e.g. S´ barabh s4ya ad 1.1.18, 1.1.19 and 1.1.20).
S´ barabh s4ya ad 1.1.20: na hi *pram n4ajyes4t4he pratyaks4e saty apram n4at sy t
*pram n4ajyes4t4he] M(=A5)K1K2U1pc; pram n4e A(=A4)I1U1ac; pram n4e jyes4t4e S1; pram n4a- S2; pram n4ajyes4t4a- T1
11 Biardeau 1964:185 translates the line in a similar way.
The older Ānandās´rama edition A (as well all the following editions except for M) reads pram n4e instead of pram n4ajyes4t4he that is supported by the M m m4s bh s4yaparis´is4t4a. Accordingly, G. Jhā, who bases his translation on the Bibliotheca Indica edition A1, translates the line as follows (Jhā 1973:40):（12）
So that when there is Sense-perception available as the means of cognising a certain thing, that thing could not be regarded as not cognisable by any means of cognition.
The expression pram n4ajyes4t4ha is appealing as a characterization of the priority of perception among pram n4as.（13） The same expression has been used by Umbeka (T tparyat4k 58.15), Kamalas´īla (Tattvasan4grahapa jik ad v.456, BB 200.16; ad v. 2140, BB 727.21; ad v. 2225, BB 751.13) and Jayanta (Ny yama jar I 282.9). S´āntaraks4ita records the expression jyes4t4hapram n4ena in Tattvasan4graha v. 456, which might have been quoted from Kumārilaʼs lost work Br4hatt44k . This expression jyes4t4hapram n4a has been also adopted by Sucarita Mis´ra (K s´ik ad anum na v. 58, III 43.22). We might not have guessed the hoary antiquity of this attractive characterization of perception but for this reading in the entire S´ barabh s4ya.
Synopsis of the s´abdanityatv dhikaran4a
1 p rvapaks4ah4: s´abdo nityah4 sūtra 1.1 karmaike tatra dars´an t 1.1.6 1.2 asth n t 1.1.7
12 Biardeau 1964:191 follows the Ānandās´rama edition A (=A4) and quotes it in a footnote as given.
13 Perception is considered by S´abara to be prior to other means of valid cognition.
S´ barabh s4ya ad 1.1.4, Frauwallner 1968:22.19: pratyaks4ap rvakatv c c num nopam n rth patt n m.
1.3 karotis´abd t 1.1.8 1.4 sattv ntare ca yaugapady t 1.1.9 1.5 prakr4tivikr4tyos´ ca 1.1.10 1.6 vr4ddhis´ ca kartr4bh mn sy t 1.1.11 2 ks4epaparih rah4: naiva s´abdo nityah4
2.1 samam4 tu tatra dars´anam 1.1.12 2.2 satah4 param adars´anam4 vis4ay n gam t 1.1.13 2.3 prayogasya param 1.1.14 2.4 dityavad yaugapadyam 1.1.15 2.5 s´abd ntaram4 vik rah4 1.1.16 2.6 n davr4ddhih4 par 1.1.17 3 siddh ntah4: s´abdo nityah4
3.1 nityas tu sy d dars´anasya par rthatv t 1.1.18 3.2 sarvatra yaugapady t 1.1.19 3.3 sam4khy bh v t 1.1.20 3.4 anapeks4atv t 1.1.21 3.5 prakhy bh v c ca yogasya 1.1.22 3.6 lin4gadars´an c ca 1.1.23 Conventions used in the edition
There are four layers of critical apparatus on each page except for the pages on which the second or the third is not necessary.
The first layer gives the edited text of the S´ barabh s4ya including the Jaiminis tra on which S´abara comments. The punctuation, with dan4d4as and commas, is mine.
The second layer is the critical apparatus for variants of the Jaiminis tra.
The third layer shows testimonia found, e.g., in the M m m4s bh s4yaparis´is4t4a.
Quoted or parallel passages of S´abaraʼs predecessors, such as the Mah bh s4ya, are included here.
The fourth layer is the critical apparatus for variants of the S´ barabh s4ya.
The critical apparatus is positive: each reading is repeated after the line number in the edition and followed by a lemma sign ʻ]ʼ. This is followed by the siglum or sigla of edition(s) and/or manuscripts which have the accepted reading. After a semi-colon, variants are given with the siglum of the edition/manuscript. For example:
vinas4t4ah4 s´abdah4] AMK1K2S2T1U1; vinas4t4e s´abde I1S1
This means that editions A and M and manuscripts K1, K2, S2, T1 and U1 have the accepted reading vinas4t4ah4 s´abdah4, while two manuscripts I1 and S1 have the variant vinas4t4e s´abde.
I have not reported the punctuation and nonsignificant variants of the editions and manuscripts, e.g. the non-application of sandhi or common orthographic variants such as gemination or degemination of consonants after or before semi-vowels, and the exchange of anusv rah4 and homorganic nasals.
However I reproduce them when a reading of an edition or a manuscript is quoted as a significant variant. With this convention I follow Isaacson [2002:155].
Acknowledgment
I am indebted to the following libraries and institutes for having given me permission to consult manuscripts: The Bharat Kala Bhavan Library of Banaras Hindu University, Varanasi; The Oriental Research Institute & Manuscripts Librar y, Kerala University, Trivandrum; The Sarasvati Bhavan Librar y of Sampurnananda Sanskrit University, Varanasi; The British Library, London; The Library of the University of Pennsylvania, Philadelphia; The Tanjore Maharaja Serfojiʼs Sarasvati Mahal Library, Thanjavur. The present research was supported by the Grant-in-Aid for Young Scientists (B), from The Ministry of Education, Culture, Sports, Science and Technology, Japan. Prof. Dr. Arlo Griffiths was very kind to go through the final draft and gave me many valuable suggestions. Last but not least I am much obliged to Prof. Dr. Harunaga Isaacson for his comments on my final draft.
Abbreviations
ac Readings before correction (ante correctionem).
pc Readings after correction (post correctionem).
(eyeskip) The variant is caused by eyeskip.
(marked) The variant is marked with a dot above the letter without correction.
corr. Correction by the editor.
ed. Reading as given in the quoted edition.
em. Emendation by the editor.
om. Omission of the reading.
Bibliography
R4juvimal : see Br4hat.
R4gvedapr tis´ khya: S´aunakaviracitam R4gvedapr tis´ khyam. Ed.
Vīrendrakumār Varmā. Delhi: Chaukhamba Sanskrit Pratishtan, 1986.
K s´ik (by Sucarita Mis´ra): A commentary by Sucarita Mis´ra on Kumārilaʼs M m m4s s´lokav rttika. A transcript preserved in the Adyar Library, Chennai, No.
63361 (38.G.5, TR 66.7). Paper. Devanāgarī. pp. 3200–3689. [According to the colophon given at the end, this is a transcript of a Grantha manuscript preserved in the Government Oriental Manuscript Library, Chennai, No. 3689 (25820).
Sucaritaʼs Kās´ikā has been published only up to the sam4bandh ks4epa. The remaining sections, i.e. sphot4a, kr4ti, apoha, vana, sam4bandh ks4epaparih ra, citr ks4epaparih ra, tman, s´abdanityatva, v kya and ved paurus4eyatva, remain unpublished.]
Jaiminim m m4s s trap t4ha: M m m4s dars´anam Jaiminim m m4s - S trap t4hah4. Ed. Kevalānanda Sarasvatī. Wai: Prājña Pāt4has´āl4ā Man4d4al4a, 1948.
Tattvasan4grahapa jik
BB: Tattvasangraha of c rya Sh ntaraks4ita with the Commentary Pa jik of Shri Kamalash la. Ed. Dvārikadāsa S´āstrī. 2 vols. Bauddha Bharati Series, 1 &
2. Varanasi: Bauddha Bharati, 21981, 21982.
GOS: Tattvasan4graha of S´ ntaraks4ita. Ed. Embar Krishnamacharya. 2 vols.
Baroda: Oriental Institute, 21984, 21988.
T tparyat4k : S´lokav rtikavy khy T tparyat4k of Um4veka Bhatt44a. Ed. S.K.
Rāmanātha S´āstrī. Madras: University of Madras, 1971.
Ny yabh s4ya: Gautam yany yadars´ana with Bh s4ya of V tsy yana. Ed.
Anantalal Thakur. New Delhi: Indian Council of Philosophical Research, 1997.
Ny yama jar of Jayantabhatt44a with T4ippan4 Ny yasaurabha by the Editor.
Ed. K.S. Varadācārya. 2 vols. Mysore: Oriental Research Institute, 1969, 1983.
Ny yama jar granthibhan4ga: Ny yama jar granthibhan4ga. Ed. Nagin J.
Shah. Ahmedabad: L.D. Institute of Indology, 1972.
Ny yaratn kara: S´lokav rttika of S´r Kum rila Bhatt44a with the Commentary Ny yaratn kara of S´r P rthas rathi Mis´ra. Ed. Swāmī Dvārikadāsa S´āstrī.
Varanasi: Tara Publications, 1978.
Ny yas tra: See Ny yabh s4ya.
Br4hat : Br4hat of Prabh kara Mis´ra with the R4juvimal pa cik of S´ likan tha.
Part I. Ed. S.K. Rāmanātha S´āstrī. Madras: University of Madras, 1934.
Mah bh s4ya: The Vy karan4a-Mah bh s4ya of Pata jali. Ed. F. Kielhorn. 3 vols.
Poona: Bhandarkar Oriental Research Institute, 41985, 31965, 31985.
Mah bh s4yad pik : Mah bh s4yad pik of Bhartr4hari. Fascicule I, Ed. G.B.
Palsule (1985); II, V.P. Limaye, G.B. Palsule & V.B. Bhagavat (1985); III, V.B.
Bhagavat & Saroja Bhate (1986); IV, Johannes Bronkhorst (1987); V, G.B. Palsule (1988); VI, G.V. Devasthali & G.B. Palsule (1989); VII, V.B. Bhagavat & Saroja Bhate (1990); VIII, G.B. Palsule & V.B. Bhagavat (1991). Poona: Bhandarkar Oriental Research Institute.
M m m4s bh s4yaparis´is4t4a: Br4hat of Prabh kara Mis´ra. Part II. Ed. S.K.
Rāmanātha S´āstrī. Madras: University of Madras, 1936. [This contains a critical edition of the M m m4s bh s4yaparis´is4t4a in the upper layer and a diplomatic transcript of a manuscript of the same in the lower layer.]
V kyapad ya: Bhartr4haris V kyapad ya. Ed. Wilhelm Rau. Wiesbaden: Franz Steiner, 1977.
V kyapad yasvavr4tti: V kyapad ya of Bhartr4hari with the Commentaries Vr4tti
and Paddhati of Vr4s4abhadeva. Kān4d4a I. Ed. K.A. Subramania Iyer. Poona: Deccan College, 1966.
Vais´es4ikas tra: Vais´es4ikas tra of Kan4 da with the Commentar y of Candr nanda. Ed. Muni Jambuvijayaji. Baroda: Oriental Institute, 21982.
Vais´es4ikas travr4tti: See Vais´es4ikas tra.
S´ barabh s4ya (1.1.1–5): See Frauwallner 1968.
S´ barabh s4ya: S´r majjaiminipran4 tam4 M m m4s dars´anam. Ed.
Subbās´āstrī. 6 bhāgas. Poona: Ānandās´ramamudran4ālaya, 1929–34.
S´lokav rttika: See Ny yaratn kara.
Biardeau, Madeleine
1964 Th orie de la connaissance et philosophie de la parole dans le brahmanisme classique. Paris: Mouton & Co.
DʼSa, Francis X.
1980 S´abdapr m n4yam in S´abara and Kum rila. Vienna: De Nobili Research Library.
Frauwallner, Erich
1961 “Mīmāmsāsūtram I, 1, 6-23.” Wiener Zeitschrift f r die Kunde S d- und Ostasiens, 5, 113-124.
1968 Materialien zur ltesten Erkenntnislehre der Karmam m m4s . Wien:
Hermann Bölaus Nachf.
Harunaga Isaacson
2002 “Ratnākaras´āntiʼs Bhramaharan ma Hevajras dhana: Critical Edition (Studies in Ratnākaras´āntiʼs tantric works III).” Journal of the International College for Advanced Buddhist Studies, 5, 2002, 151–176.
Houben, Jan E.M.
1995 The Sambandha-Samuddes´a and Bhartr4hari s Philosophy of Language.
Groningen: Egbert Forsten.
Jhā, Gan4gānātha
1973 S´ bara-Bh s4ya, Translated into English by Ganganatha Jha. Vol. 1. Baroda:
Oriental Institute.
Kataoka, Kei
2004 Koten Indo no Saishiki K i-ron: S´ barabh s4ya & Tantrav rttika ad 2.1.1–4 Genten K tei Yakuch Kenky . Tokyo: Sankibo Press.
, ,
5
1.1.6 ] AMI1K1S1S2T1U1V1; K2
3 ] Jaiminis¯utra 1.1.5: ; S¯´abarabh¯as.ya
ad 1.1.5, Frauwallner 1968:24.3:
4 ] M¯ım¯am. s¯abh¯as.yapari´sis.t.a45.17:
4 · · · ] M¯ım¯am. s¯abh¯as.yapari´sis.t.a45.17–18: ---
---
5 · · · ] M¯ım¯am. s¯abh¯as.yapari´sis.t.a45.18: ---
6 ] M¯ım¯am. s¯abh¯as.yapari´sis.t.a 45.22: ; see also
´Sabara’s reply in 1.1.12: “ ”
3 ] AMI1K1K2S1S2T1U1; V1(marked) 3 · · · ] AMI1K1K2S1S2T1U1; om. V1(eyeskip) 3 ] AMK1
K2S1T1U1; I1(marked) ; S2 4 ]
AMK1K2S2T1U1; I1S1 4 ] I1K1K2S1T1; AMS2U1 4 ] AMI1K1K2S1U1; S2; T1 5 ] AMI1K1K2S1S2T1U1; V1 5
] K1K2; AMI1S2U1V1; S1;
T1 6 ] AMI1K1K2S1S2T1U1; V1 6 ] AMK1K2S1S2T1U1V1; I1 6 ] K1T1; AMI1S1
S2U1V1; K2 6 ] K1K2; AMI1S1S2T1U1V1
5
1.1.7 ] AMI1K1K2T1S2U1V1; S1
1 ] Cf.Ny¯ayabh¯as.yaad 2.2.18, 109.7:
2 · · · ] Cf. Ny¯ayas¯utra 2.2.18, 109.2:
; Ny¯ayabh¯as.ya ad 2.2.18, 109.3:
; Vai´ses.ikas¯utra 2.2.30, 22.2:
; Vai´ses.ikas¯utravr.tti ad 2.2.30, 22.3–4:
; Vai´ses.ikas¯utra 2.2.33, 22.12: ; Vai´ses.ikas¯utravr.tti ad 2.2.33, 22.13:
5 · · · ] Cf. Vai´ses.ikas¯utra2.2.29, 21.19:
; Vai´ses.ikas¯utravr.tti ad 2.2.29, 21.20–21:
,
5 ] M¯ım¯am. s¯abh¯as.yapari´sis.t.a46.14:
1 ] AMI1K1K2S1S2V1; T1; Uac1 ; Upc1 1 ] AMI1K1K2S2U1; ¯ S1; T1; V1 2 ] AMI1K1K2S1S2T1U1;
V1 2 ] AMI1K1K2S1S2T1U1; V1 3 ] AMI1K1
K2S1S2T1U1; V1 3 ] AMK1K2S1S2U1; I1; T1; V1 3 ] AMK1K2S2T1U1;
I1S1V1 5 ] K1K2;
AMS1S2U1; I1;
T1;
,
, ,
“ ” “ ” “ ”
5“ ”
1 ] Cf. Vai´ses.ikas¯utravr.tti ad 2.2.30, 22.1:
,
1 · · · ] Cf.Ny¯ayas¯utra2.2.18, 109.2:
; Ny¯ayabh¯as.yaad 2.2.18, 109.3–4:
( )
5 · · · ] Mah¯abh¯as.ya I 1.12 ; also quoted inBr.hat¯ı266.8–9
andR. juvimal¯a266.15
1 ] AMI1K1K2S1T1U1V1; S2 1
] AMK1K2S2T1U1; I1S1; V1 2
] AMI1K1K2S1S2U1V1; T1 2 ]
AMK1K2I1S1S2U1; T1; V1 2
· · · ] AMK1K2S1S2T1U1; · · · I1; V1
3 ] AMK1K2S1S2T1U1; I1 3 ] AMK1K2T1Uac1 ; I1;
S1; S2;
Upc1 5 “ ] K1K2T1;
AMI1S1S2U1 5 ] AMK1K2S1S2T1U1; I1 5 ”
“ ” ] I1K1K2T1Upc1 ; AMS1Uac1 V1; S2 6 ] AMI1K1K2S1S2U1V1; T1 6 ] K1K2T1; AMS1S2Uac1 V1; I1; Upc1 6 ] AMI1K1K2
5
“ ”
1.1.9 ] AMI1K1K2T1U1; S1;
S2; V1
2 · · · ] Ny¯ayaratn¯akara518.16–17:
,
6 · · · ] Cf. Ny¯ayabh¯as.ya ad 2.2.40, 119.5:
6 “ ” ] M¯ım¯am. s¯abh¯as.yapari´sis.t.a48.14:
7 ] M¯ım¯am. s¯abh¯as.yapari´sis.t.a48.17:
2 ] K1K2T1; AMI1S1S2U1V1 2 ] AMI1K1K2S2T1U1V1; S1 2 ] AMI1K1K2S1S2T1U1; V1 2 ] K1
K2T1; AMI1S1S2U1V1 3 ] AMI1K1K2S2T1U1V1; S1 3 ] AMI1K1K2S2T1U1V1; S1 3 ] AMI1K1K2S2
T1U1V1; S1 3 ] AMK1K2S2U1V1; I1; S1; T1 3 ] AMI1K1K2S2T1U1; S1; V1 4 ] AMI1K1S1S2T1U1V1; K2 4 ] AMI1K1K2S1
T1U1V1; S2 4 ] AMI1K1K2S2T1U1V1; S1 4 ] K1K2T1; AMI1S1S2U1V1 6 ” ] AMI1K1K2S1T1Upc1 ; S2Uac1 (eyeskip) ; V1
6 ] AMI1K1K2S1S2U1V1; T1 7 ] AI1K1K2
S1S2T1U1V1; M 7 ] AMS2U1; I1K1K2S1T1; V1 7 ] AMI1K1S1S2T1U1; K2; V1
7 ] K1K2T1; AMI1S2U1V1;
,
“
5”
1.1.11 ] Ipc1 K1K2T1; AMIac1 S1S2U1V1
2 ] M¯ım¯am. s¯abh¯as.yapari´sis.t.a48.17–18:
3 · · · ] Cf.Vai´ses.ikas¯utra2.2.34, 22.15: ;
Vai´ses.ikas¯utravr.tti ad 2.2.34, 22.16–17:
, ; V¯akyapad¯ıyasvavr.ttiad
I v. 102 (Iyer v. 99), I 164.12–165.1:
4 ] M¯ım¯am. s¯abh¯as.yapari´sis.t.a48.18–19:
1 ] K1K2T1; AMI1S1S2Uac1 V1; Upc1 3 ] AMK1K2S1T1U1V1;
I1; S2; 3 ] AMK1K2S1S2U1V1; I1;
T1 3 ] K1K2; AV1; M ;
I1S1; S2; T1; Uac1 ; Upc1 3 ] AMI1K1K2S1S2U1V1; T1 3 ] K1T1Upc1 ; AMI1K2S2Uac1 V1; S1 4
] AMI1K1K2S1; S2V1; T1;
U1 4 ] AMK1S1T1U1Vpc1 ; I1K2S2; Vac1 4 , ] K2S2T1Upc1 ; AMI1K1Uac1
V1; S1 5 ] AMI1K2S2U1; K1T1;
S1; V1 5 ] AMI1K1K2S1T1U1V1; S2 5 ] AMK1K2S1T1Upc1 ; I1; S2Uac1 V1 5 ] K1K2T1;
AMS1S2U1V1; I1 6 ” ] I1K1K2S1T1;
