MUSLIMS IN BUDDHIST PHILOSOPHICAL TEXT

It is already well-known … that written [Indian, vdK] sources from about the eighth century do not use terms which are used today as generic terms to refer to Muslims. … At least in the middle of the thirteenth century, there is clear evidence of familiarity with the term Musalamana (literally, ‘one who submits to Allah’), and of concepts which relate to the practice of Islam Thus writes B. Chattopadhyaya in his recent, thin but highly informative booklet.1 He also mentions there the till now earliest attested occurrence of the Sanskrit term musalamana , that is, ‘‘Muslims,’’ in the Indian subcontinent, which is found in the Veraval inscription of Arjunadeva, which is dated 1264. Of course, we know that Islam entered the subcontinent in a significant way in the second decade of the eighth century when Sindh was conquered by Muh: ammad bin Qasim. As D. Pingree, A. Wink and others have shown, Muslim traders had plied their trade in the subcontinent somewhat before this time, particularly in the western coastal towns of Gujarat, Konkan and Malabar that lie on the Arabian Sea.2 Therefore, there can hardly be any question that by the end of the seventh century Islam was not an entirely unknown quantity in certain parts of the subcontinent and that by this time its western seaboard had become home to several Muslim communities. It is difficult to be certain whether the Sanskrit neologism musalamana is derived from an Arabic, Persian, or some other source. Persian musulman is a reflex of Arabic muslimun , the plural of muslim. In any event, ‘‘the word musulmam or musalman is already well established in New Persian by the time of the earliest texts we have,’’ notably in the anonymous Hudud al-’Alam of 982, a geography of the [then] known world.3 Dating from not later than the first half of the eleventh century, the K alacakra corpus of esoteric Buddhist texts confronts Islam in several ways and even makes use of the Muslim ‘‘year of the migration’’ (hijra) – tradition has it that it began on 16 July, 622, when Muh: ammad and his followers broke up camp in Mecca to leave for Medina – as its fundamental calendrical constant, the so-called vahnau [3] khe [0] ’bdhau [4] (Tib. me mkha’ rgya mtsho), that is, 403 years after the hijra (A.H.); the number implicit in vahnau khe ’bdhau must be read from right to left. Of interest is that it nowhere employs the designations musalamana , etc. to refer to Muslims. Rather, the terms it uses are mleccha, tayin [and mleccha tayin ] and feminine tayini . 4 It now appears that an expression like but certainly not identical to musalamana had made its way into Sanskrit much earlier, in fact, several centuries earlier than has been supposed hithertofore. Indeed, Avalokitavrata uses the term in his commentary on Bh aviveka’s Prajn aprad ipa[= pp] that may be dated around 700.. The general consensus is that the famous Buddhist writer Bh aviveka [or: Bhavya, Bh avaviveka] flourished from circa 500 to 570. Among the treatises that are uncontroversially ascribed to him is the pp, a very substantial commentary on N ag arjuna’s (ca. 2nd c.) versified Mulamadhyamakak arik a [= MMK].5 Not one single Sanskrit manuscript of this work is extant, but we have two different translations: a Chinese version that was prepared by Prabh akaramitra in 630–2 and a circa 800–20 Tibetan version by the ‘‘Indian Mah ay ana M adhyamika’’ Jn anagarbha and Cog / Lcog ro Klu’i rgyal mtshan – hereafter, he will be referred to as ‘‘Cog ro.’’ Comparing the two versions, we notice that they presuppose at times quite different readings of the original Sanskrit manuscripts on which they were based. It is often alleged that the Chinese translation is generally of an inferior quality, but I am not altogether convinced of the cogency of privileging for this reason the Tibetan rendition and by and large ignoring the former, as is by no means infrequently done. Indeed, the Tibetan and Chinese passages of the text that I consulted for this paper measure up pretty well to any standard. The pp, in turn, was the subject of at least two studies, the first of which was written by a *Gun: adatta (Tib.Yon tan byin). Now lost, it is only known to have existed at some point by way of the fragmentary evidence found in the second study, namely, Avalokitavrata’s very large t: ika -commentary.6 A possible indication of the terminus ad quem of *Gun: adatta’s floruit may be found in the Chinese Buddhist canon. There we learn that the Indian Buddhist commentator Gongdeshi was responsible for an exegesis of the Vajracchedikas utra , which a consortium of literati including Dipoheluo (*Div akara) – this Central Indian monk lived from 613 to 688 and arrived in China in 676 – had rendered into Chinese.7 The Korean monk Wonch’uk [= Ch. Yuance] (613–96) cites this exegesis several times in his large commentary on the Sam: dhinirmocanasutra , where its early ninth century translator Wu Facheng alias ’Gos Lo ts a ba Chos grub rendered his name by ‘‘Yon tan byin.’’8 This is not to say that gongdeshi was not be interpreted differently. Citing a nineteenth century Tibetan author and pointing out that Chinese gongde is also used to translate Sanskrit sr i, G. Tucci opted for reconstructing the author’s name as *S´r datta rather than*Gun: adatta. It now turns out that the gongdeshi-dpal sbyin / byin (*sr idatta) reconstruction goes at least back to the one we find in Mgon po skyabs’ eighteenth century history of Buddhism in China, which is in part based on the Zhiyuan fabao kantong zonglu, a comparative Sino-Tibetan catalogue of translations of Buddhist texts, that was prepared at the Mongol-Yuan winter capital of Dadu in 1285 to 1287.9 This is all fine and well were it not for the fact that we have much earlier evidence that the Tibetan name of the author of this commentary was Yon tan byin. As is well known, the 812 / 824 Lhan dkar ma [or: Ldan dkar ma] catalogue that was included in the Tibetan Buddhist Tanjur-canon as well as the 830 ’Phang thang ma catalogue are invaluable witnesses for the earliest Tibetan versions of Buddhist scripture that were prepared during a period of some 60 years, though there is some evidence of later interpolations in the recently published manuscript of the ’Phang thang ma. Now, it so happens that a translation of a study of the Vajracchedikas utra in two bam po units of text by none other than a Yon tan byin is registered in the ’Phang thang ma, without underlining that the translation was done from the Chinese.10 Dar ma rgyal mtshan (1227–1305) alias Bcom ldan rig[s] pa’i ral gri appears to register such a commentary in his circa 1270 catalogue of translated scripture, albeit without giving the author’s name, whereas Bu ston Rin chen grub (1290–1364) [wrongly] lists it under ‘‘Yon tan ’od zer byin’’ in the catalogue appended to his chronicle of 1322–6.11 This translation did not survive and was for this reason not included in the extant editions of the Tibetan Tanjur. ‘‘Yon tan byin’’ would of course be the equivalent of ‘‘*Gun: adatta’’ and not ‘‘*S´r datta,’’ since, to my knowledge, yon tan is not attested as a possible translation of sr i. In other words, then, if the author of the pp and Vajracchedikas utra commentaries were one and the same and if gongdeshi reflects gun: adatta, which I believe to be more than merely likely, then *Gun: adatta must have flourished before the middle of the second half of the seventh century, at the latest.

Only available in the Tibetan translation from the pen of the same team to whom we owe the rendition of the pp, not a single Sanskrit manuscript of Avalokitavrata’s work appears to have survived, and it was never translated into Chinese. What I take to be the author’s colophon relates that he came from the Sa ke ta na area (yul sa ke ta na nas byung ba), that he was a brahmin, and that he was learned in all the domains of knowledge, as he writes, in the manner of a large S al[a] tree [= Shorea Robusta Gaertn.f.] (shing sa la chen po lta bu ). The S al[a] tree grows in many of the northern hilly areas of the subcontinent and this notice may very well indicate that Avalokitavrata hailed from northern India, possibly, from what is now Bihar State, though I must confess that I do not catch the drift of the simile. But I am unable to identify *Saketana. It is possible, as was suggested to me by M. Witzel, that it refers to S aketa, an earlier name for present day Ayodhy a. Another hint of his possible north Indian origins may be found in his deliberations on a passage in the first chapter of the pp where he juxtaposes the speech of the barbarians (kla klo’i skad, *mlecchasabda ) with the speech of those living in Magadha (dbus pa’i skad, *magadhasabda, magadh i).12 At the same time, we should also be cognisant of the very good possibility that, when he was writing, M agadh still enjoyed the status of a prestige vernacular. As with the pp, the translation of Avalokitavrata’s commentary is registered in the Lhan dkar ma as well as in the ’Phang thang ma. 13 Neither indicate the identity of the translator[s] of either treatise. The first catalogue explicitly to note Cog ro in the present context, albeit without his Indian counterpart, is the one Dbus pa Blo gsal Rtsod pa’i seng ge (ca. 1265–1355) prepared for a Tanjur manuscript of Snar thang monastery in circa 1310–20; the names of both translators appear in Bu ston’s 1335 catalogue of the Zhwa lu Tanjur manuscript.14 The translators’ colophon states, as does the one of the pp, that their translation was done at the order (bka’ lung) of the Tibetan emperor who is styled dbang phyug dam pa’i mnga’ bdag rgyal po dpal lha btsan po, that is, ‘‘Lord Noble Great Sovereign King, Lustrous Divine Mighty One’’15 – here literally rendered by ‘‘Mighty One,’’ btsan po can of course also be translated by ‘‘Emperor,’’ that is, as the attested semantic equivalent of Chinese tianzi, Son of Heaven.’’16 It is not entirely obvious who might be indicated by this imposing string of epithets; there are four possible candidates: Khri srong lde btsan (r.?-ca. 797), Mu ne btsan po (r. ?-803), Khri lde srong btsan (r. 803–815), or Khri gtsug lde btsan (r. ca. 815–41) alias Ral pa can. Cog ro’s name is absent from the list of individuals mentioned in the first section of the Sgra sbyor bam po gnyis pa lexicon. Judging from a comparison of a fragment of the latter found in Ta pho monastery with the text in the Tanjur [plus two fragments found near Dunhuang], the first section of this bipartite lexicon, which comprises a decree of sorts, initially dated from 795 [or, less likely, 783] and when this decree was reconfirmed in 814, it was accompanied by a deep revision and enlargement.17 The early epigraphical monuments, too, do not record his name. But the consensus as to his floruit in admittedly much later Tibetan literary sources seems to have been that he was active in the late eighth and the first half of the ninth centuries. For example, the twelfth century historian Nyang ral Nyi ma’i ’od zer (1124–92) situates him toward the end of Khri srong lde btsan’s reign; Dar ma rgyal mtshan does the same; and Sne’u / Snel Pan: d: ita Grags pa smon lam blo gros (13th–14th c.), who was familiar with Dar ma rgyal mtshan’s oeuvre, asserts that he was alive and well during the era of Khri gtsug lde btsan.18 The ostensibly late recensions of the Sba bzhed that is attributed to Sba Gsal snang (ca. 800) retrospectively note Cog ro as a member of a triumvirate of key-translators – the other two are Ska ba Dpal brtsegs and Sna nam Ye shes sde – in connection with an era prior to Khri gtsug lde btsan’s reign.19 We learn from the Lhan dkar ma and the ’Phang thang ma that Cog ro was not only a translator, but also an author in his own right. The first registers a work by him in its rubric ‘‘Commentary of Mah ay anasutra[s],’’ whereas, somewhat anomalously, the latter subsumes the same treatise under the one of ‘‘Sutra-commentary Translated from the Chinese’’:20 1. A substantial commentary (rgya cher ‘grel pa) of the Sam: dhinirmocanasutra in [forty bam po units of text] Absent from the Lhan dkar ma, the ’Phang thang ma contains at the end the rubric ‘‘Commentaries of sutras and exegetical works; what was written by Emperor Khri srong lde btsan.’’ This rubric ends with the statement: ‘‘These were written by Tibetan masters; smaller texts were not recorded,’’ and contains the following treatises from Cog ro’s pen:21 2. Nges pa’i don dbu ma; a treatise on Madhyamaka thought [bam po unit of text] 3. Dbu ma’i don, a summary-cum-petition (gsol ba btab pa) of/to the Prajn ap aramit a (dbu ma’i don dbu ma slob dpon klu’i rgyal mtshan gyis shes rab kyi pha rol du phyin pa la / mdor bsdus te gsol ba btab pa) [bam po unit of text] 4. Chos kyi gzhung spyir bstan pa, the brgal lan bzhi phrugs [co-authored with Ska ba Dpal brtsegs]; a four-fold reply to controversies surrounding Buddhism [bam po unit of text] 5. Phreng ba’i rgyan [one bam po unit of text] 6. Chos gtan la dbab pa’i mdo, an ?important [text] (gal) written by Master Klu’i rgyal mtshan [shu log (< sloka) unit of text]; on the essentials of Buddhism Dar ma rgyal mtshan had access to versions of the Lhan dkar ma and the ’Phang thang ma, but he recorded not six but eight works for Cog ro:22 1. Gal (sic!) lan bzhi phrugs [co-authored with Ska ba Dpal brtsegs] 2. A commentary (rgya cher ’grel pa) of the Sam: dhinirmocanasutra [forty bam po-s] 3. Nges don gyi dbu ma 4. Dbu ma’i don sher phyin las btus te gsal [sic!] btab pa 5. Theg chen kun rdzob dang don [?dam] bstan pa 6. Theg chen dbu ma’i tshul bstan pa [two hundred shu log units of text] 7. Chos kyi gzhung phyir [read: spyir] bstan pa phreng ba’i rgyan [one bam po unit of text] 8. Chos gtan la dbab pa’i mdo’i gal

Dar ma rgyal mtshan’s nos. 5 and 6 occur between the Dbu ma’i don and the Chos kyi gzhung spyir bstan pa in the ’Phang thang ma which, however, does not explicitly attribute them to Cog ro, and his no.7 collapses nos. 4 and 5 of the ’Phang thang ma into one single work. While Dbus pa Blo gsal’s catalogue does not list the titles of any specimen of Cog ro’s oeuvre, Bu ston has the following five in the catalogue appended to his chronicle:23 1. A commentary on the Sam: dhinirmocanasutra [forty bam po-s] 2. Nges pa’i don dbu ma [bam po unit of text] 3. A summary-cum-clarification (mdor bsdus te gsal ba) of the Prajn ap aramit a [literature] [bam po unit of text] 4. Phreng ba’i rgyan [one bam po unit of text] 5. Chos gtan la dbab pa’i gal To be sure, the reason why Dar ma rgyal mtshan and Bu ston, in his catalogue appended to his chronicle, register these among many other lost writings from the imperial period should probably be sought in the fact that their catalogues did not correspond to any actual collection of texts. This would explain, of course, why Dbus pa Blo gsal and Bu ston did not record them in their catalogues for the Snar thang and Zhwa lu Tanjurs, which indeed were textual corpora that were actually present in these monasteries. Whatever the case may have been, the above data suggest that circa 750–80 would not be an altogether unreasonable terminus ad quem for Avalokitavrata and his t: ika -commentary. A few final words. Contrary to the Lhan dkar ma, the published text of the ’Phang thang ma has an afterword that, for several reasons, should be of some importance to the cultural historian, despite the considerable likelihood that it was added much later to the catalogue itself. It is, for example, interesting to note that Cog ro is there said to be a ‘‘great translator who renders the sense [of the original texts]’’ (don sgyur gyi lo tsha ba chen po), whereas his colleague Ska ba Dpal brtsegs is designated as a ‘‘great translator who renders the wording [of the original texts]’’ (sgra sgyur gyi lo tsha ba chen po). The categories of don sgyur or sgra sgyur do not occur in the Sgra sbyor bam po gnyis pa. Further, the colophons of the translations of the pp and Avalokitavrata’s commentary prefix Cog ro’s name by the ‘‘Tibetan editor-translator’’ (bod kyi zhu chen gyi lo tsa ba ), and we meet with the same in his extraordinary corpus of translations from the vinaya literature, much of which he did together with Jinamitra, a Kashmirian Mulasarv astiv ada monk, whose work only the canonical recension of the Sgra sbyor bam po gnyis pa notes as having been registered in a catalogue. All of them include the very same reference to the Tibetan emperor. An exception is Cog ro’s and Jn anagarbha’s translation of Buddhap alita’s (ca.5th c.) study of the MMK, where this is lacking.24 J.K. Panglung has noted that the Ta pho fragment of the Sgra sbyor bam po gnyis pa does not include the reference to the translations having been edited (zhus) or to the expression zhu chen, 25 both of which we find in the corresponding section of its canonical recension. This would indicate that Cog ro’s editorial activities did not take place until the era of Khri lde srong btsan, at the earliest. Further, whereas the Ta pho fragment has yab myes, the corresponding canonical text has just [sngon] lha sras yab. 26 The latter expression clearly means ‘‘[the preceding] son of the divine, [my honorable] father.’’ The intent of the term yab m[y]es is more difficult to determine. It occurs several times in the epigraphical record as a dvandva compound in the sense of ‘‘father and grandfather,’’ and as a collective noun meaning more generally ‘‘ancestors’’, with the clear implication of a farther remove in time.27 Whether or not one opts for the latter, as does J.K. Panglung, who is also inclined to place this fragment in the era of Khri srong lde btsan, the consequential implication is that Tibetan translations of the Ratnameghasutra and the Lank_ avat aras utra , the two sutras mentioned in this passage, predate him by at least two generations.28 This would mean that we must seriously consider a reevaluation of the onset of the translation of Buddhist texts into Tibetan. Cog ro seems also to have been active on fronts other than Madhyamaka philosophy and Vinaya texts. For example, he and [?a] Jn anagarbha are mentioned as translators of several esoteric texts contained in the [Sde dge print of the] Rnying ma’i rgyud ’bum – these are not listed in the Lhan dkar ma and the ’Phang thang ma –, and several authors overtly affiliated with the Rnying ma, ‘‘the Old,’’ school of Tibetan Buddhism inconsistently note him in connection with inter alia a role he may have played in the transmission of the teachings of Vimalamitra, one of this school’s enduring patriarchs.29 Avalokitavrata’s terminus a quo is perhaps less difficult to ascertain. We are helped by the fact that he mentions Sthiramati (ca. 510–70) and Candrak rti (ca. 600–50) among those individuals who wrote commentaries on the MMK, though these are not listed in chronological order.30 Though contemporaries, we know that Sthiramati criticized Bh aviveka and that the latter was also taken to task by Candrak rti. Avalokitavrata quotes from Sthiramati’s work several times in a critical fashion, but not once from Candrak rti. This is strange, for Candrak rti had also voiced his fundamental disagreement with Bh aviveka on several occasions in his study of the MMK. In the absence of any evidence to the contrary, the terminus a quo of his floruit could have fallen around the middle of the seventh century. In other words, then, there is no question that he composed his work sometime between 650 and 780. But I believe we can we narrow this a little further. Avalokitavrata shows ample familiarity 27 with Dign aga’s (ca. 480–540) Praman : asamuccaya, as does of course Sthiramati, and even once expressly refers to it by title.31 The prevailing opinion that he also knew Dharmak rti (ca. 600–60) and was aware of his further development of Dign aga’s logical and epistemological theories is not based on the fact that he cites Dharmak rti by name or his writings by their title. Rather, if anything, this notion appears to be derived from several passages in his work where, for instance, he signals the three kinds of logical reasons (hetu) of an inference, namely, the ones of non-perception (mi dmigs pa, anupalabdhi), essential identity (ngo bo nyid, svabhava ), and causality (bras bu, karya ).32 Dign aga does not know this triad, and the earliest extant Buddhist work in which we do encounter it is the first chapter of Dharmak rti’s Praman : avarttika . It may not be important or even particularly relevant, but this is of course not the order in which these three are there discussed by Dharmak rti. In fact, the sequence of these three adopted by Avalokitavrata – I do not think it unreasonable to assume that the sequence in the translation reflects the one of the original – is found in the Praman : aviniscaya , II: 28, and in the Nyayabindu , II, but not in the Hetubindu or Vadany aya , which have: svabhava – karya – anupalabdhi. Later, in his comment on a passage of the pp’s seventeenth chapter ad MMK, XVII: 21, he reiterates his own earlier sequence of the three, and suggests there that there were those who claimed this triad as well as three valid means of cognition, including the kind of knowledge derived from credible scripture or from a credible person (yid ches pa’i lung gi tshad ma, *apt agamap- raman : a).33 Given that, if he does not countenance the latter as a separate epistemological category, Dharmak rti is at best dubious about its validity and place alongside perception and inference, it is unlikely that Avalokitavrata had him in mind, the more so since he would then have to be, in his view, a partisan of the idea of the socalled ‘‘non-perishing pheno-menon’’ (avipran: a sa ) that plays a role in the various Buddhist theories of karma.34 As U.T. Kragh pointed out, avipran: a sa is sometimes associated with the Sam: mat ya [and on a rare occasion with the Mah asam: ghika] school of Nik aya-Buddhism, but there is a persistent problem with identifying the individuals behind the objection in MMK, XVII:12, and their proposal of the notion of avipran: a sa . Contrary to the prevailing view that the commentarial literature does not identify them, Avalokitavrata glosses the opposing ‘‘others’’ (apare, gzhan dag) mentioned in the relevant passage of the pp with the Bye brag tu smra ba (*Vaibh as: ika)school of Nik aya-Buddhism.35 In fact, he does not once throughout the entire discussion even allude to the Sam: mat ya, let alone the Mah asam: ghika. Though purportedly contemporaries, Dharmak rti and Candrak rti do not seem to have known of one another’s work. Or if they did, they do not provide any conclusive evidence for this in their extant oeuvre. Avalokitavrata nowhere suggests, or at least I have not come across a single occasion where I found it to be the case, that he was familiar with the writings and ideas of any of the great late seventh or eighth century Buddhist thinkers that have come down to us. Slim as these data admittedly are, they do make me incline to place Avalokitavrata’s floruit closer to the beginning of the middle of the seventh than to the middle of the eighth century. Hence, I suggest that he should be assigned to circa 700. To be sure, it is not in the least disconcerting that he does not seem to have been known to the otherwise well informed Chinese scholar-traveler Yijing (635–713/4). But it is a bit peculiar that such very well-read late eighth century thinkers S´ antaraks: ita and Kamalas la also do not once mention or refer to him in their oeuvre. Indeed, as far as I am aware, save for Atisa’s (ca. 982–1054) Madhyamakopadesa titled (nama ) Ratnakaran: d: odghat :a and the *Bodhimargaprad ipapanjik a that is perhaps wrongly attributed to him as it stands,36 the later Indian Buddhist literature does not once take notice of him or his massive work. Bh aviveka wrote37 in the pp’s first chapter ad MMK, I: 1a‘‘not [arisen] from itself’’ (na svato), that nature (ngo bo nyid, svabhava ), God (dbang phyug, isvara), the Person (skyes bu, purus : a), primal matter (gtso bo, pradhana ),38 time (dus, kala ), and N ar ayan: a [sred med kyi bu = Vis: n: u],39 etc. are ‘‘bad causes’’ (rgyu ngan pa). He argues that these six causes are bad, that is, lack reality, in the sense that they can neither logically nor ontologically function as agents that are capable of causing animate or inanimate things into being. Other than simply calling them ‘‘those who claim the Person as a cause’’ (skyes bu rgyur ’dod pa dag), Bh aviveka does not associate the adherents of the purus : a as a creator with those whom Avalokitavrata specifically calls ‘‘Ved antav adin-s’’ (rig byed kyi mtha’ smra ba dag),40 although he does cite in this connection the R: g-veda or the Svet a svatara-Upanis: ad and the I sa-Upanis : ad. In point of fact, and this can by no means be dismissed as an entirely insignificant datum when it comes to authenticating other texts that are attributed to him, he no-where uses the designation *Ved anta[v adin] in the pp or, for that matter, in his *Hastaratna [or, less likely, *Karatalaratna] that is
only available in the Chinese rendition of Xuanzang (600/2-64).41 To
this sextet, Avalokitavrata adds the following:42 Tshangs pa – Tibetan
tshangs pa is used to render brahman, brahma and brahma –, the
lord of living beings (skye dgu’i bdag po, prajapati ), Manu (ma nu),
what is ordained by heaven (gnam gyis bskos pa, daiva), Dru b a na of
the Par sig-s (*paras ika) (par sig rnams kyi dru ba na ), and Yu na of
the Bar bar-s (bar bar rnams kyi yu na). Unfortunately, he glosses
only the ontological and, in the relevant instances, the theological
positions implied by the terms mentioned by Bh aviveka and, with the
exception of sporadic mentions of Tshangs pa, he does not elaborate
here [or elsewhere] on the other six he thus rather nonchalantly injects
into the discussion. The Bar bar-s, the last of these six, are the barbarians
(<Sanskrit barbar); it is not entirely clear whether the toponym
Barbara mentioned, for example, in the mid-tenth century
R as:t:rakut :a grant of Kr:s: n: a III (939–67)and by al B run (973–1048)
has anything to do with barbar.
43 But it appears that it does, since
these sources do locate Barbara in areas where foreigners lived.
Now [a] Bhavya drew attention to a mysterious deity called
Yon akadeva (Tib. Nam mkha’i lha) in the Madhyamakahr: dayakarik a
[= MHK], IV: 12, and its Tarkajval a [=TJ] autocommentary that are
not uncontroversially attributed in toto to Bh aviveka, the author of
the pp.44 The printed texts of the Tanjur as well as the text of the
‘‘Golden Manuscript’’ contain the occasionally problematic MHK and
TJ translations by Atisa and Nag tsho Lo ts a ba Tshul khrims rgyal
ba (1011/2-ca. 69), which date from circa 1050. A significant source of
large portions of the MHK/TJ other than the texts found in the later
printed editions of the Tanjur are the chapters on S am: khya,Vaises: ika
Ved anta, and M m am: s a, that is, chapters six to nine, as well as chapter ten on an aspect of Nirgrantha, that is, Digambara Jaina philosophy, which Bo dong Pan: chen ’Jigs med grags pa (1375–1451) alias Phyogs las rnam rgyal included in his encyclopedic De kho na nyid kyi ’dus pa, which he conceived writing in circa 1404.45 Although the manuscript of these pieces is in many places horribly disfigured through scribal ignorance, it can only be ignored at one’s peril. An interesting case in point why it should be consulted when working on these chapters of the MHK/TJ is presented by the reading of MHK, IX: 163c, and the corresponding comment in TJ, both of which are only available in Tibetan. This line in the MHK and its citation in the TJ read gang phyir mdze [’dze] mi [ni] ni [mi] sogs kyis [kyi] // in the various canonical recensions46 and the situation is not different in the TJ’s prose commentary. Chr. Lindtner evidently opted to read mdze mi and ni as an empty, metri causa syllable, and thus translated it by ‘‘leper.’’47 On the other hand, previously, Sh. Kawasaki had already rightly recognized the name of the M m am: saka patriarch Jaimini in syllables three to five and, indeed, his interpretation is unambiguously supported by Bo dong Pan: chen’s text, which has in every instance dzai mi ni. 48 Showing his intellectual independence while retaining his respect for the tradition, Bo dong Pan: chen does not always slavishly follow the TJ’s arguments in those parts of this study where he argues against the positions taken by these non-Buddhists. On occasion, he explicitly voices his disagreement with the arguments presented in the TJ, as, for example, in connection with MHK, VI: 28, where he states out right that the intent of the verse is explained differently in the TJ and that he did not see its comment to be a good one (legs pa ma mthong).49 Lastly, the remarks with which he concludes his series of refutations are quite telling as to what he thought about the transmission of the MHK and the TJ, and their Tibetan translation. He thus ends with the declaration that the [Tibetan text of the] TJ has errors, unclarities, and is confused (nor zhing mi gsal ’khrugs pa’i rnam pa can), and entertains the idea that these problems derive from the less than satisfactory job done by the translators.50 ’Gos Lo ts a ba Gzhon nu dpal (1392–1481) was also among those later Tibetan scholars who had recognized that all was not well with the way in which the texts of their Tibetan translations had been transmitted. This, I submit, can be gleaned from Zhwa dmar IV Chos grags ye shes’ (1453–1524) suggestion in his 1517 biography of his teacher to the effect that he had made some unspecified editorial revisions in the translation of the TJ [as well as of Kamalas la’s (8th. c) Madhyamakaloka ], albeit evidently without having been able to consult one or the other Sanskrit manuscript of these texts.51 Bhavya’s treatises were never translated into Chinese. The Lhan dkar ma and the ’Phang thang ma register incomplete Tibetan versions of both and it bears repeating that their entries suggest that the MHK-versetext might have actually been called the Madhyamakahr: dayatarkajval a . 52 Be this as it may, I am basically inclined to agree with Y. Ejima’s arguments and those of V.V. Gokhale and S.S. Bahulkar to the extent that a portion of these texts was originally titled Tattvamr :tavat ara and that this was a work most likely written by Bh aviveka.53 True enough, matters are rather complicated and one needs to be a great deal more nuanced about the question of their authorship. For this reason, I intend to revisit the conundrum of the oeuvre of Bh aviveka/Bhavya on a future occasion. While the fourth chapter of the MHK/TJ is titled On What is Real for the Sr avaka [Nik aya-Buddhism], it was in fact written with the express purpose of vindicating and establishing Mahayana Buddhism and its scriptural foundations as being indeed the word of the Buddha. There we learn that the Yon akadeva faithfuld held that he was created by Tshangs pa and that he taught a certain religious doctrine, which included the killing of ants in a golden vessel by piercing them with a golden needle, the killing of cattle and other animals, and the practice of incest (nyal po byas). The specious allegation was that, not admitting the notion of the four truths/realities [of Nik aya-Buddhism], adherents of the Mahayana accept another truth/reality and, inasmuch as it is different from the one Nik aya-Buddhism acknowledges, this truth/reality is like the one taught by Yon akadeva. Chr. Lindtner’s translation of nyal po [not: nal po!] by ‘‘having intercourse with one’s parent’’ is not quite accurate, for nyal po appears to be a general term for any kind of incestuous relationship. He correlated yonaka with yona, which is the Pali term for Sanskrit yavana, and I believe there is very little room for doubting that, well, Tibetan yu na is also related to it. This passage occurs in the context of one of a series of critiques levelled by the S´r avaka[= Nik aya]-Buddhists against the authenticity of Mahayana Buddhism and its scriptural basis. As reported by him, their sustained attack54 is launched by the allegation that the views espoused in what they believed to be the so-called Mahayana scriptures very much resemble those of the *Ved anta and the nihilists (nastika ) – it is probably not entirely insignificant in view of the problems that beset the compositional history and authorship of the MHK and TJ that this is the first time that the term *Ved anta makes its appearance in these two texts. This is followed by the allegation that one does not even encounter the term Mahayana in any work belonging to the Tripit:aka, a survey of the 18 different Nik aya-Buddhist schools, and a host of buddhological and soteriological questions such as inter alia the idea of a permanent tathagata , the tathagatagarbha and the rise of the use of mantras, all of which apparently ran against their sensibilities. Bhavya’s rebuttal of the allegation of any resemblance between the doctrines surrounding Yon akadeva and Mahayana Buddhism in MHK, IV: 68, as well as in the corresponding comments in the TJ, is not as strong as one might have reasonably expected in view of the seriousness of the allegation and its unpleasant insinuations.55 Lastly, with the exception of the reference to the Yon akadeva, the points raised about and the reactions to these by the Mahayana Buddhists are to some extent similar to the ones we come across in the third and fourth chapters of Vasubandhu’s Vyakhy ayukti and in the first chapter of Sthiramati’s Mahay anas utr alam : karabh as : ya or -vyakhy a , in particular.56 It is of course of some importance and can hardly be  accidental that the fourth chapter of the Tibetan translation of the Vyakhy ayukti and the fourth chapter of the Tibetan translation of the MHK [and the TJ] have two verses in common, as P. Skilling pointed out.57 These should have been found beginning with line 34c of V.V. Gokhale’s Sanskrit text, but they are not.58 From the work of Y. Ejima and others, we know that there are a large number of yet to be explained discrepancies between the Tibetan version[s] of the MHK and the sole surviving Sanskrit text. But it should be also noted that this is not all that these chapters literally have in common. In fact, the prose text that follows these two verses is shared by the Vyakhy ayukti and the TJ as well!59 Indeed, these two chapters have so many points of intersection that it will be well to study them together and not in isolation of one another. That the two verses are found in both the Vyakhy ayukti and the MHK [though extracted from the TJ!] was, as should be expected, also known to the Tibetans. For instance, Gung thang Dkon mchog bstan pa’i sgron me (1762–1823) cites both and comments on the slight differences in their readings that ‘‘apart from the somewhat different translations, it appears that [they] had the same Sanskrit text (’gyur khyad tsam las rgya dpe la gcig tu yod par snang).’’60 Though we cannot enter into its discussion, it is noteworthy that he draws several important implications from the different translations of the last two lines of the second verse.

The above list of six positions anent na svato has a reasonably close parallel with a passage that we find in the eighteenth chapter of the pp ad MMK, XVIII: 1c–d where, however, Bh aviveka registers but the four of isvara, pradhana , time, and N ar ayan: a as agent-creators (byed pa po, kartr:) or, what amounts to the same thing, causes (rgyu, karan : a).61 These enumerations also have several parallels in the third chapter of the MHK/TJ. 62 For instance, apropos of ‘‘[arising] from itself’’ (svato),the TJ ad MHK, III: 137–8, lists the following five: isvara , purus : a, pradhana , time, and N ar ayan: a. On the other hand, MHK, III: 223, substitutes ‘‘atom’’ (an: u) for ‘‘time’’ and has Vis: n: u instead of N ar ayan: a, while MHK, III: 248, has the five of Kesava (sred med) [= Vis: n: u], isa (dbang phyug), purus : a, pradhana , and an: u. Of the other designations, Avalokitavrata added to the first of Bh aviveka’s lists, the P aras ka are [most probably still for his time] Persians or those living in the subcontinent who were of Persian descent and had preserved some of their customs, so that the name Dru b a na [or: Dru v a na] is possibly a reference to the Zoroastrian deity Zervan/Zurv an.63 Avalokitavrata notes the P aras ka on three other occasions in his work, that is to say, in his comment on PP ad MMK, I: 1, anent God, and in two related passages of the pp ad MMK, XXII:10c–d.64 In the 61 S

first, he states in an aside that the P aras ka worship the water-deity (chu lha,*Varun: a). This might be a reference to the Mithraic deity Apa˛m Nap at.65 The second occurs via a passage from the pp where Bh aviveka has an interlocutor cast aspersions on several aspects of what enlightenment in a Buddhist sense entails, including the notion of the presumed omniscience of [a] Buddha or Tath agata [or buddha and tathagata !] and that, specifically,

de bzhin gshegs pa’i gsung rab ni thams cad mkhyen pa ma yin pas smras pa yin te / byed pa po dang bcas pa’i phyir dper na/bye brag pa la sogs pa’i bstan bcos bzhin no// The Word[s] (*pravacana) of the Tath agata [= Buddhist scripture] were [oral texts initially] spoken by one who is not omniscient, because they have a [human] agent/ composer as, for example, treatise[s] of the Vaises: ika [tradition], etc. Bh aviveka attributes this comment to ‘‘some[one]’’belonging to the M m am: s a tradition (dpyod pa can dag las kha cig gis smras pa). This is not the place to do so, but these and

Bh aviveka’s other related remarks anent these two lines [and Avalokitavrata’s long comment] bear a study in which these are compared to inter alia what the formidable M m am: s a philosopher Kum arila argued [contra the omniscience of the Buddha and contra the infallibility of his teachings] in the relevant passages in, for example, the Codana chapter of his S´lokavarttika [I: 2, 116–46, 169] and his Br: hat:t: ika , which survives only by way of quoted fragments.67 Kum arila is the earliest known M m am: saka to have taken issue with these Buddhist dogmas, but it does not of course necessarily follow from this that it is he against whom Bh aviveka is arguing. Indeed, were it otherwise, we would then either have to confront a serious problem with the widely accepted dates for Bh aviveka or cast doubts on the textual integrity of this passage of the PP – the same might then also hold for the relevant passages of the MHK and TJ! 68 There is, to be sure, some reason to believe that Dharmak rti was aware of Kum arila’s work and his critique of Dign aga, and it has also been argued that, in his Br: hat:t: ika , Kum arila reacted to the [early] Dharmak rti’s criticism of him.69 Not wishing to skirt potential text-historical problems that may be found with the Tibetan version of the PP [and the Sanskrit manuscript on which it is based], it should be pointed out that the three references to the views of exponents of the M m am: s a school or the Jaimin ya (Rgyal dpog pa) occur only in this chapter of [the Tibetan version of] the PP ad MMK, XXII:10c–d,70 and that, furthermore, it appears to be the only extant study of the MMK in which the M m am: s a are taken to task for the issues that are there addressed.71 As is perhaps to be expected with the same team of translators and the high probability that the manuscripts of these texts had an identical filiation, the Tibetan translation of Avalokitavrata’s commentary ever so closely follows the Tibetan text of this chapter of the PP.72 On the other hand, the PP’s Chinese translation takes here a radically different route, for the passages under discussion are not found in the comments on MMK, XXII: 10, as in the Tibetan version, but rather in the explanations of MMK, XXII: 8! Indeed, the comments on MMK, XXII: 8, onwards have a number of immaterial as well as rather significant differences in both translations.73 For one, the Chinese translation contains two quatrains from the lun (*sastra ). Athough they are marked in exactly the same way as the text marks the verses of the MMK, that is, as ru lun jie shuo, both are absent from all the other versions of the MMK, XXII.74 Only the second is found in the pp’s Tibetan translation and, although it is there unsourced, Avalokitavrata identified it as deriving from Aryadeva (ca. ?3 c.); it is in fact Catuh: sataka, XVI: 23.75 For another, whereas the Chinese translation quotes a verse from an unidentified sutra (ru jing jie shuo), the PP’s Tibetan counterpart gives this passage in prose and does not at all indicate that it is quotation, let alone from a sutra!76 Further, in connection with the Chinese translation in general, it is noteworthy to point out that U.T. Kragh has already drawn attention to the fact that it is not only evidently based on a different Sanskrit manuscript of the PP than the Tibetan translation, but also that it on occasion contains interpolated passages from, among other authorities, Qingmu’s Zhonglun (*Madhyamakasastra ), which was translated into Chinese by Kum araj va (344–413) in the beginning of the fifth century.77 In other words, then, its textual history is rather complicated and requires much further inquiry. What is more, and this is of no uncertain significance for one of the principal arguments that has been advanced in favor of holding that the author of the PP and the MHK/TJ is one and the same person, the Chinese translation does not contain the reference to a ‘‘section, chapter’’ (skabs, *prakaran: a) of the Dbu ma’i snying po’i de kho na nyid la ’jug pa (*Madhyamakahr: dayatattvavat ara ) that we uniquely find in the 25 chapter of the PP’s Tibetan translation.78 The corresponding passage in Avalokitavrata’s commentary even introduces at this point the title Rtog ge ’bar ba [= TJ] as well as the section/chapter-title of Rnal ’byor spyod pa pa’i de kho na la ’jug pa [= *Yogac aratattv avat ara ], and explicitly states that it was written by Legs ldan ’byed (*Bh aviveka).79 Of further interest is also that, in the above quote [at n. 66], the ‘‘some[one]’’ who belonged to the M m am: s a tradition refers to at least one *Vaises: ikasastra , the earliest extant specimen of which would be the commentary by Candramati or Maticandra (ca. 6th c.). On the other hand, if we need to maintain, as I think we should, a strict parallel with what was said about the Buddha’s pravacana, then the *Vaises : ikasastra may very well have to be identified with Kan: ada’s Vaises: ikasutra . Rejecting the assumptions underlying this remark, Bh aviveka turns the tables and begins to take his opponent [or: opponents] to task for holding the Vedas to be revelatory oral texts (sruti ) that had no [potentially fallible] human hand in their composition and states that, as a matter of fact, their author or authors was or were demonstrably fallible. He writes in the PP that:80 P

PP – Tibetan translation: rig byed ni byed pa po tshul khrims ’chal bas byas par shes par bya ste/’tshe ba dang/ bgrod par bya ba ma yin par ’gro ba dang/chang ’thung ba chos su ston pa’i phyir/dper na par sig la sogs pa’i bstan bcos bzhin no// One should know that the Veda was composed by author[s] with bad ethics and morals (tshul khrims ’chal ba, *duh: s ila), because, as a [form of] religion (*dharma), it teaches violence (’tshe ba, him: sa ) [, that is, ritual sacrifice], going for something that should not be gone for (*agamyagamana = incest), and the drinking of alcohol;81 for example, like the treatise[s] of the *P aras ka (*paras ikasastra ), etc.
 PP – Chinese translation: Again, the Weituo (Veda[s]) were composed by evil people who broke [their] vows, because they talk about killing living beings (sha sheng) and offering [them] to Tian (?deva, Heaven/God), close relatives living together and doing bad deeds (qin chu xiexing), and drinking alcohol, etc., like, for example, the words of Puoximujia (*P aras ka) people. The ninth chapter of the MHK that deals with the M m am: s a contains a parallel passage to this. Namely, in MHK, IX: 31, the author mentions, in the very same context, violence to animals and the drinking of alcohol, and also draws attention to the text[s] of the Persian [Zoroastrian] Maga or the Maga brahmins (magasastra ) 82 – it is worthy of note that Bh aviveka nowhere uses the term maga in his pp or his *Hastaratna/*Karatalaratna. The author of the TJ’s comment on this verse of the MHK identifies the *P aras ka,83 etc. as those who practice him: sa , drink alcohol, and engage in incest, which he characterizes as sleeping with one’s ‘‘mother, sister and/or daughter’’ (ma dang sring mo dang bu mo). And he states rhetorically, as if he would not already know the answer to this question: ‘‘What is the difference between the Maga treatise and the Veda…?’’ Commenting on the PP’s reference to the *P aras ka, Avalokitavrata draws parallels with the M m am: s a by mentioning four different Vedic sacrifices (mchod sbyin, *yajna ), the caturm asya- (zla ba bzhi pa[’i] mchod sbyin) and the ?pasv alambha/pa subandha - (phyugs kha ’dog pa[’i] 84 mchod sbyin), the gosava- (go sa ba[’i] mchod sbyin) and the sautraman : i-yajna (sau ta ma ni[’i] mchod sbyin),85 where, in his opinion, similar unseemly acts are affirmed; he writes:86 dper na par sig la sogs pa’i bstan bcos ni byi smyil la sogs pas srog chags la ’tshe ba dang/bu sring dang nyal ba la sogs pas bgrod par bya ba ma yin par ’gro ba dang/lkog tu myos kyang bla’i zhes bya ba la sogs pas chang ’thung bdag chos su ston pa’i phyir/ byed pa po tshul khrims ’chal bas byas par shes pa byas pa de bzhin du/dpyod pa can dag gi rig byed kyang zla ba bzhi bzhi zhing zla ba bzhi pa zhes bya ba’i mchod sbyin 82
dang/zla ba drug drug cing phyugs ’dog pa zhes bya ba’i mchod sbyin byed pa’i tshea phyugs kha ’dog cing mchod sbyin byed pas srog chags la ’tshe ba dang/go sa be zhes bya ba’i mchod sbyin byed pa’i tshe ma dang bu sring la sogs pa dang lhan cig tu gcer bur phyung te/phyugs bzhin du rkang lag bzhi sa la btsugs shing rtswa za ba ltar bcos te mngal gyi sgor lces ’dag pa dang/bshang pa’i lam du snom pa dang/’khrig pa lhag par spyod pa la sogs pa dang/bu med pa la mtho ris su ’gro ba med do zhes zer zhing rang gi dbang pob dul bar bya ba dang/mtho ris su ’gro ba’i lam ni bu yod par bya ba yinc no zhes phyugs bzhin du ma sring la sogs pa dang ’chol bar spyod pa la sogs pas bgrod par bya ba ma yin par ’gro ba dang/sau ta ma ni zhes bya ba’i mchod sbyin gyi tshed chang dang btsos pa’i chang dang bag med par ’thung ba la sogs pas chang ’thung ba de dag thams cad mtho ris su skye ba’i rgyu dge ba’i chos yin par stone pa’i phyir/byed pa po sdig pa can tshul khrims ’chal pa zhig gis byas par shes par bya’o// a. SDE: byed pa de’i tshe; GM, byed pa’i tshe. b. SDE: rang gi dbang po; GM: dbang po. c. GM: ma yin. d. The expected phrase would have been: byed pa’i tshe. e. SDE: bstan; GM, ston. Because, for example, treatise[s] of the *P aras ka etc. teach [as their] own religion (dharma): [1] Harming animals by ?byi smyil etc., [2] Going for something that should not be gone for by sleeping with [one’s] siblings (bu sring) etc., and [3] Drinking alcohol by the so-called highest one (bla’i), though secretly intoxicated, etc., one was made aware that [they] were written by author[s] with bad ethics and morals. Likewise, because also the Veda[s] of the M m am: sakas teach that all these [following practices] are good-dharma cause[s] for [re]birth in heaven: [1] Harming living beings (srog chags, *pran : aka) [is explained] by killing, that is, sacrifing animals (?pasv alambha/pa subandha ) when performing every four months the [quarterly] caturm asya sacrifice and every six months the ?pasvalambha/ pasubandha sacrifice, [2] Going for something that should not be gone for [is explained] by [a], when performing the gosava-sacrifice, one disrobes together with [one’s] mother, siblings, etc and, like cattle, goes on all fours, as if eating grass, licks the entry of the womb [=vagina], sniffs at the anus, copulates etc. and by [b] perverse behavior, etc. with [one’s] mother, sister, etc. like cattle, that is, alleging that one does not go to heaven when without a son, one says that one’s organ needs to be pacified and that the path going to heaven requires that one has a son, [3] Drinking alcohol [is explained] by when the sautraman : i sacrifice is performed, one drinks alcohol, alcohol that is ?distilled (btsos pa’i chang), one drinks immoderately, etc., one should be aware that [the Vedas] were written by evil authors with bad ethics and morals. Bhavya states something similar in the TJ after a verse that is not part of the MHK as we know it. He places this verse in the mouths of adherents of the Veda (rig byed pa, *vaidika) immediately following
his initial assessment of MHK, IX: 31, and his rhetorical query whether here is any difference between the Maga treatise and the Veda, because, after all:87 rig byed tshe ba gang bstan pa // bstan bcos gzhan du smad gyur kyang // des ni sdig byas mi ’gyur bar // gzhan du skye bdag gis bstan to // Though the Veda that teaches violence is belittled in other texts, that it does not create sin is taught elsewhere by *Praj apati (skye bdag < skye dgu’i bdag po). And this is followed by citations of a litany of Vedic injunctions and references to a host of other practises, many of which running parallel those to which Avalokitavrata has drawn attention, which the author held as being condoned in the Vedic texts. Just as I will refrain from entering into a discussion of these otherwise fascinating expository remarks and critiques, so I will also not deal with the other arguments presented in this long passage of the PP and the much longer one in Avalokitavrata’s commentary surrounding these issues, not to mention the other corresponding passages in the MHK and the TJ. Suffice it to say for now that while Bh aviveka himself does not dwell on the phrase ‘‘going for something that should not be gone for’’ or the drinking of alcohol in the PP, he does there discuss at some length the question of the him: sa of sacrifice/worship and the context of the various grammatical categories used in the description of the sacrifice on which the M m am: saka commentators have bequeathed upon us a rich hermeneutic legacy.88 The exploration and analysis of these will have to be reserved for another occasion. Towards the end of these disquisitions, Bh aviveka notes an attempt to rebut his kind of criticism of sacrifice and the consumption of the sacrificed animal’s flesh, which he also dismisses as ‘‘not good’’.89 And it is at this juncture that Avalokitavrata writes:

…zhes bya bas ni dpyod pa can dang / [par sig dang/mo go sha dang / mu sul man
dang / tar sag la sogs pa dag gi gzhung]
91 las phyugs pa la sogs pa dud ’gro rnams
mchod sbyin du spyad pa dang/mis bza’ ba’i rkyen du bya ba yin pas de dag mchod
sbyin du spyad pa dang/mis zos pa la sdig pa med do zhes zer ba’i gzhung ston te/…
…[the] statement [in the pp]…indicates passage[s] from the [textual] traditions of the
M m am: s a, the *P aras ka, Mo go sha, Mu sul man (< ?Persian: musulman ), Tar sag
(<Middle Persian: tarsak /tarsag ), etc. that allege: ‘‘Since animals such as a cattle etc.
will be enjoyed in the sacrifice and will function as a part (rkyen du bya ba)
92 of what
is eaten by a person, there is no sin in having enjoyed sacrificing them and in having
[them] being consumed by a person.’’
Avalokitavrata thus extends Bh aviveka’s critique of the M m am: s a
and the *P aras ka to a series of other religious groupings with
which he finds similar faults. Of these, I am unable to say anything
about mo ga sha. As my colleague P.O. Skjaervo kindly informed
me, the expression tar sag is no doubt derived from Middle Persian
tarsak /tarsag and refers to Christians. If not the earliest, this is one
of the earliest references to Christians in an Indian Buddhist
treatise, albeit via its Tibetan translation. Given its linguistic origins,
it cannot have much to do with the Christians of Kerala, in
South India, who lay claim to being the oldest Christian communities
of India, as they were allegedly directly inspired by the
Apostle Thomas.
The expression mu sul man that is sandwiched between these two
is, to my understanding, the oldest attestation in a Sanskrit work,
if through the medium of its Tibetan translation, of what is in all
probability a reflex of New Persian musulman . The conquest of
Persia by the Muslim Arabs under the leadership of the Omar ibn
al-Khatab coincided virtually with the assassination of Yazdgird
III (r. 632–51/2), the last Sassanian ruler. It now appears that the
Arabic terms muslim, muslimun and thence Persian musulman may
be of a later vintage than had been assumed not all that long ago.
So far, the earliest datable epigraphical occurrence of muslim is met
with in the famous Kufic inscription in the Dome of the Rock, in Jerusalem, which dates from A.H. 72 (691).93 R.G. Hoyland has recently maintained that the earliest Arabic inscription in which it occurs is the A.H. 123 (741) inscription at ’Ayn al-Jarr,94 although he does suggest [in n. 64] that ‘‘though not dated, very likely earlier is an inscription from Qusayr Amra seeking forgiveness for ‘the crown prince of the Muslim men and women.’’’ Whatever the case may be, it can now be recognized that the notice of mu sul man by the learned Buddhist scholar Avalokitavrata, however uninformative this mention is otherwise, is one the earliest literary testimonies outside the Koran attesting to the way in which adherents of Islam referred to themselves. Furthermore, it is to my knowledge the first attestation of an awareness of Islam and the fact that it does not proscribe the consumption of meat in that part of the Indian subcontinent that lies to the east of the Indus River