REVIEWS

Paul Horsch, Die vedische Gdth~- und Sloka-Literatur. Verlag, 1966. 516 pp.

Bern, Francke

We are here confronted with a work on the Vedic gathSs and Mokas~ studied on such a large scale as unprecedented. 2 There is a n excellent survey of this kind of stanzas by L. Renou in his article entitled "Les vers ins6r6s dans la prose v~dique", Asiatica (Fs. Ft. WeUer, 1954), pp. 528-534. But the stanzas inserted in the Vedic prose in general are not coextensive with what our author understands under the VGS in this book. As basic materials for his research Horsch has collected from the Brahma.nas a n d older Upanis.ads all the anonymous stanzas 3 explicitly labeled as gSth5 or Moka by the texts themselves (see esp. pp. 348-352). The limitation, however, of the VGS thus defined in distinction from the mantras of the Sa.mhitas (cf. p. 1) is subject to certain exceptions. A RV-stanza 1.164.2, for instance, is included in Horsch's collection on account of its being cited as a gloka in Pragna Up. 1.11 ( = no. 175, cf. p. 350), while the stanzas in, e.g., AB VIII.27, though essentially glokas, are excluded from the collection because they are not introduced as such by a proper formula (see p. 428 with n. 2), or B,~U IV.4.23 introduced by a formula, tadetadrcdbhyuktam, in spite of its content and form, is not regarded as a regular gloka (see p. 172 under no. 153). The difficulty of making a strict demarcation between the VG~ proper and those that belong substantially, though not formally, to the same category has been fully recognized by Horsch and many of the stanzas of the latter class are mentioned in the commentary of the regular ones (cf. p. 349). Though too much importance may seem to have been attached to the introductory formulae, 4 the stanzas selected according to a rigid standard are numerous and multifarious enough to answer the purpose of the present book. In this work, it is said, the chief weight has been placed on the following three points (see pp. 1-3): firstly to investigate more exactly than hitherto the relation of the VGSliterature to Buddhism and Jainism, secondly to indicate the after-effect of this literary genre in Hinduism and thirdly to inquire into the origin, that is, the institutors of the gfithic verses. Horsch maintains that these anonymous fragments of stanzas, which originated mostly from secular and popular circles, already contain the rudiments of the later Hindu literature, show many associations with the new turn of thought 1 Hereafter abbreviated as VGS. For brevity's sake we also speak of the G(fithh)literature or -tradition. 2 Cf. the same author's condensed yet lucid summary, "Vedische GfithAs und ~lokas", WZKSO, 6 (1962), pp. 1-14. a As well as those, the authorship of which is ascribed to mythological beings. 4 For the textcritical importance of these clich6s, see pp. 455-458.

28

REVIEWS

emerging about the middle of the first millennium B.C. and form, as it were, the missing link between Brahmanism and ~ramanism (cf. p. 4). It may be added in advance that Horsch's chief object repeatedly asserted throughout the work is to prove that the ultimate source of both the epic and the drama of India is intimately connected with the G literature, which is derived in its turn from the non-hieratic or unorthodox spheres such as the Ks.atriya-milieu or the Vrgtya-community. In the first Part ("Die Randgebiete der vedischen Literatur", pp. 5-71) Horsch spreads out a wide panorama of the orthodox as well as marginal domains of knowledge listed in various Vedic texts. No less than twenty-five marginal spheres are enumerated and each of them is fully explained with regard to its meaning and definition, its mention in the lists and actual examples, while the ample and apposite bibliographical references in the foot-notes bear testimony to the author's scrutiny a n d scholarship. 5 As the exceedingly rich contents cannot be briefly described, I merely draw the readers' attention to the items that are closely related to the main subject of" this book: gdtha (together with its special subgroup ndrdgarnsi) and gloka (pp. 9-10, 11-12, 31-32, 53) in the first place, next itih~sa and pura.na (pp. 12-15, 23, n. 2, 48-49, 52-53). After having referred to the pros and cons of the 'Streitfrage' whether the various branches of knowledge were already laid down in the form of collections or compendia for the relevant study at the time when they were first mentioned as such (pp. 47-51), Horsch tries to put to the rights the point at issue and advances his own opinion t h a t the itihasa-purd.na 6 seems to have early tended to form a collection while the gffthas a n d glokas had the least chance of being gathered into a work (pp. 52-53). In order to determine the origin of the marginal branches of knowledge Horsch examines the context in which their chief lists appear and fiads the latter closely connected with the domestic rites, the ks.atriya-ritual and the gilpas as forerunners of the corresponding Hindu-writings, that is, the Sm.rti and special works, and, moreover, Horsch indicates the fact that they are adduced by the unorthodox Vrhtyas as well as by such supporters of the gloka-tradition as Y~jfiavalkya and Nfirada. All this. clearly points to the secular and popular environment opposite, though mostly standing near, to the hieratic Brahmanism (pp. 54-55). Horsch tries further to endorse the above results by collecting similar lists of sciences and arts from the vast literature of Buddhism (pp. 55-64) and Jainism (pp. 64-67). Out of many useful informations we owe to Horsch's astounding assiduity I should like to retain here the following statements. Against our expectation the older Buddhism knows only the three Vedas without including the A V, and it makes no mention of the Upani.sads. T h o u g h a special position is reserved for the itihdsa and purd.na, the glokas in the technical sense, that is, as the non-mantra verses inserted in the Vedie prose, occur in one of the Buddhist lists, so also the gdthffs outside the so-called navatiga-classification (of. p. 217, n. 1) of the Buddhist canon (p. 64). Lastly a brief but welcome survey is given of the post-Vedic development of the literary and scientific branches from PA.nini over Kautilya and Manu down to Madhusfidana Sarasvati (ca. 1500 A.D.), many of which go back to the Vedic period and were comparatively early systematized. "'Die Smrti-Texte sind in der vedischen Periode verwurzelt, so vor allem Itihfisas und Purfin.as, wiihre~d manche Dharmaghstras aus dem herrenlosen Vorrat popuF,irer, ethischer Maximen gesch/Spft haben" (p. 71). The second Part ("Sammlung der vedischen Gfithfis und Slokas", pp. 73-212) is devoted to a collection of 194 VGS which are in accord with the definition as stated in 5 I miss, however, B. Faddegon, "The catalogue of sciences in the Ch~ndogyaUpani.sad (VII, 1, 2)", AO, 4 (1926), pp. 42-54. s Mentioned as the fifth Veda first in ChU VII.1.2 and its parallel passages, see pp. 12, 52.

REVIEWS

29

t h e beginning of this review. The distribution of the verses to the texts is quite uneven. The largest portion is occupied by the AB (nos. 1-50), next in number are the JB (nos. 53-80), SB (nos. 88-110) and B~U (nos. 121-153), while the other texts (KB, PB, JUB. GB, Azf, ChU, etc.) offer a small number of examples (cf. also pp. 458-461 : Concordance of the same verses occurring in different texts). Each stanza is given in the original, translated and annotated. In doing so Horsch has not spared himself the trouble of emending textual corruptions or metrical deficiencies, explaining ditticultt words or phrases and citing parallel or similar verses from Hinduistic and Buddhist texts. Every Sanskritist should be grateful to Horsch for his interpretation of many stanzas of ritual and speculative import that are often not very easy to understand. Taking in consideration the importance of this Part which forms the foundation for the next Part or the main body of the work I should like to make some remarks of a philological nature, while passing over insignificant misprints and slips of the pen. Under no. 2 (AB V.30.3) Horsch writes: "Agnihotra soil nach Sonnenaufgang vollzogen werden: A B 5, 29,8; vor und nach Sormenaufgang: 5,30,1, worauf sich der Vers bezieht." But AB V.30.1 clearly condemns the time before the sunrise for the morning agnihotra and prescribes the time after it. The last line of our verse, divaivanyaj juhuyan naktam anyat, merely refers to the morning and evening agnihotra, not to the morning agnihotra before and after the sunrise; cf. also nos. 3 and 5. Decisive is the precept tasmad udite hotavyam V.29.8 and 30.4. This is the only time allowed by the Aitareyins, s. Weber, ISt., X, p. 329 (cf. IX, p. 292), somewhat compromised by .4SS 11.4.24, see Dumont, L'Agnihotra, p. 179. - Ad no. 4 (AB V.30.11), line a Horsch remarks: "Schon Weber ... konjizierte yo anstelle von so" (similarly also WZKSO, 6, p. 5), but it is not so in a but in d that should be emended to yo, see Geldner, ZDMG, 65 (1911), p. 306 ("S. 360 f." Horsch, p. 76 being a mere misprint), Oldenberg, NGGW (1911), p. 464, n. 2 (Kleiae Schriften, p. 1418). And it is not Geldner but Charpentier (ZDMG, 64, p. 65 if). who compared for the first time Jataka 488 (Bhisajataka) with the epic version, while Geldner has the credit of having added the Aitareya verse to the discussion of the 'Lotusdiebstahl'. Further the statement: "abhi~astfit, apaharfit vedischer Subjunktiv ~in the main clause~ (Keith S. 69)" is rather misleading as the first word is an imperative in -tat (s. Whitney, Gr., w end), correctly designated so p. 365. - Under no. 52 (KB XXVII.1) Horsch says on account of a curious form yamaki: "Keine volkstfimliche Form, wohl ffir yamfini wie in RV 4, 18, 3b", where .gamani is used, but cf. p. 364. It is simpler, however, to admit the form as an isolated forerunner of the type pacataki (: pacati) of the native grammarians, see Renou, Asiatica, p. 529 ( = yami), Gr. scte., p. 247 (: w cf. Wackernagel, Ai. Gr., III, p. 447 end (: w a, 13, Anm. where pacataki should be read instead ofpacakati). Under no. 66 (JB II. 73-74) Horsch connecting ahndya (d) 'ffir lange Zeit' with anugatya (c) rightly prefers to take the verse as an expression of 'endgOltige Uberwindung des Todes', but the same verse is later referred to in support of the punarm.rtyu,doctrine when he states: "Naeh Strophe 65/iberwindet man auf den G6tterpfaden den Tod; der folgende Vers jedoch prfizisiert: ~ff/ir lange Zeit>;' (p. 297). - That sm.rta.h in no. 113 (GB II.2.5), line b means 'fiberwacht' or 'fiberdacht' does not convince me, nor Horsch's comment: "Weil der Brahman-Priester das Opfer vom S~iden her fiberwacht." Such a well-guarded yaj~a cannot be blamed as hin&iga, etc. in c and d. There is no direct bearing on the brahman-priest's protecting the sacrifice from his seat to the south of the ahavaMya, cf. the passage quoted by Horsch, p. 176 (no. 156), The main purport of the verse is that the sacrifice bending to the south 7 is regarded or called daks.it~atat3 'southward', that is, turning to the direction of Yama and death, and For dak.si.naprava.na in contradistinction to udakprava.na, cf. e.g. Caland, Die altind. Todten- u. Bestattungsgebrauche, p. 31, Caland-Henry, L'Agnist.oma, p. 7 (w Caland, n. 1 ad ,,tpSS x.20.1.

30

REWEWS

therefore defective and not divine. - Horsch's interpretation of no. 178 (PU V. 6) involves two assumptions: (1) M~trd mrtyumatyah in a is to be read mOtrdm.rtyu ~ with the double sa.mdhi for ~ amrtyu ~ and (2) anaviprayuktd.h in b = aviprayuktdb.. Though the meaning of the verse is by no means quite clear to me, I venture to offer my solution of the riddle. It is, of course, beyond doubt that a-kdra, u-k&a and ma-k&a are understood by tisro rn~trd.h. But why do they contain death (rn.rtyumatyat.O? Simply because ma-kdra suggests mrtyu. Strange though it may be, even satyam is said to implicate amrta and martya (or an.rta), s The three rndtrds are used (prayukta) combined together (anyonyasakta) or separately (anaviprayukta), ~ but when used without knowledge they may bring deadly effect. If, however, correctly employed (samyakprayukta) in three abhidhydna-exercises, bdhya corresponds to P U V.3 (ekam~tra), madhyama to lb. 4 (dvimdtra) and abhyantara to lb. 5 (trimdtra), the knower has no fear. - Under no. 185 (MaitU VI.34) Horseh, after having mentioned var. lec. sam.sdra.h, explains sa.msdram in a as accusative or adjective. But in consideration of other cases of the lifzgavyatyaya in this Upani.sad ~ should be taken as a neuter nominative, cf. no. 190 (MaitU VI.34), line c: etaj j~6nan.7 ca rooks.an3 ca. 1~ Horsch himself admits that sa.msdro is here identified with citta, see p. 300. The third Part ("Literatttrgeschichte") is divided into five chapters, the first of which ("Allgemeines") treats the concepts ofg~thd (pp. 213-223) and Moka (pp. 223-229) and the relation between them (pp. 229-249) from the viewpoint of the history of literature. Horsch thoroughly pursues the development, use and notion of both terms in Brahmanism, Hinduism, Buddhism and Jainism through the ages as well as in the Prakrit, and modern Indo-Aryan languages. Among many interesting assertions and obervations I should like to mention here only one point that is important for the Vedic students, that is, a gradual retreat of the term gdthd (and much rarer yaj~ag~th~) giving way to a younger term Moka. The monopoly of the latter is complete in the old prose Upanis.ads (see esp. pp. 225 and 234). According to Horsch this depreciation and exclusion of the g6thds because of their popular and secular character was, however, restricted to the higher and radical circle of the Brahmins, while the evidence of the Epics and Purfin. as as well as of Buddhism and Jainism speaks for the preservation of the old value of the term (see esp. pp. 229-233, 243-244, 247-249; cf. also pp. 451-452, 456-458,468). In the second chapter ("Inhalt", pp. 250-305) Horsch gives a detailed description of the contents of the VGS cited in Part n . After having given a glance at the yaj~agdth~s 'ritual stanzas', brahmodyas 'riddle-stanzas' and vdkovdkyas 'dialogue stanzas' (pp. 250-251, cf. also pp. 18-19, 217-219), Horsch dwells at length on a special kind of the gdthds: ndrdda.msfs 'Mfinnerlobverse' (pp. 251-284, cf. also pp. 9-12). The kings praised in these verses (AB nos. 38-50, SB nos. 96, 99-110) are investigated one by one as to their land and people as well as their officiating priests (pp. 253-275, cf. the synop: tic table pp. 251-253). The following points result from this historical study (pp. 270280): the model of the ndr~a.msis can be found in the ddnastutis of the R V, the geographical scene can on the whole be sought in Madhyadega, most of the kings seem to have lived ca. 1200-900 B.C. while the eulogical verses were later composed, say till about the seventh century B.C., the historicity of the persons concerned, with one or two exceptions, enhances the value of this sort of research, and, lastly, the original

8 Cf. ChUVIII. 3.5:t~nihavaetdnitrg.~yak~ara.nisa-ti-yamiti, tadyatsattadamrtam, atha yat ti tan martyam, atha yad ya.m tenobhe yacchati; in the parallel passage of B A U V.5.1 ti represents an.rta. 9 In support of his explanation ana- : a- Horsch refers to Wackernagel, Ai. Gr., II, 1, p. 80: w b, Anm. where, however, one finds rather a negative verdict: "Aber das Ai. bietet keine sichern Beispiele." 10 For other examples, see N. Tsuji, "Some linguistic remarks on the Maitri Upani.sad", Is. Yamaguchi (Kyoto, 1955), pp. 96-97.

REVIEWS

31

source of this literary genre derives itself from the gathas of the AB sung at the rdjas~ya 'royal consecration'. After a short exposition of the true nature of the so-called ndr~dat?lsdni in SSS XVI. 11 (pp. 281-283) and a reflexion on the relation of the ndrdda.msis to the Great Epic Horsch takes up the next topic: itihdsa (pp. 284-295). In search after the real epic forerunners in the G-literature Horseh discusses the famous ~una.h~epa-legend - once again after Ft. Weller (1956) and H. Lommel (1964) - naturally with special reference to the inserted stanzas (cf. AB and S~S nos. 7-37) whilst the N~rada-g6thds (pp. 291292), the Indra-gdthds (p. 292) and the Vi~v~mitra-episode (pp. 293 f.) form a supplement. A cursory view is further given of the itihdsa-motif in other gdthds (p. 294) and of the Buddhist parallels of the ndrddamsis, brahmodyas and dkhydna-gdthds (p. 295). More importance than hitherto should in Horsch's opinion be attached to the cosmological and speculative stanzas (pp. 295-305), some of which announce the new epoch of the Indian thought and reveal the fundamental ideas of the Upani.sadic philosophy (cf. esp. B.,(Unos. 121-153, ChUnos. 154-162): sa.msdra, karman, brahman and dtman, the dreamless sleep and emancipation, the germs of the later S~.mkhyaYoga and the maya-doctrine, etc. The greater part of the third chapter (pp. 306-365) is assigned to the problem of the dkhydna and the beginnings of the Indian drama (pp. 307-357). As it is important to know what Horsch means by the term dkhydna I shall quote here his own definition: "Im folgenden verstehen wit unter ~khy~na den formellen Aspekt einer Erz~ihlung: eine Mischung aus Prosa und eingestreuten Gfith~s, wobei letztere in der Regel den dramatischen H6hepunkt, oder den Dialog, oder Sinnspr/iche und/ihnliches repr~isentieren" (p. 308)? a The dkhydnas in the Br~hma.nas are of two kinds: exegetic (Type: Purfiravas and Urva~i, SB XI. 5. 1) and independent (Type: Suna.h~epa-legend, AB VII. 13-18) (p. 316, cf. also p. 341, no. 2). Reviewing the much debated problem of the sa.mvdda-hymns of the R V Horsch concludes that Geldner's 'Balladenhypothese' (Philologenversammlung, Marburg 1913) is the most plausible among various theories expounded by Oldenberg, v. Schroeder, Hertel, etc., inasmuch as it necessitates no connecting prose; and that Thieme's thesis of the 'Legendenzauber' (ZDMG, 113, 1963, p. 69 ft.) is also to be taken into account (pp. 308-316, cf. also p. 341, no. 1)? 2 Passing over many interesting discussions and digressions and making use of Horsch's own summary of the chief results (pp. 341-343) I epitomize here as briefly as possible the main line of his argument concerning the beginnings of the Indian epic and drama. Horsch assumes various stages of development for both and complex factors that contributed to the development of the drama. He admits that both arose from popular-religious currents, the oldest expression of which is found in the VGS (p. 328). The independent type of the 6khydna (s. above) together with the naraga.msis and itih6sa-gdthds is to he regarded as the Vedic forerunner of the Indian epic. The mere recitation being sufficient to produce the grava.naphala 'fruit of listening' (cf. also p. 469), this kind of 6khydna was never presented as a play in its proper sense though we hear of dance, song and recitation accompanied by lute. Since about the fifth century B.C. we may reckon with the beginnings of the great epic poetry, the tradition of which is connected with the G-literature. About the same time the actors and theatrical art, first held in contempt, began to win the recognition of the higher classes (witness Phn. ini and then Agvagho.sa). In short the whole development can be reduced 11 On the functions of the VGS, s. further pp. 314, 348, 352. 12 In the main I agree with Horsch's view. O f course we need not stick to the designation ballad, but we must at any rate endeavour to interpret the R V-hymns just as they stand without having recourse to the hypothetical links in prose. For Horsch's criticism of L. Alsdorf's opinion (JOIBar, 13, 1964, pp. 195-207), see pp. 343-344.

32

REVIEWS

to main roots: the G-literature, out of which the epic and later the kcivya arose, and a lowly class of mimes, dancers and actors, who were originally despised by the monks and Brahmins. The aesthetic standard of the VGS is not uniform. The poetical value of the ya]Tta. g~thas, brahmodyas and n6raga.msfs cannot be regarded as high; quite different, however, are the dialogue- and itihasa-verses, maxims like the N~rada-gathas (AB VII. 13), the Indra-gath~s (ib. 15) and the speculative glokas of the older Upanis.ads(pp. 352-353, cf. also pp. 357-359). Lastly a study of the metrical (pp. 359-363) and grammatical features (p.364 f.) concludes this chapter on the formal side of the VGS. In the next or fouth chapter ("Herkunft", pp. 366-448) the origin of the G-literature in its connection with the social milieu is fully investigated. All the materials historical and legendary are focussed on the centralidea that the ks.atriya-elass played the most important part in the development of thisliterary genre. After a brief survey of the mythical speakers ofg6thas, Ngtrada (9. 368 f.) and Indra (9. 369 L), two personages who occupy the most significant place in the G-tradition, viz. Vi~vAmitra (pp. 370-380) and Y~jfiavalkya (pp. 380-401), are dealt with at length. Vigv~rnitra of the Snnah. gepa-legend is not identical with the seer of the R I/, and is himself a legendary figure that took over not only the name but also certain features of the older ~.'s.i.This fact, however, does not hinder to discern the ks.atriya-element in the episode and surroundings of Vigvgtmitra, who can be regarded as 'Hanptexponent der altvedischen, protoepischen G~thhTradition' (9. 379-380). In this connection Gfithin 'possessor of the gAthfis' as epithet of Viw father and descendants is also very meaningful (pp. 376 f.). The next chapter contains a very interesting essay on Ygtjfiavalkya in his relation to the G-literature (pp. 380-401). Though the discrepancy between the ritualist YAjfiavalkya and the representative of the Upanis.ad-philosophy early attracted the scholars' attention, TM Horsch's assumption of three Yfijfiavalkya's is something quite new: a historical person as a ritual authority in SB I-V, a legendary expounder of the ritual in lb. XI-XIII and the foremost philosopher in the B,gU (pp. 380-391). 14 Chronologically Horsch is inclined to place the Y~jfiavalkya-parts of this Upanis.adin the sixth century B.C. (9. 391, but see also p. 474 end with n. 1). 15 In conclusion Horsch states that behind the legendary features of the great mystic Yajfiavalkya there stand out always three aspects in marked relief: lirsfly his relation to the ks.atriya Janaka, secondly to the east of India and thirdly to the philosophical gloka-tradition. All this is judged to point to the sixth-fifth century B.C. Next the often discussed nature of the Vratyas and the question of the Vr~tyastomas are once again treated in view of demonstrating a close relation between the Vratyas and the G-tradition (pp. 401-420). Similarly connected with the G-literature as well as with the ks.atriya-milieu are the bards known by various names: gdth~kdra, vi.ndgdthin, sara, mdgadha, vaidehaka, ku~flava, etc. (pp. 420-427). Lastly the ks.atriya-problem (pp. 427-448, cf. also p. 479 f.), one of Horsch's chief themes, is collectively set forth. A particular stress is laid on the contribution of the k.satriya-class to the development of the mental and cultural life of ancient India. In Horsch's opinion the contribution x3 To mention only a few older names: Oldenberg, Buddha, 9 p. 33, n. 1, Die Lehre der Upanishaden, p. 209; Deussen, Sechzig Upanishad's, 3 p. 426; el. also MacdonellKeith, Vedic Index, I1, p. 189. 14 Logically persuasive. In spite of Horsch's able exposition, however, I can not conceal that I feel still somewhat disconsolate over this solution. In a domestic scene at his parting from his beloved wives (B.gU II, 4, IV. 5) and a dramatic spectacle at the end of a great philosophical disputation in King Janaka's court (III. 9.27) Y~jfiavalkya seems to stand out too vividly to be explained away by the hypothesis o f a 'Deckname'. 1~ On Horsch's view of Y~tjfiavalkya's relation to Pgm.ini and Buddha, see pp. 391399 and cf. also p. 464 f.

REVIEWS

33

of the k.satriya-teachers to the formation of the fundamental Upani.saddoctrines should not be underrated. They were also most intimately connected with the epic as well as the speculative dlokas. The fifth or last chapter ("Entwicklungsgeschichte", pp. 449-482) sums up a history of the G-literature in a broad sense, that is, a history of the gath~s from the IndoIranian period, in which they simply meant religious poetry in general, down to the time when, as the fountainhead of the sm.rti-works (Epics, Purdnas on one side, on the other metrical Upanicads and Dharrnagdstras), they furnished Hinduism with its foundation (p. 449). In the first place the transition from the stanzas quoted in various texts of the Hindu literature (including the younger Upani.sads and Kalpastitras) to the compilation of metrical works is historically traced (pp. 449-455). Next the form of the introductory formulae and the use of the term gdth8 or ~loka are evaluated for textcritical purpose (pp. 455-458). By means of a concordance of verses occurring in different texts it is shown that the remarkable independence of the Vedic schools in this matter proves at the same time the independence of the G-tradition (pp. 458-461). The sixth century B.C., being a critical point in the history of Indian thought, is also important for the chronological study of the G-literature. From this viewpoint the relation between Buddhism and flramanism in general proves a difficult problem (pp. 462-467, 473-478). Out of many I have chosen here only one example worthy of special mention: both the dtmavdda of the Upanicads and the anattavada of Buddhism are derived from a common source closely connected with the G-tradition (p. 463 f.). After a sketch of the post-Vedic appraisal of the VGS (pp. 468-470) Horsch discusses the elusive problem of their chronology (pp. 471-473). I would here retain his view that the period of the compilation of the SB should tentatively be dated at 800600 B.C. Summarizing the intellectual reform which began in the seventh century B.C. and reached its climax in the second half of the following century under Buddha, Jina and others and taking into account a substantial contribution of the warrior-nobility to the transformation of ideas in the Upani.sads, Horsch boldly speaks of the 'K.satriyaRevolution' (p. 479 f.). This first revolution, however, was accomplished in a still limited circle. 16 There followed, Horsch maintains, a second, so to speak, 'democratic' revolution which, from the literary point of view, resulted in the production from the fifth century B.C. onward, of the epic texts for Hinduism and the Jfitaka-stories for Buddhism. In a sense this is just a renewal of the old G-tradition and the Vedic akhyana was transformed into the monumental epics (p. 481 f.). And Horsch's monumental work ends with the following words: "In diesem Sinne bildet die G~th~Literatur das Durchganstor vom Vedismus zum Hinduismus, sie dient als Schlfissel zum Verst~indnis der so komplexen indischen Geistesgeschichte" (p. 482). From the above synopsis I hope it has become quite clear that on account of its important contents, remarkable assertions and rich bibliographical notes, this is a work of lasting merit, which will remain one of those books every vddisant must read. All the more it is regrettable that owing perhaps to the circumstance mentioned in his Preface, p. VIII init. some inaccurate references and misprints have been left uncorrected. As most of them, however, are of such a sort as can be easily rectified by careful readers and do not detract from the value of the work, I abstain from enumerating them. 17 Instead 16 For the process, see esp. p. 480. But it is very difficult to answer the question why just the ks.atriyas played a leading part in that age of mental fermentation, cf. p. 480f. 1T On the whole the Index of Indian Termini (p. 503ff.) suffers from omissions though the criterion of selecting items is not clear. It registers e.g. gdthaka (p. 328 in fine) but not g~th~dargin (Sa.mhUB) nor g~thakara (p. 421 init., where Pfi.nini 3, 2, 23 should be read instead of 26). It is rather strange that Horsch spells always

34

REVIEWS

I should like to comment on one passage of interest (p. 431). Relying for JB 1.285 on Caland's conjecture kegina.h sdtyakdminab (Gen.) for kegind s~tyakdrnind (Instr.) ms. (see Das JB in Auswahl, no. 100: p. 110, p. 111, n. 2; so also the Nagpur edition), Horsch speaks of a struggle between Ahinas AAvatthi and Kegin Dfirbhya after Ke~in Shtyakamin's purohitaship and suspects with Caland, op. cit., p. 112, n. 14, that the names Ke~in D~rbhya and Ke~in S~ttyak~uain were wrongly exchanged. But the manuscript-reading need not be changed and the passage means: (King) Ke~in D~rbhya drove out (i.e. dismissed) Ahinas ,~gvatthi from the purohitaship on account of (i.e. in favour of) Ke~in S~,tyak~mi (this being the only correct stem-form warranted by the tradition). It seems that of the two Brahmins who served Ke~in Dfirbhya's father (JB ILl00: Caland op. cit., no. 133) the younger one Kew S~tyak~mi alone won the new king's favour, whence Ahinas ,~vatthi's lamentation, a8 Tokyo

N. Tsuj

C. G. Hartman, Emphasizing and Connecting Particles in the Thirteen Principal Upanishads ( = Annales Scientiarum Fennicae, series B, No. 143, 2). Helsinki, Suomalainen Tiedeakatemia, 1966. 180 pp. If one thinks of the never-failing interest aroused by the Upanishads and the voluminous literature published on their philosophy, it seems surprising that so little work has been done on the language of these texts. Despite the great advances made over the last few decades in the field of Vedic and Br~hmana language studies there has been no major work specifically devoted to the language of the Upanishads since A. Fiirst, Bet Sprachgebrauch der iilteren Upanis.ads (G6ttingen, 1916). The present work deals with one of the most striking features of the B r a h m ~ a s and Upanishads, the use of emphasising and connecting particles, and it is therefore welcome. The work begins with a survey of the definitions of the term 'particle' as given in the standard Sanskrit grammars and dictionaries. It seems a pity that no recent general linguistic literature is referred to, either here or elsewhere in the book - such references would have given weight to Hartman's own conclusions regarding the lack of independence of particles. The first major section of the work gives a detailed account of the use of the thirteen particles api, u, uta, eva, kila, khahl, ca, nu, vdva, vai, sma, svid and ha in the thirteen principal Upanishads, that is the Upanishads which figure in Hume's translation, though the order in which they are examined is that adhered to by Radhakrishnan

(B!'had-dra.nyaka, Ch~ndogya, Aitareya, Taittirfya, ida, Kena, Katha, Pragna, Mun.daka, Md.nd.~tkya, Svetdgvatara, Kaugftakf Br6hma.na Upanis.ad, Maitri). Hartman adheres mainly to the text of Radhakrishnan's edition and translation. He refers also to Deussen, Hume, R6er, Senart and Renou, but van Buitenen's edition of the Maitraya.nfya Upanis.'ad (The Hague, 1962) does not seem to have been available to him. The particles examined by Hartman are both emphatic and connecting and form a distinct group; they are certainly the most frequent and charaeteristic in the Brahman.as and Upanishads, but there are other important particles with claims for inclusion in this group, particularly hi. H a r t m a n makes a complete inventory of the passages in which the thirteen particles occur, classifying their use by their significance and their (except p. 491, line 2) the famous Indologist's name S. n. 1, where his well-known book Le th~dltre indien whereas the same work is mentioned under E. Senart x8 Cf. N. Tsuji, " O n and around Ke~in D~rbhya" (Tokyo, 1966), p. 134f.

L6vi as L6vy, see e.g., p. 310, (Paris, 1890) is referred to, in the Bibliography (p. 492). (in Japanese), Fs. Kanakura

REVIEWS

35

position in the sentence. This represents a very thorough and sound piece of research, but there are some minor points which seem debatable. The use of the term 'epithetic' (pp. 21-24) in the sense of 'following' is not altogether satisfactory. Some interpretations are open to doubt, e.g.p. 23: "In a few cases yady api is not used concessively ('even if', 'even though') but is simply translated by a conditional 'if'. Chhndogya V.24.4 tasm6d u haivan.~vid yady api ea.nd.6layocehi.st.am prayacchet ~tmani haiv6sya tad vai~v6nare hutam sy~d." Hartman is here guided by Radhakrishnan's translation "Therefore if anyone who knows this should offer the remnant of his food to a Ca.nd.~da, it would be offered in his Universal Self." It is quite certain that api here has a very strong concessive meaning and should certainly have been translated by 'even if' "even if he gave ... to a Ca.n.d~la (the lowest form of humanity)". This is corroborated by Senart's translation "quand m~me il donnerait h t m ca .n.d~la les restes du sacrifice..." In the quotation (p. 29) from the Taittiriya Upani.sad II.6.1. the particle u would hardly seem to "imply alternatives: either ... or". The alternative is implied by the particle uta which occurs in this sentence. The occasional references to the later history of the particles are so vague as to be worthless, e.g.p. 25 (api): "The particle is often found in various kinds of P~li literature", and p. 54 "'ca is frequently found in various kinds of Phli literature". But these and other objections are in fact very minor, they do not alter the main conclusions regarding the use of the particles in the Upanishads. The second main section lists all the combinations of particles and analyses their occurrences, while the third section deals with the general discussion of particles and their use in Upanishadic style. In this section Hartman analyses the thirteen principal Upanishads paragraph by paragraph, demonstrating how the philosophical (or speculative, or descriptive) sections of the Upanishads contain more particles than the historical and metrical sections. The conclusions are of particular interest, showing how api, uta, kiln, khaht, vava and svid occur exclusively in the philosophical parts, while ha is most usual in historical narrative. Perhaps not enough attention is paid here to dialogue. The repeated use of particles, emphasising and calling for attention, gives life to dialogue in the Brahma.has and Upanishads - the particles also give rhythmic balance to the clause. It is therefore not altogether accurate to regard them as punctuation marks (p. 168): "Generally speaking ha and u serve as comma, semicolon or full stop, nu and svid as question mark..." Hartman's work represents a conservative and scholarly study of an interesting aspect of the language of the Upanishads. It will be useful for reference and lays a foundation for future historical and comparative studies on the use of emphasising and connecting particles. Melbourne

L.A. Schwarzschild

Debi Prasanna Pattanayak, A Controlled Historical Reconstruction of Oriya, Assamese, BengaR and Hindi ( = Janua Linguarum, Series Praetica, XXXI). The Hague, Mouton & Co., 1966. 90 pp. Fortunately the 'unbridgeable gulf' that some linguists see between descriptive and diaehronic linguistics has not become very significant as far as the study of Indian languages is concerned: traditional historical language studies have been refreshed by the modern advances made in descriptive linguistics, while descriptive studies have gained in depth from the older historical linguistics. The present work illustrates this. It is a diachronic study based not on the evidence o f older texts in the languages concerned, out on the comparison of modern synchronic data. The author very ably follows the methods of phonemic comparison initiated by G. H. Fairbanks.

36

REVIEWS

The work begins with a phoneme inventory of each of the languages studied. In the case of Oriya, the standard coastal dialect is taken as a norm, for Assamese the language of the Nowgong district (eastern Assamese), for Bengali the standard colloquial Bengali of the Calcutta area, and for Hindi standard colloquial Hindi. The author maintains a skifful compromise between the requisites of pure descriptivism and the requisites of historical analysis: he makes it clear for instance that the phonemic status of r in Hindi is in fact due to borrowed items which have created a contrast between intervocalic .d and r. Some indication of consonant clusters is given for each language. It would have created certain difficulties, but it would also have b o a of great interest, if either this section or the section describing correspondences in consonant clusters, had indicated the basic possibilities of clustering in each language, and what clusters are in fact found only in borrowed words: this would have emphasised the main differences between the four languages in phoneme occurrence and arrangement. The main part of the work consists of an analysis, mainly in tabular form, of the phonemic correspondences between the four languages in 545 items of vocabulary (a representative list drawn up mainly by G. H. Fairbanks). Lists of correspondences are drawn up for each phoneme and also for consonant clusters. A few of the words are irrelevant or only partially usable, as shown by Pattanayak: this is due to the absence of a cognate word in one or other of the languages being studied, and to borrowings. From the tables of correspondences the author deduces the protophonemes of 'Proto-Colloquial OABH', the language from which modern Oriya, Assamese, Bengali and Hindi are derived, and he lists the phonological changes that have taken place in each of the modern languages. He is in fact reconstructing the sound-system of a hypothetical proto-language contemporaneous with many Prakrit and Apabhra.mga texts. This hypothetical 'Proto-Colloquial OABH' is very much like the hypothetical Proto-Romance reconstructed by Romance linguists, and contemporaneous with 'vulgar' and late Latin texts. The lists of phonemic correspondences in OABH leave no doubt as to the basic accuracy of the main deductions. In a few cases the evidence ofthe corpus of 545 words is naturally too slim to allow any firm conclusions. This is noticeable for instance in the discussion of the initial retroflex plosives. Pattanayak's conclusions are sound: "Consideration of pattern makes us suspect Hindi dentals as correspondences to retroflex consonants in Oriya and Bengali" (p. 25). But the phonemic evidence from one Hindi informant over a corpus of 545 words cannot give any idea of the true complexity of the developments, and cannot give any inkling of the fact that the initial retroflex in Hindi conveys a certain amount of emphasis and has 'prosodic' significance, as has been shown by A. Master, "Initial Cerebrals in Indoaryan', P. If. Gode Commemoration Volume, p. 10. The lists of correspondences and their analysis is followed by a table which summarises the development of the proto-phonemes in the four languages under study. F r o m the results of the correspondences the author is also able to give a statement of relative chronology. He illustrates this statement by drawing a family-tree, and he shows how certain changes, such as vowel assimilation and the merging of short and long i and short and long u brought about the separation of Proto-OAB from the Hindi branch of the original OABH. Subsequently further sound-changes brought about the separation of Oriya from Assamese-Bengali. These deductions are of great interest and put the whole of the comparative study of these languages onto a more rigorous footing. This family tree is excellent and clear to follow at a glance, but naturally cannot account for overlapping innovations and non-definite splits. It would have been particularly interesting if Pattanayak had indicated some of this complexity in further diagrams, as has been done so brilliantly for Panjabi, Hindi, Bengali and Marathi by F. Southworth, "Family-tree diagrams", Language, XL (1964), p. 562. This would have thrown more emphasis on the true intricacy of language development.

REVIEWS

37

The Controlled Historical Reconstruction of Oriya, Assamese, Bengali and Hindi is a most valuable contribution to comparative Indo-Aryan Linguistics: it sheds a new light on the comparative phonology of the Central and Eastern Indo-Aryan language groups, a field that has received little attention since the great works of Grierson and S. K. Chatterji. One would like to see similar studies in the field of morphology and syntax. The author mentions in his Introduction the possibility of ultimately "demonstrating the relation of the Prakrits and the Proto-Language in a neat and scientific manner". Such a study would form an important sequel to the present work - - the relations with the literary Prakrits would help the reader to visualise more clearly the historical and geographical background and the position of the as yet tabular Proto-Colloquial O A B H within the general development of Indo-Aryan. Melbourne

L . A . Schwarzschild

Jan Gonda, A Concise Elementary Grammar of the Sanskrit Language, with Exercises, Reading Selections, and a Glossary, translated from the German by G o r d o n B. Ford Jr. Leiden, Brill, 1966. VII + 152 pp. This translation of Gonda's Karze Elementar-Grammatik is based upon the fourth edition (1963) of the G e r m a n version, in which the inaccuracies of the former editions with regard to the sandhi have been corrected. It is to be welcomed that this book, now perhaps the most widely used in introductory courses, has also been made accessible to students in the United States and Canada. The grammar may be supposed to be too well known to need any comment. The translation has been done carefully, although it is at times somewhat mechanical. Thus on p. 93 1.1, where G o n d a had explained e~o "naya krfd.ann dsta by the English translation "he was sporting with her", the indication "Eng." before the translation is redundant in this version. Slight mistakes of the German version have been slavishly copied, e.g., p. 52 1.20 "3rd pl.impf." for "3rd pl. pres. impf."; p. 94 1.17 pithikdm for pit.hik6m; p. 99 1.9 from the bottom "pres. act. ind." for "pres. mid. ind."; p. 119 1.15 yen~syabhihata iti for yendsy abhihata iti; p. 125 1.12 sydmo for sydma; p. 134, where gd- has been inserted in the fourth edition but gd- + upa-~- (cf. updg~t, p. 107 1.2), which could easily have been inserted in this entirely new book, is lacking; p. 149, second column, where the words from saYhvatsara- to sarhh.rr appear after sathmdrjat instead of before sarhkdga- (as is the normal alphabetical order which is followed in the case of pdthsu- p. 140, mdrhsa- p. 143). Understandably some new printer's errors have crept in, e.g., p. 99 1.9 ksal- for k.sal-, p. 117 1.18 -locandc for -locand~, 1.19 srayatdm for grayatdm. So much for the translation. I take this opportunity to add a few remarks which may deserve consideration in view of later editions. In this connection it may be remarked that the habit (dating from Pischel) of taking illustrative sentences from poetry, however attractive they may be, often necessitates some comment when the sentences are given as mere prose. Thus for kuto gata bhavanta.h (p. 101 1. 18f.) kuta dgatd bhavanta.h would seem more idiomatic (cf. also p. 16 1. 14 devo gacchati which is "a god is going" rather than "coming"). Similarly the next sentence (p. 101 1. 19) ajagamapunar vedma Sdvitrfsaha mantribhi.h has a wordorder which would be uncommon in prose; cf. punar-dgacchati, punardgamanam). However, the exercises certainly serve the immediate purpose of grammatical drill, and it can be maintained that considerations of idiom must be reserved till later. P. 101 1. 8 narab sumand dvi~am apagyat: the meaning "intelligent" which is given for sumanas- in the vocabulary, is correct for p. 102 1. 1, but here the word is apparently used in the commoner meaning "benevolent".

38

REVIEWS

P. 101 1. 13 from the bottom: pf~rvatit hi sakha me "si sarhbandhi ca, a line taken from the Nala episode (Mhbh. III. 76.14 crit. ed.) shows the use ofparvam with a present (like ha smapur~ in the older language), which should have been noted in 1.5 from the bottom and on p. 141 s.v. pftrvam. P. 113 1. 6janucalanayogya-: is calana- here "movement, creeping" (p. 135)? The combination with yogya- rather points to the meaning "rocking, swaying" (Mort. Will.), "auf den Knieen schaukeln" (Pet. Dict.). ; of. l. 9janupracalanaparo. P. 113 1.4 from the bottom: in bhaye v~ yadi v~ hars.e satilprapte yo vimargayet we have rather the common construction with a second v6 being replaced, on account of its enclitic character, by yadi vd (ef. Latin s~'-ve) if it is to precede the noun. Cf., e.g., Speyer, Sanskrit Syntax, para 428 Remark. P. 115 1.8: "Never give up hope, yath6 samudre "pi ca potabha~ge / s~thyatriko v6~chati tartum eva." What does tartum mean here? G o n d a paraphrases it as "he does not remain quietly at home". At first sight, he seems to have the strong support of BiShtlingk, who in the first edition of his lndische Spriiche, Nr. 1072 rendered this by "wiinscht ... gerade wieder zur See zu gehen." Since, however, the last two lines are obviously meant as an illustration of the general sentence dhairy6t kad6cid gatim apnuy6t sa.h (i.e., who does not despair may be saved), I take the last line to mean: "just as a seafarer who suffers shipwreck, even though he is in mid-ocean, desires to save his life" (t.r- "gliicklich davonkommen, sich retten", Pet. Diet.). This is also more or less B6htlingk's translation in the 2nd edition (Nr. 2636): "Wiinscht doch ein Seefahrer, werm ihm auch im Meere ein Schiffbrueh widerfahren ist, seinen Bestimmungsort zu erreichen". P. 115 1.5: nayatat.~ (also in the German version of the ~ a m m a r ) is a printer's error for nayathatz that has escaped notice. P. 124 1.13: note the plural for the dual (which the Sanskrit commentator explained otherwise: bahuvacanarh parivardpek.saya). P. 125 1.1: the meaning "old, dilapidated", which the vocabtdary gives forjfr.na- is correct for p. 106 1.4 but here it means of course "digested". Leiden

F . B . J . Kuiper

Tile Begh~nings of Indian Philosophy. Selections fi'om the Rig Veda, Atharva Veda, Upani~ads, and Mah6bhdrata. Translated from the Sanskrit with an Introduction, Notes and Glossarial Index by Franklin Edgerton ( = Unesco Collection of Representative Works - Indian Series). Cambridge, Mass., Harvard University Press, 1965. 362 pp. in-8 ~ After an introduction of thirty-two pages giving a survey of the doctrines included in the texts, translations are presented, partly or in full, of fifteen hymns of the t~gveda and sixteen of the Atharvaveda, of Books III and IV of the B!'had6ra.nyaka, of Chapter VI of the Ch6ndogya Upani.sad, of the whole of the Kat.ha Upani.sad, and of selections from the Bhagavadgit~ as well as from the lVlok.sadharma. The selections from Vedic scriptures comprise material of great philosophical and historical relevance. The book is described in the translator's preface as "a kind of scholarly last will and testament". Readers will certainly consider with great attention and respect the great philologist's bequest. The translation is sober, and where the translator deviates from predecessors his option always seems at least reasonable or worthy of discussion, if not an improvement. The strength of his philology seems to lie in finding out meanings of words and sentences. In a few cases, however, where the text calls for an interpretation from a wider context of thought, Edgerton seems to remain all too dependent on achievements of

REVIEWS

39

earlier scholars, even such as were inferior to him in their philological ability. I will note a few examples of dependence on questionable interpretations put forward by Paul Deussen. Brh. 4.3.10 mdtrd is translated by 'material', B!4t. 4.5.15 and BhG. 2.14 by 'matter'. Now 'matter' is a highly abstract and, in the sense meant here, even a philosophical term. If mdtrd ever had this meaning, it would be astonishing that philosophers did not take advantage of this philosophy already preformed in the language. I think that in all the three passages the word has the meaning, frequently occurring, of 'a particle'. In BhG. 2.14 it is certainly not easy to understand as is borne out by the desperate attempts of commentators to interpret it in a sense fitting in with their own philosophy. Perhaps the 'particles' meant here are particles of the five elements. In the tat-tvam-asi text of Ch6nd. VI, Deussen translated tat satyam by 'das ist das Reale'. Edgerton follows him, saying: 'that is the Real'. However, it seems more than doubtful that satya here should mean 'real'. In the later Vedanta satya certainly is a term to denote 'the truly Real' as opposed to what is ontologically 'false' (mithy6, anirvacaniya). But this concept cannot be assumed to have existed as early as the Chandogya Upanis.ad. The tat-tvam-asi text itself (section 16) shows that at the time of its composition the old belief in Truth as a magical power was still alive. What the whole of the Sixth Chapter of this Upani.sad intends to teach is the doctrine of the Existent (sat) which as an imperceptibly subtle but all-powerful entity is operative in every thing and in every person. Now this doctrine is reconciled with earlier doctrines of omnipotent and all-pervading entities by identification of the central concept with the Self (dtman), which had been the object of many upani.sadic speculations, and with Truth (satya), which, originally under the name of.rta, had from the earliest times been respected as a magical and cosmic power. Such identification - the oldest form of what I have called 'inclusivism' when speaking of later Hinduism - legitimatizes the new doctrine and recommends it to the sympathy of advocates of the earlier beliefs and speculations. Truth and the Self - this under the name ofjfi,a dtmd - are moreover specially demonstrated to be identical with the Existent in sections 15 and 11, respectively, whereas the doctrine of the Existent being the essence of the universe (aitaddtmyam idaJ.n sarvam) was expounded in the preceding sections 2-4 and the doctrine of its being the essence of man (tat tram asi) in sections 5-8. So all the constituents of the refrain aitaddtmyam ida~.n sarvan.7 tat sat)'a~?~ sa dtmd tat tvam asi are treated specially in the chapter, and as satya evidently means truth in section 15, it cannot but have the same meaning in the refrain. At Chand. 6.11 Edgerton translates jiva by 'the life' (Deussen: 'das Leben'). But fiva is an adjective as is borne out, e.g., by the expression jivd 6tm6 occurring in the same section. When a noun is lacking,jiva is to be translated by 'the Living One', which after all is a notion somewhat different from 'the life'. Again following the lead of Paul Deussen, Edgerton believes that what I have called 'the ethics of tat tvam asi' is expressed in BhG. 13.28. For a criticism of this view, which was first conceived by Schopenhauer and presupposes his philosophy, see my article "Schopenhauer and die Ethik des Hinduismus", in Saeculum, Jahrbuch fiir Universal geschichte, vol. 12 (1961), pp. 366-399. Edgerton writes in his preface : "The majority of the Sanskrit texts here translated have been the object of intensive study by me for decades, and have been worked over again and again in seminars with advanced Sanskrit pupils ..." It is astonishing that such repeated work should never have led the great scholar to question the unity of texts like the Ka.tha Upani.sad and the Bhagavadgitd. "'The Gita seems a unit", he states laconically (p. 197). But I think that at least the fact that one and the same subject is sometimes found treated in the Gird at different places in a different manner makes it impossible to maintain the unity of the poem. If Garbe's and Rudolf Otto's attempts do not seem convincing, the problem cannot be brushed aside by acquiescing

40

REVIEWS

in uncritical views. I think it is time that 'higher criticism' should be undertaken in treating texts like the Upanisads and the Bhagavadgfta (cp. my article "Zur Methode der geschichtlichen Erforschung der anonymen Sanskritliteratur des Hinduismus", in Zeitschrift der Deutschen Morgenliind. Gesellsch., vol. 11, 1962, pp. 483-492). Edgerton is certainly right in stating, contrary to the opinion of H. Oldenberg: "The Upani.sads ... seek to know the real truth about the universe, not for its own sake; not for the mere joy of knowledge; not as an abstract speculation; but simply because they conceive such knowledge as a short-cut to the control of every cosmic power" (p. 28). Knowledge 'for its own sake' has never been an Indian ideal, not even in later times. In speaking of the philosophical passages of the Epic, Edgerton remarks: "Like the Upani.sads, they are still tentative, fluid and unstable, and often inconsistent with themselves and each other. But they mark another step, if not a very long one, in the direction of the systematic codifications of classical Hindu philosophical systems'" (p. 35). However, it seems that after Frauwallner's investigations the theory, established by Deussen, of a 'transition philosophy' (Obergangsphilosophie) in the Epic is no longer tenable (nor can the other theory of a 'mixed philosophy', Mischphilosophie, be maintained now with any claim to probability). Rather, the problem is one of literary genus. Almost ~11 primary literature from the time when the S~flkhya was a living philosophy has been lost; some fragments can however be recovered from the works of Buddhist arid Jain authors. In all probability the philosophical passages, which after all belong to the latest strata of the Epic, as well as some later Upani.sads were composed during the time when Sfifikhya schools were flourishing. They are not preparatory to, but reflective of, what may be called the 'classical' period of the Sfiflkhya system. This must be kept in mind when these passages are utilized for clarifying the history of Indian philosophy. They are not primary sources for our knowledge of the Sfifikhya system. They include ideas of the philosophy current at their time - as the Pur~.nas do at later times. They may even have preserved some small texts that were used in schools for instruction. But their outlook is not the same as that of the philosophical schools. This is the reason why, e.g., the word s~likhya has a peculiar meaning in the Epic. If the word becomes in the Epic a "term to describe the method of gaining salvation by knowledge" (Edgerton, p. 36), this seems to be due to its reception into the literary genus of the Epic, which included popular philosophy and practical wisdom. Therefore, this use of the word is no more an indication of an earlier meaning than the 'fluid and unstable' character of the 'philosophy of the Epic' points to an earlier stage of development of the S~flkhya system. Rather, the word has been accommodated to the intellectual atmosphere and intentions of the authors of the Epic, at a time when the S~flkhya system had already reached a high stage of development and was being taught in flourishing schools. A reflection of contemporary philosophy is perceptible in the Epic, which in this respect does not differ from the Pura.nas, in which reflections of philosophies current after the decline of the Sfiflkhya can be traced. Miinster (Westf.)

Paul Hacker

C. Hooykaas, Sf~rya-Sevana, The Way to God of a Balinese ~iva Priest ( = Verhandelingen der Koninklijke Nederlandse Akademie van Wetenschappen, Afd. Letterkunde, Nieuwe Reeks, Deel LXXII, No 3). Amsterdam, Noord-Hollandse Uitgevers Maatschappij, 1966. 224 pages, 33 plates. In 1964 Dr. Hooykaas, of the London School of Oriental and African Studies, pub-

REVIEWS

41

lished a book called Agama Tirtha, Five Studies in Hindu-Balinese Religion. It has been reviewed in the Indo-Iranian Journal, volume X, no 1 (1967). The book which is under review now is a sequel to the other one, and it will probably be followed by some more volumes dealing with the same subject: religion of Bali. S~'ya-Sevana, Sun Worship, is the name of the principal rite of the pgdandas, the (mostly) Shivaite brahmin priests of Bali. It is, in a simple form, performed daily in the priestly manses, called griyas (Sanskrit g.rha). More elaborate rites of the same character belong to the periodical festivals celebrated in the numerous Balinese temples. Dancing, music, theatricals and offerings are the most prominent features of the temple worship, however. Cremation ceremonies occupy an important place in the social and religious life of the Balinese. P~dandas performing rites of the S~rya-Sevana type are usually asked to grace these ceremonies with their presence, but the other features, such as gorgeous pageants and merrymaking, attract far more attention of the public attending a cremation than the mysterious rites of the lonely bralarnin priest. The essence of the Srarya-Sevana rite is the making of Holy Water. The priest, sitting on a mat in a small open pavilion, says Sanskrit mantras, prayers and hymns, makes sacral gestures (mudras) and handles some small brass implements. First he purifies himself, then he induces the God, Shiva-Sun, to enter through his head into his heart and further into the cup of water, which thereby is hallowed. Finally ShivaSun is prayed to return to His divine abode, leaving the priest's body as it was before. At the end of the ceremony the priest sips a little of the Holy Water, which is further used to purify places, food and persons by sprinkling. The importance of this rite in Balinese Hinduism is apparent from the fact that the religion is commonly called /~gama Tirtha, Holy Water religion. Dr. Hooykaas has treated the subject very thoroughly and methodically. The philological study of the numerous palmleaf manuscripts containing notes on the Sfirya-Sevana mantras and ~lokas occupies many pages. They contain tables and lists of variae lectiones of Sanskrit and Javanese-Balinese texts, provided with English translations. The second chapter (pp. 44-129) and the Appendices (p.p. 158-186) are mainly philological studies. The introduction (chapter I, pp. 9-42) contains notes on the principal elements of the rites: Sanskrit mantras, hymns and mudr,Ls. The last paragraph of the Introduction (pp. 35-42) contains a "Summary of SQrya-Sevana" which is a description of the actions and words of the priest. Beside these descriptive chapters, which are divided into numerous paragraphs the book contains two chapters dealing with the religious significance of the rite Chapter III (pp. 131-140) contains "A Balinese Priest's Detailed Comment", with an English translation. It is, of course, interesting to learn what a Balinese scholar offers as explanation of the numerous actions, mantras and hymns of the rite. Viewing the subject from a broad standpoint, Dr. I-Iooykaas has offered in the fourth chapter (pp. 141-156) a "Comparison of Saiva Ritual in South India and Bali." He has found four books containing descriptions of Shivaitic rites performed by priests in South India: Instrument and Parpose, Studies on Rites and Ritual in South lndia by Carl Gustav Diehl (Lund, 1956), "Das P'ajhvidhinirfipa.ha des Trimalla" by Fausta Nowotny (lndo-lranian Journal, I, 1957), Rauravagama, Volume I by N. R. Bhatt (Pablications de l'Institut Francais d'Indologie, no 18, Pondich6ry) (Paris, Maisonneuve, 1961) and Le Rituel quotidien givaite de l'Inde du Sud selon Somasambhu by H616ne BrunnerLachaux (same Pablieations, no 25) (1963). These books have enabled him to compare the Balinese rites with the Indian ones. Moreover Dr. Hooykaas's S~rya-Sevana contains, spread throughout the book, two valuable contributions towards the knowledge of Balinese religion. The first one is the collection of drawings of Balinese priests performing their rites which were made by the Swedish artist Tyra de Kleen in cooperation with Mr. P. de Kat Angelino, a

42

REVIEWS

Dutch magistrate residing in South Bali in the early twenties. The drawings were first published in a Dutch book, Moedra's op Bali, Handhoudingen der Priesters (The Hague, 1922). German and English editions appeared in 1923 (Folkwang Verlag, Hagen) and 1924 (Kegan Paul, Trench, Trubner & Co., London, and E. P. Dutton & Co., New York). Another, smaller collection of drawings of mudras, made recently by a Balinese artist, has been added to the artistical work by Tyra de Kleen. The second contribution towards the study of religion in Bali, incorporated in S~rya-Sevana, is the complete English translation of all Sanskrit and Balinese texts and words. Moreover an "Index to .~gama Tirtha and Sfirya-Sevana" has been appended (pp. 192-224). The Index contains a glossary of Javanese-Balinese, Balinese and Sanskrit words found in the texts, with references to the pages of the two books where explanations are provided. Dr. Hooykaas was able to supplement the "Glossary of Sanskrit from Indonesia" by J.Ensink and J.A.B.van Buitenen (in VAK, no 6~ December 1964, Deccan College, Poona) by some newly found words. Old Javanese, Javanese-Balinese and Balinese lexicography is also enriched by the references to religious texts. The combination of the Index of Names (Balinese, Dutch and English) with the Glossary of Words (Balinese, Javanese, Sanskrit and English) is very practical. Two or more separate indexes would have meant time wasted for the interested reader looking for information. Compared with the book under review, Dr. Hooykaas's previous book, ,~gama Tirtha, Five Studies in Hindu-Ba6nese Religion, may contain more varied readingmatter on different subjects which are interesting for anthropologists in general. Surya-Sevana, however, is an exhaustive study of one subject, the most important rite of high-class Balinese priests. The thorough philological studies of the mantras and prayers in the Introduction, the second chapter and the Appendices are remarkable scholarly achievements. They constitute a complete Balinese prayer book. Published separately, these sections of SOrya-Sevana would form a useful textbook and manual for Balinese students of religion and aspirants for the priesthood. Such a book does not yet exist in Bali. In the Preface (p. 5) Dr. Hooykaas mentions the appearance, in 1966, of a Balinese edition of a part of Sylvain L6vi's "Sanskrit Texts from Bali" of 1933. Dr. Hooykaas's work contains much more material than the older book, and it is wholly based on the best Balinese information available. It would certainly appeal to religiously minded Balinese scholars. The appearance of a Balinese prayer book compiled by a Dutch orientalist would be a noteworthy instance of friendly cooperation of scholars belonging to different cultures. However, the perusal of those sections of Sftrya-Sevana which contain the Sanskrit mantras makes great demands upon the perseverance of non-Balinese readers. Dr. Hooykaas has been aware of this fact, and therefore he has inserted the "Summary of Sftrya-Sevana" (pp. 35-42). This paragraph contains most of the material which is interesting for non-Balinese students of Balinese religion. The fourth chapter, "Comparison of Saiva Ritual in South India and Bali", is the most interesting one, seen from a general point of view. Dr. Hooykaas's conclusions, summarized in the paragraphs E and F, need not be repeated here. Suffice it to say that in both countries the "unification of the worshipper's soul with the Siva Soul" seems to be the essence of the ritual. In South India an anthropomorphic concept of the Deity is an important feature of religion. In Bali it seems to be mostly absent from the daily worship, whereas the Holy Water, which is so very important in Bali, is much less so in South India. Dr. Hooykaas's expectation (!o. 156) "that more critical editions of South Indian agama will shed light not only on the development of South Indian ritual but also on the period when the present-day Balinese ritual was borrowed" is likely. However, in the present reviewer's opinion, it seems possible that the Balinese ritual developed in its own way, diverging from its Indian prototype, during the four or five centuries

REVIEWS

43

of isolation following the conversion of Java to Islam about 1500 A.D. The priestly rites as described in S&'ya-Sevana do not satisfy the Balinese people's need for religious accompaniment of the stages of life and death. Beside the Hindu-Javanese ritual there exists a popular religiosity of indigenous, pre-Hinduistic origin. It seems probable that its manifestations, such as ancestor worship and sumptuous disposal of the dead, have been influenced by Indian ideas in the course of the centuries. There is no reason to assume, however, that the Indian rites have forever remained the same since the time when they were introduced into the Archipelago. Some differences between South Indian and Balinese rites, as they are at present, might be the result of the influence exercised by autochthonous Balinese religious concepts on the development of Hinduistic ritual in Bali. S&ya-Sevana is an excellent expos6 of a special Hinduistic aspect of Balinese religion, the Siva cult. The sub-title, The gray to God of a Balinese Siva Priest, has been added in order to place this fact into due relief. It is hoped that Dr. Hooykaas will continue his valuable series of publications on Bali. In some of his next books the autochthonous aspect of the Balinese religion might come to the foreground. Some features - exorcism, ancestor worship, trance, potlatch - have been mentioned already in the Conclusion of the present book (p. 155). The image of Balinese culture, presented so vividly in Dr. Hooykaas's publications, would be deceptive if Sarya-Sevana is not counterbalanced by descriptions of the other, popular and indigenous, aspects of the religion of the artistic inhabitants of the beautiful island. Leiden

Th. G. Th. Pigeaud

Heinrich von Stietencron, lndische Sonnenpriester. Sdmba und die Sdkadvipiya-Brdhma.na. Eine textkritische und religionsgeschichtliche Studie zum indischen Sonnenkult (=Schriftenreihe des Siidasien-lnstituts der Universitiit Heidelberg. Hrsg. von Hermann Berger u.a., Bd. 3). Wiesbaden, 1966. [Dissertation, Mfinchen 1965.] Seit den Zeiten von Wilford (1808), Windischmann (1832), Wilson (1840), Reinaud (1849), Weber (1853 ~ 7) und Aufrecht (1859) interessierte man sich ffir die Nachrichten des Bhavis.yapur~n.a, sp/iter auch ffir die des letzterem teilweise parallelen Sdmbapur&.la, iiber die Grfindung indischer Sormentempel und fiber die Berufung der sonnenanbetenden Maga-Brahmanen aus Shkadvipa nach Indien. H. yon Stietencron unternimmt es in seiner Miinchener Dissertation, diese Nachrichten neu aufzuarbeiten. Von Stietencron liefert von den ihm relevant erscheinenden Passagen beider Purh.nas zunfichst den Sanskrittext (pp. 29-120) und dann eine deutsche IJbersetzung (pp. 125214). Soweit das mtiglich ist, sind beide Versionen zu einem einzigen Text verarbeitet. Ansonsten sind sie parallel nebeneinandergestellt. In denjenigen F/illen, in denen nur eine Version vorliegt, ist unachtsamerweise tells ganzseitiger Druck (SP, pp. 67-74, 162-170), teils halbseitiger (BhP, pp. 81-120) und teils zweidrittelseitiger (BhP, pp. 175-214) verwendet, was sehr verwirrend ist. Es folgen zwei wissenschaftliche Kommentare. Im Kommentar I "Zur Geschichte der Geographie" (pp. 217-247) wird darauf hingewiesen, dab der gleiche Vorgang, die Grfindung eines Sonnertheiligtums dutch Sfimba, in verschiedenen Passagen des SP mit verschiedenen Ortlichkeiten in Verbindung gebracht wird. In einer ~ilteren Textschicht (SP 24 = BhP 1,127) wird dem S~mba die Grfindung des Heiligtums Mitravana bei Mfilasth~na/Multan zugeschrieben, in einer jiingeren Textschicht (SP 42) die des Tempels yon Tapovana/Kon. hrka in Orissa. Der Kommentar II (p. 248-272) widmet sich der Religion tier Sormenpriester, d.h. der Maga[-Brahmanen] des SP und der nut im BhP auftretenden, in Magadha zu

44

REVIE~VS

lokalisierenden Bhojaka, die teilweise ebenfalls Maga genannt werden. Die nach von Stieteneron spiitestens im 1. Jahrhundert v. Chr. eingewanderten Maga des SP sollen nichtzoroastrische Mithraisten gewesen sein (19. 277), die aber Berfihrung mit synkretistischem Zoroastrismus gehabt haben (p. 235). Die Bhojaka-Maga des BhP abet, die etwa am Anfang des 7. Jahrhunderts n.Chr, nach Magadha eingewandert sein soUen, sind, wie yon Stieteneron meint, eine zoroastrische Priestergruppe zurvanistischer Priigung (p. 243), welehe von den Parsen versehieden ist und die niehtzoroastrische Maga-Tradition adoptiert hat. Das Buch selffieBt mit einem englisehen Summary (pp. 273-282). Die Textedition ist keineswegs so kritisch, wie es der Untertitel verspricht. Ist es dem Leser nach einigen Schwierigkeiten gelungen, die Angaben fiber ihre Grundlagen auf p. 19 zu finden und zu entziffern, so stellt er fest, dab nur eine einzige Handschrift verwendet ist. Es handelt sich um die SP-Handschrift Ind.Off.Libr. 3619, die gemeinsam mit der gedruckten Ausgabe yon 1899 den SP-Text stellt. Ffir das BhP als ganzes sind iiberhaupt nut die beiden gedruekten Ausgaben yon 1897 und 1959 herangezogen, von denen die erstere nahezu unbrauehbar ist. Sie sind in bescheidener Weise ergiinzt dutch die yon Aufrecht vor fiber hundert Jahren (1859) in seinem Katalog der Bodleian Library, p. 31 ft. ver6ffenflichten Ausz6ge aus den Oxforder BhP-Handsehriften. Die Konkordanz mit diesen Ausziigen muB man sieh selbst erarbeiten: Aufrecht, p. 31 * = 1,72,4-6; p. 32 = 1,117, 8-56; p. 32 f. = 1,139,28-45; p. 33 = 1,139,70-81 und 1,140,32-50. Die Benutzung wird durch die Trelmung yon Sanskrittext (92 Seiten) und IJbersetzung (90 Seiten) zu einer aufreibenden Miihe. Dieser kann sich auch derjenige nicht entziehen, der fliegend Sanskrit liest, da die (Jbersetzung wegen ihrer wissenschaftlichen Anmerkungen nicht ganz auger Acht gelassen werden kann. Wer mit dem Buch arbeiten will und deshalb Wert auf Ubersichtlichkeit legen muB, wird abet noch an vielen anderen technischen Miingeln verzweifeln. Sie zu verhindern wiire Aufgabe der sechs Herausgeber gewesen. Der Stil der Kommentare ist versehwommen. Aussagen fiber ein und denselben Gegenstand sind in der Regel fiber das ganze Werk verteilt und in immer neuer Abwandlung wiederholt. Will man sieh an die Ansichten von yon Stietencron herantasten oder will man sich auch nut tiber einen simplen Tatbestand informieren, so mug man nicht selten die Einleitung und die Noten zur l~bersetzung, dazu die beiden Kommentare und auch noch das Summary einsehen. Man vergleiche dazu z.B. die .~al~erungen fiber den avya~ga-, den heiligen Gfirtel der Sonnenpriester, die auf acht relevante von dell f'tinfzehn im Register angegebenen Stellen des Buehes verteilt sind. In der Regel findet man sich besser zurecht und wird vollst~indiger informiert, wenn man die beiden wichtigsten Vorgiinger von von Stieteneron benutzt, n~imlieh J. Seheftelowitz, "Die Mithra-Religion der Indoskythen und ihre Beziehung zum Sauraund Mithraskult", (Act.Or., 11, 1933, pp. 293-333), und R. C. Hazra, Studies in the Upapur6t.tas I (Calcutta, 1958, pp. 29-109). Leider werden sic yon yon Stieteneron nur summarisch und negativ zitiert. Andere Gewiihrsleute werden iiberhaupt nur mit dem Namen genannt (p. 15: Mironow, Barth, Khareghat, Biswas). Albrecht Weber wird im Literaturverzeiehnis, p. 9 mit seiner Sehrift "lAber die Magavyakti des K.r.soad~sa Micra" (Monatsber. Akad. Berlin, 1878) zitiert. Auf p. 14 folgt der unklare Hinweis: "In einem Zeitraum yon mehr als 25 Jahren bescNiftigte er sich immer wieder mit dem Problem der Maga." ErwRhnt werden jedoeh nur noch auf p. 19649 seine Indischen Skizzen, deren Erscheinungsdaten (Berlin, 1857) der Leser selbst ausfmdig zu maeben hat. V6Uig vergessen ist dagegen Webers wichtige Arbeit "Ober zwei Parteisehriften zugunsten der Maga, resp. (~akadvipiya Brfihmao.a", Monatsber. Akad. (Berlin, 1880). Unter diesen Umsti~nden ist es mehr als schwierig, die eigene Leistung yon yon Stietencron festzustellen. Das meiste zerflieBt dem Leser unter den Hiinden. Einige Fortschritte linden sieh wolff in der Auseinandersetzung mit Hazra fiber die Frage, ob

REVIEWS

45

dem Autor der ~Utesten Textschicht des SP neben Mitravana mad einigen anderen alten SonnenheiligtiJmern, welche mit diesem konkurrierten, auch schon die Tempel von Tapovana und K~lapriya bekannt waren (p. 221 ft.). Es wird hier die insbesondere in der deutsehen Indologie beliebte Kritik der literarisehen Strata betrieben. Zu eigentlich historischer Kritik schreitet yon Stieteneron aber wohl nur p. 229 f. vor, wo er nachweist, dab der berfihmte Sonnentempel von Multan nicht Mitravana identisch, sondem dessen Nachfolger an anderem Platze ist. Kombiniert man Einleitung, p. 13 mit Kommentar p. 235 mad Summary, p. 278, so ergibt sich, dab von Stietencron der alten These folgt, die Magas seien sakischer Herkunft. Sie beruht auf der Ansicht, ~dkadvipa, der Ort der Herkunft der Maga, sei eine mythisierende Bezeichnmag for gakasthhna 'Sakenland' und meine in diesem Zusammenhange Sist~n. Sicher ist jedoch nut, dab es sich bei den Magas um Nordl~inder handelt, wie sich aus der Anweisung des Var/ihamihira (Brh. Sa.mh. 58,46) ergibt, das Bildnis der Sonne solle solcher Leute Tracht tragen (kurydd udicyave$am). Man kann also auch an die Kug~n denken. Nichts Entscheidendes gibt allerdings die obligate Nennung der Miinzlegende Mttpo 'Mi0ra' der Miinzen der grol3en Kug/in auf p. 233 aus. Selbst wenn von Stietencron meine eigenen friiheren Bemfihungen um die Frage, welche Aussagen die Legenden der Kug~n-Miinzen fiber die Religion der K u ~ n machen, nicht zitieren wollte, so h~itte doch wenigstens er selbst in die Tiefe des Problems gehen mfissen. Seit den Zeiten yon A. Weber ist das Material immerhin erheblich vermehrt worden. Nfitzlich fiir den konkreten Nachweis eines iranisch-indisehen Synkretismus w/ire es wohl gewesen, wenn sich von Stietencron mit den inzwischen leicht zug/inglichen ku~Lno-sassanidischen Miinzen beseh/iftigt h/itte. Ieh denke insbesondere an die Bezeichnung des Siva als baktr. Bop~ooctvSo tct~ctfioauf kughno-sassanidischen Schiisselmfinzen des Kughn-S~h Hormizd parallel zur Bezeiehnung des Ahura Mazd/i bzw. des Feuers als phi. bwrz'wndy yzdty = burzavand~ yazdg auf sassanoiden Miinzen des gleichen Herrschers, von mir erw~mt in Baktrische Sprachdenkmiiler, I, 52 mater NumKS 9. Daran, dab die Maga des SP eine iranische Religionstradition fortsetzen, hege ich keinerlei Zweifel. Die yon von Stietencron nieht gebiihrend gewordigten Naehriehten des A1-Birfmi (India, transl. Sachau, I, p. 20) mad des Var/ihamihira (B.rh.Sa.mh. 60,19) fiber die Maga-Priester sprechen ja eine deutliche Sprache. Die Frage, ob sie wirklich vorzarathustrische Mithraisten waren, wird man allerdings noehmals durchdenken miissen. FOr ganz unglaublich halte ich aber die These, die Bhojaka-Maga des BhP seien Zoroastrier gewesen. Zuzugeben ist nat0,rlich, dab der im BhP wie auch im SP belegte avyat~ga- 'heiliger Gortel' auf aw. aifliid~hana- zurfickgeht und somit iranischer, vielleicht sogar zoroastrischer Herkunft ist. Absolut unzoroastrisch ist jedoch die HersteUung des avyanga- aus Schlangenhaut (BhP 1,140,25). Einem jeden Zoroastrier miil3te sie als ganz besonders schauderhaftes Vergehen gegen Gott und als schreekliche Sch/idigmag yon Leib mad Seele gelten. Es kann also sehwerlieh ein Zweifel daran bestehen, daft in die Nachriehten des BhP fiber die Bhojaka-Maga ganz verschiedene Dinge eingeflossen sind. Das ist auch daran zu erkennen, dab sie einmal mit geflochtenem Haar (BhP 1,139,58) mad andere Male mit geschorenen K6pfen vorgefiihrt werden (BhP 1,117,56; 146,17; 147,12). Keine der beiden Praktiken weist auf zoroastrische Tradition. Das Scheren des Haupthaares gemeinsam mit manchen anderen von yon Stietencron p. 270 f. genarmten Eigentornlichkeiten laBt sogar vielmehr an buddhistisehen EinfluB denken. Unter Berfieksichtigung gerade dieses Umstandes seheint es mir recht klar, dab die in Magadha beheimateten sonnenanbetenden Bhojaka-Maga des BhP die Maga-Tradition des SP nicht zuletzt auch um des etymologiscben Anschlusses von Magadha an Maga willen iibernommen haben. Dieser kommt aueh in der Etymologie maga.m dhydyanti te yasmdt tena te magadhd.h sm.rtd.h BhP 1,117,53 zum Ausdruck.

46

REVIEWS

Trotz all dem steht der negativen Beurteilung des Quellenwertes des BhP zu der hier diskutierten Frage noch eine wesentliche Schwierigkeit gegenfiber. Im SP, dem wit alte Tradition zuzuerkennen bereit sind, finden sich iranische Reminiszenzen nur in bescheidenem MaBe, wahrend das aus dem SP entleihende BhP diese Reminiszenzen erheblich vermehrt. Neben dem auch durch Brh.Sa.mh. 58,47 bezeugten und oben schon genannten avya~ga- erscheinen im SP auch homa- = aw. haoma- und parahoma= aw. para. haoma- und schlieBlich die Nachricht von den vier Standen Maga-, Magyaka, Mdnasa und Mandaga. Das BhP bietet zusatzlich z.B. rardman- = aw. barasman-, patiddna- = aw. paitLddna-, acas.n.a- = rap. yazign. Neben dem Namen Jaradastra- = aw. ZaraOugtra- liefert es auBerdem noch eine Reihe yon anderen Zoroastrismen, wie die Einteilung des Tages in fiinf Abschnitte oder die Sitte, die Mahlzeiten schweigend einzunehmen. Paradox w~ire es, diese textgeschichtlich jfingeren trod sprachlich teilweise sehr verderbten Erweiterungen des BhP mit yon Stietencron als uralte zoroastrische lJberlieferung zu erklaren. Im Zusammenhang damit ist die Fussnote 9 bei yon Stietencron, p. 196 bemerkenswert. In ihr werden die vier "Veden" der Bhojaka-Maga behandelt, namlich Veda, Vidvavada, Vidud und A~girasa. Die ersten drei hatte A. Weber mit Yasna, Vispered und Videvdat identifiziert. H. Hoffmama sieht nun, wie yon Stietencron hinzuffigt, in At~girasa eine missratene Entsprechung des Nirangistan. H/itte er damit recht, so hatte er auf diesem Wege wider Willen die Nachrichten fiber die zoroastroiden Bhojaka-Maga als Geschichtsf~ilschung des 18. Jahrhtmderts erwiesen, als ganz rezenten Versuch, die Tradition der Bhojaka a n die der Parsen anzuschlieBen. Die Bezeichnung des vierten Veda nach dem Nirangistan setzt n/imlich eine besonders hervortretende Bedeutung dieses letzteren voraus. Eine solche konnte abet das Nirangistan nicht etwa schon in sassanidischer Zeit gehabt haben, da es ja nur einen der 21 Nasks des awestisehen Kanons darstellte. Relativ grOl]ere Wichtigkeit konnte es erst nach Verlust des sassanidischen Kanons erlangen, und auch dann erst, nachdem sein Text wiederentdeckt und im Jahre 1720 dutch Mobad Jamasp yon Iran nach Indien gebracht worden war.. DaB eine Geschichtsf/ilschung vorliegt, scheint mir sicher. Sucht m a n jedoch einen Terminus post quem ffir sie, so liegt das Jahr 1720 offensichtlich zu spilt. Deshalb ist es wohl besser, Aflgirasa einfach als verderbten Ausdruck fiir eine Samlung yon magischen Formeln (nap. nTrang) zu betrachten. Dabei sollte man nicht vergessen, dab nfrang ein typisch persisches Wort ist und dab damit die Ubernahme yon den Parsen naheliegt, in deren Oberlieferung auch heute noeh die Nirangsammlungen eine wichtige RoUe spielen. Aber auch dann noch bleibt die Frage im Raume stehen, wie die BhojakaMaga des BhP Kontakt mit den Parsen bekommen und sich hinreichend fiber sie und ihre Lehren informieren konnten. Dazu m6chte ich meinen, dab die Begegnung im Zusammenhang mit den religi6sen Disputationen am Hofe des Kaisers Akbar stattgefunden haben k6nnte. Als terminus post quem erg~be sich dann das Jahr 1577 n.Chr., das mir nicht mehr so ganz unm6glich erscheint. Von Stietencron gibt an, in seinem Buche sei "nunmehr das wichtigste purhn.ische Material fiber die Maga ... zusammengefagt" (p. 20). Ware das richtig, so k~Snnte man verzeihen, dab "urn den Purh.na-Stil zu bewahren ... die weitschweifige Form der Erzahlung nicht gekiirzt" wurde. Es ist abet doch so, daB die Weitschweifigkeit auf der einen Seite zur Unterlassung auf der anderen geftihrt hat. Die wichtigen Passagen fiber die Begleiter des Sfirya, darunter Rhjfia und Srau.sa (aw. Ragnu und Sraoga) sind ausgelassen, wie m a n aus den dunklen Andeutungen p. 20 noch nicht erkennen kann, sondern erst festzustellen vermag, wenn man die p. 253 gegebenen Zitate vergeblich im Text yon yon Stietencron zu verifizieren sucht. In ~ihnliche Schwierigkeiten kommt man, wenn man den mehrfachen Andeutungen fiber eine Mihira-Verehrung der Maga (p. 253 etc.) nachgeht. In den yon yon Stietencron vorgelegten Textpartien findet sich Mihira lediglich als Familiermame (gotrat.n mihiram BhP 1,139,33). Weiteres, wie etwa

REVIEWS

47

BhP 1,70,3, muB man fiber Scheftelowitz, p. 295 usw. suchen, ganz abgesehen von den einschl~igigen auBerpurg.nischen Materialien, wie etwa Mbh. 3,3,61. Das Verdienst des Buches, das iibrigens materiell hervorragend ausgestattet ist, liegt darin, dab es den purfin.ischen Text durch Transkription und eine sehr ansprechende deutsche Obersetzung auch bequemen Lesern zug~inglich macht. Ansonsten macht es auf ein altes wissenschaftliches Problem aufmerksam und zeigt, dab dieses einer neuerlichen Bearbeitung harrt. Es geht dabei nicht nur u m die Umwendung altbekannten Materials, sondern auch um die Gewinnung von neuem. Zun~ichst w~iren wenigstens die von von Stietencron vernachl~issigten Oxforder Handschriften des BhP zu kollationieren. Lind schlieBlich di)rfte es sich lohnen, den Spuren der Sgkaldip-Brahmanen nachzugehen, die heute noch existieren und fiber die mir ein Gewfihrsmann berichtete, sie betfitigten sich als Astrologen. Mainz

Helmut Humbach

Richard N. Frye, Bukhara. The 3/Iedieval Achievement. Norman, Oklahoma, University of Oklahoma Press, 1965. XIII, 219 S. $ 2,75. Im Jahre 1954 hatte R. N. Frye seine englische Obersetzung yon Nargakbi's Tarikh-i Bukh6rd vorgelegt. Aus der Arbeit an diesem Thema ist nun auch dieses kleine, etwas schnetl geschriebene, vulgarisierende und deshalb verdienstliche Buch entstanden. In ihrn gibt F. ein teilweise recht farbiges Bild von der Geschichte dieser faszinierenden Stadt bis zum Mongoleneinbruch. Dabei betont er naturgemiiss die Rolle Bucharas als Zentrum der neupersischen Renaissance des 10. Jahrhunderts: "The golden age of the city coincided with the rise of the New Persian language and literature and with the ecumenical development of Islam, and Bukhara played an important role in both." Als wichtige Voraussetzung ffir das Aufkommen der Renaissancebewegung im Osten betrachtet Frye p. 100 den Umstand, dass dem Osten die stabile zoroastrische und sassanidische Tradition des Westens fehle. Mehrere Bemerkungen zeigen, dass Frye, Vorg/ingern folgend, fiberhaupt die Existenz eines unverf/ilschten Zoroastrismus in Buchara bezweifelt. So schreibt er p. 20 von Tughshada, dessen Gebeine nach Buchara gebracht worden sind: "It is probable that [he] was some sort of a local Zoroastrian." Die yon der Sage tiberlieferte Totenklage ffir Siyfivug l~sst ihn p. 27 schliessen: "It would be hazardous to identify the prevailing faith in Central Asia with the state religion of the Sassanians." Und p. 36 fiussert er: "Even the existence of Zoroastrian Mobads ... in Bukhara who could read Pahlavi is open to question." Hierzu ist eine Bemerkung yon A1-Biruni zu beachten, die sich in India p. 125, 2.3 (trsl. Sachau, I, p. 249 sq.) findet. Nach dem Weltbild der Magier von Sughd wird die Welt vom Berge 'rdy' (aw. haraiti barg) umgebem Ausserhalb yon diesem befindet sich der Owm (aw. haoma-). Im Zentrum aber steht der Berg grnTr, ein Pendant des indischen Meru. Die Transkription dieses grnyr durch Girnagar bei Sachau hat, wenn ich l~inrcichend informiert bin, die Erkenntnis verhindert, dass hier ein 'Berg der Berge' gemeint ist. Der Sache nach zu vergleichen ist b.sogd, sm'yr ?r"n 7wt'w 'Sumeru, K6nig der Berge', TS 6.47. Die Lautform grnTr ist jedoch nicht sogdisch. Herkunft aus dem Persischen ist allein schon aus lexikalischen Griinden nicht zu diskutieren (gar : kOf). So ist man wohl zur Annahme genStigt, dass grnTr etwa ein garinu-Tari reprfisentiert, welches ein altes awestisches *gairina.m gairig* fortsetzt. Dabei scheint es, dass grn?,r lautliche Eigentiimlichkeiten des Awestischen hewahrt hat, obwohl seine Vorform in unserem Awesta tiberhaupt nicht bezeugt ist. Das weist auf nine zwar yon der sassanidisch-persischen Restauration unabh/ingige, abet doch sehr best~indige Tradition.

48

REVIEWS

Merkwfirdig ist auch A1-Birunis Beschreibung des Mihr~n/Indus, India p. 130, 12.13 (Sachau, I, p. 260), die in Zusammenhang mit dem hapta h~ndu 'sieben Str/Sme, Siebenstromland' der awestischen L~inderliste (V. 1, 18) zu betrachten ist: "As the name union of the five rivers (mu~tami' al-anhar al-hamsa) occurs in this part of the world [in Panjab], we observe that a similar name is used also to the north of the above mentioned mountain chains, for the rivers which flow thence towards the north, after having united near Tirmidh (kama ida p,tama 'at 'inda Tirmi_da) and having formed the river of Balkh, are called the union of the seven rivers (ma~tami' al-anhdr as-sab'a). The Zoroastrians of Sogdiana have confounded these two things; for they say that the whole of the seven rivers is Sindh and its upper course Baridi~ (fa-qala anna ~umlata l-anhdri s-sab'ati Sindu wa aClahu Baridig).'" Nicht nur die Pehlewiiibersetzung (haft hincl~kdn), sondern auch die Komposition der awestischen L[inderliste (WZKSO, 4, 1960, p. 36 sqq.) zwingt uns dazu, hapta handu im Panjftb, oder, wegen der dort herrsehenden ungewiShnlichen Hitze (araOwimca gar~mdum), besser in Sindh zu suchen. Auch den Magiern von Sughd war diese Lokalisation bekannt, doch verfiigten sie nach A1-Biruni noch tiber eine andere Tradition. Diese ist, weist sie doch ins ostiranische Stammland, ohne Zweifel altertiimlicher. Mainz

Helmut Humbach

Karl Jahn, RashM al-Dfn's History of India. Collected Essays with facsimiles and indices ( = Central Asiatic Studies, X). The Hague, Mouton & Co., 1965. Rashid al-Din Fa.zl Allfih was physician to the Mongol sult.fialAbfiqfi Khan (663-680 H./ 1265-1281 A.D.) and wazir to both GhfizS.n Khfin (694-703 H./1295-1304A.D.) and Uljhytfi (703-716 H./1304-1316 A.D.). In about 703 H./1303-4 A.D., he completed his famous general history entitled Jdmi' al-tawdrikh. The work is widely known for its account of the Mongols but its last section deals with a history oftheyugas (world ages), the ancient and medieval history of India and of Kashmir and the life and teachings of Buddha. It is naturally of great interest to the students of Indian history and culture. Professor Karl Jahn has therefore rightly named this section "Rashid al-Din's History o f India". It contains two parts (qisms); the first divided into ten and the second divided into twenty chapters (fas.ls). Of it Professor Jahn has published facsimiles from the following three manuscripts: 1. An incomplete Arabic version of Jdmi' al-tawdrikh belonging to the Royal Asiatic Society, A 27 (714 H./1314-15 A.D.), facsimiles pp. 1-48. 2. An incomplete Persian version preserved in Topkapi Sarayi (Istanbul), 940Hazine 1654 (717 H./1317 A.D.), facsimiles pp. 51-94. 3. Another Persian version belonging to the British Museum, Add. 7628 (837 H./ 1433 A.D.), facsimiles pp. 97-153. Professor Jalm has assigned the following reasons for publishing Rashid al-Din's History of India in facsimile: (1) the number of really reliable and complete Mss. is very small; (2) the portions taken from al-BirOni have already been scrupulously edited and translated; (3) the hitherto unknown sections have been dealt with and their contents examined in the present collection of essays (pp. xv-xvii). The publication of facsimiles tends to save scholars from being misled by any errors which editors may commit, either in matters of fact, or in the establishment of the text. Nevertheless, only those who are trained in reading complicated manuscripts can make adequate use of facsimiles. Professor Jahn has himself rightly pointed out: 'Only the first four chapters (fa~ls) of the first p a r t . . , are based on al-BinSmi, ... As early as the fifth chapter, however, parts only earl be traced to al-Bin-mi, while borrowing from

REWEWS

49

Jftzj~ni has patently also taken place; and it is in this chapter that Rashid al-Din begins to speak... The sixth chapter contains entirely new historical material and deals with Kashmir and its rulers. After a geographical introduction, founded on al-Birfini, there follows a short extract from the history of the Kings that is of great importance, particularly for the Mongol period. Quite new are the following four chapters (in the Arabic version five) that treat of the forms and events of the four world ages (yugas). As is the case with the subsequent second part of the History of India, they can be traced back to Rashid al-Din's authority, the Bakshi K~malashri, about whom further information from either historical or literary sources has as yet not come to light ...' (p. xiii). In view of the importance of the fresh material embodied in Rashid al-Din's History of India, a critical and annotated edition on the basis of the manuscripts noticed by C. A. Storey (Persian Literature, Section II, Fasciculus 1, London, 1935, pp. 72-75), and some other like the two in the Ra.zfi Library R~mpfir (India), would undoubtedly be welcomed by students of ancient and medieval Indian history and culture. For the time being, the detailed analysis of the contents (except for chapter V, part I), a general index to the facsimile manuscripts and an index to introduction and essays, which the editor has presented, should help readers to make some use of the text. The essays incorporated in the present work, now amended and arranged, are certainly a great help to all those who are interested in the subject. It seems that the Turks had never been able to completely annihilate Buddhism in the lands conquered by them. During his wanderings, al- .Hall~.j (857-922 A.D.) is believed to have stayed, in about 902 A.D., with Buddhist monks (Bakshis), in Turkist~.n (~ RisMa-i-lqbMiya, British Museum Mss. Or. 9725 f. 46a). The conversion of a considerable section of the il-Kh~ns to Buddhism posed a threat of considerable magnitude to orthodox Islhm in their territory. According to 'AI~' al-Dawla Simn~ni (1261-1336), an eminent .stifi and Muslim revivalist, the Buddhists tried to assimilate Islhm in their own fold by misinterpreting the theological principles oflsl~.m and by projectinga highly distorted picture of Buddha and his teachings. They emphasised that all the Saiyids and holy men of Islam practised idolatry. In 686H/1287 A.D., he held polemical discussions with some Bakshis from India, Kashmir, Tibet and Uyghurist~.n who lived at Sul_tS.niya, the new Mongol capital, under the patronage of Gh~.z~.n Khan. 'Aal~,' al-Dawla claims to have routed them in all the debates that were held there and to have considerably impressed G__hh~iz~nK._hh~.n. ('Al~'al-Dawla, AI'Urwa, British Museum Mss. Or 3649ff. 142a.) Great champion of orthodox Isl~tm as he was, 'AI~.' al-Dawla exercised considerable influence over a large section of the Muslims in Persia and, through his disciples, over Kashmiris and on Indian Muslims. He and his disciples occasionally preached orthodoxy in a most militant manner. The account of his debates with the Bakhshis in his Al-'Urwa liAhlal-Khalwa sets out those beliefs of the Buddhists which he sought to controvert. The works of his disciples also incidentally reflect the beliefs of contemporary Buddhists. Though Al-'Urwa was not completed until 721 H./1321 A.D., the polemics of 'Al~t' al-Dawla and his disciples must have been known to Rashid al-Din. The fact that he relied mainly on K~.malashri, the Kashmiri Bakshi, for his account of Buddha and his teachings, and on A1-Birfini for his account of Hinduism, tends to show that he was anxious to draw upon the most authentic sources and that his objectivity was not greatly affected by comtemporary Muslim revivalist trends and prejudices. A comparison of the teachings of Buddha, as gleaned from Rashid al-Din's History oflndia, with the principles of Buddhism that can be culled from Al-'Urwa might give us a better understanding of Buddhism in Persia at that time. We may perhaps tentatively assume that the Mahay~.nist Buddhists of Central Asia and Iran at that time did not stress the Shaktistic or T~ntristic practices. K~malashri's account certainly gives ground for this suggestion. As far as the importance of Rashid al-Din's History of

50

REVIEWS

India is concerned, one cannot dispute the conclusions of Professor Jahn that 'the work may be considered as a substitute for the certainly numerous Buddhist writings, of which nothing has been handed down to us but which were undoubtedly in widespread circulation in Iran at the time' (p. xxxiii). Since Professor Jahn has published only his collected essays, he has not paid attention in this book to the importance of the chapter relating to the Sul~ns of Delhi in Rashid al-Din's History of India. (Part I, chapter V, Facsimiles RAS, pp. 14-20, Ts, pp. 67-75, and BM, pp. 112-118). The chapter seems to have been based on Mongol sources not accessible to contemporary Indian scholars. Though the account of the Sul_tfins of Delhi embodied in the present work is very brief, the information relating to Sultfin Balban and his successors down to 703 H./1303-4 A.D. is unique and exceedingly valuable. Rashid al-DLu is the first writer to assert that Ulug.h_hKhan organised a rebellion against Sul.tfin Ra.ziya (1236-1240 A.D.) and assassinated her; the weight of evidence and tradition, however, does not support him. Ulugbh Khin, who after his accession assumed the title of Sultfin Ghiyis al-DLu Balban (1266-1287 A.D.), had organlsed regular campaigns against the Mongols in the reign of his son-in-law SultAn Ni.sir al-Din Mal).mfid (1246-1266 A.D.) and had deputed his son Prince Muh.ammad to fight against them when he himself ascended the throne. These campaigns insured that Balban's name was well known to the Mongols. The Tabaq6t i-N6s.iri of Minhij bin Sirfij Jfizjini, dedicated to Sultan Nil.sir al-Din Mah. mild, comes to an abrupt end after bringing the account down to Sha'bfin 658 H./July-August 1260, though the author continued to live well into Balban's reign. The manner in which Balban came to the throne was not dealt with by Z.iyi al-Din Barani, whose history of the Sul.tfins of Delhi runs from the reign of SultAn Balban, to Sul~n Firfiz Shah Tugbluq's sixth year (758 H./1357 A.D.). Only 'Isfimi, the author of Futab al-Sal~in, who composed his work in 1349-50, gave a detailed account of the circumstances under which Balban killed his father-in-law (Futa.h al-Sal6tin, Madras, 1948, p. 163). Rashid al-Din has categorically accused Ulugh_ Khfin of assassinating his father-in-law, and has added two Persian couplets to make his statement more effective. In the Arabic version of the History of India the couplets are retained in the original Persian facsimile (p. 18). Equally interesting is Rashid al-Din's account of the rise of Sultan Jalil al-Din Fir~z Khalji (1290-1296 A.D.). He says that, in the reign of the son of Sult_in Ghiya_s al-Din Balban (obviously a slip; he was succeeded by his grandson, Mu'izz al-Din Kaiqubfid), Hulagu Khfin summoned his governor of Sindh, Malik NA.Siral-Diu son of Wafa Malik (name not mentioned in the General Index to Facsimile Mss.; pp. 18, 73, ll7) to his court and killed him because of the intrigues of his enemies. Malik Firfiz Kfah, who governed the fort of Sindh on behalf of Malik Ni.sir al-Din repaired to Delhi and entered the service of the son of Sul.tfin Gh_iyi_s al-Din (sc., his grandson, Kaiqubad). The Sultfin commissioned him to guard Multin, the frontier of his kingdom, and protect it from the depredations of the Mongols. For some time he served there. Those who were envious of him, accused him of being friendly with the Mongols. The Sul.tfin summoned him to Delhi. The wazir dealt with him in a high-handed manner. Malik Firfiz Kf~h retaliated by killing the wazir and rode his horses through the Sultan's palace; an episode illustrated in the facsimile (p. 73). The Sultin's retinue were scattered; and he himself was an epileptic. Malik F~rQz killed the Suit.fin. The members of his armed forces raised one of the infant sons of the deceased Sultfin to the throne. Malik Firfiz won the obedience of the whole army; then killed the boy king and himself ascended the throne. The circumstances under which Sul.tfin 'Al~t' al-Din Khalji ascended the throne (1296-1316 A.D.) after assassinating his father-in-law Snltin (Jalil al-Din) Fir~z have have been very vividly described by Rashid al-Din. A miniature in the Topkapi Sarayi manuscript illustrates the event (p. 74) in a very naive but realistic manner. The miniature in the Arabic version (p. 19) does not correspond with the facts; it shows

REVIEWS

5l

Sultan Fir~z bound and blindfolded, kneeling before 'Ala' al-Din, who sits upon a throne. An executioner stands over him, sword raised, while courtiers look on. The account of Suit.fin Alfi' al-Din's reign is brought down to 703 H./1303-4 A.D. The work helps us in correctly fixing the date of the famous Mongol invasion made by Quflumla Khwfija son of Duwah son of Baraq (pp. 20, 75, 118). Professor Jahn's edition has already beea warmly received by Indian scholars. Professor Budh Prakash, who in January 1966 read a paper on Rashid al-Din's Jami" al-tawfirikh in a seminar held under the auspices of Jhmi'a Millia Delhi, profusely drew upon Professor Jahn's essays. A.N.U., Canberra

S.A.A. Rizvi

A Critical Pali Dictionary, begam by V. Trenckner: Vol. II, continuing the work of D. Andersen and H. Smith, comprising the material collected by W. Geiger, edited by an international body of P~li scholars, and published by the Royal Danish Academy of Sciences and Letters, in commission by E. Munksgaard, Copenhagen [with the financial assistance of UNESCO on the recommendation of the International Council for Philosophy and Humanistic Studies], Fascicle 3 (1965): ~pattikusalat~-dyu, pp. 97-136. Dan. kr. 22,00. This fascicle consists of articles contributed by six scholars belonging to four different nationalities: F. Moller-Kristensen (apattikusalata-apa), B. Vimalabuddhi (apakaapathaka), J. A. B. van Buitenen and E. Pauly (lama-amisesana), E. Pauly (amukharh~modeti), and K. de Vreese (aya-ayu). The work is indeed a tribute to international teamwork. A uniformly high standard has been achieved. Pfili scholars from all over the world have lent their skills to this project, not only from Scandinavian countries but also from Ceylon, Czechoslovakia, England, Germany, Holland, USA; and now scholars from Burma, Japan and Thailand are ready to give their full co-operation. It would be useful if the fascicles would contain lists of publications relating to CPD,1 and also of recent publications which contain materials important for PHi lexicography. In this connection we can already draw attention to the following reviews of the different fascicles of Vol. II: (Fasc. 1) A. Minard, BSL, LVI, 2 (1961), p. 62; P. Horsch, OLZ, LVII (1962), Sp. 189-190; F. Edgerton, JAOS, LXXXII (1962), pp. 90-91; C. Haeber, ZDMG, CXIII (1963), pp. 353-355; (Fasc. 1-2) A. K. Warder, 11J, VII (1964), pp. 238-240; (Fasc. 2) P. Horsch, OLZ, LX (1965), Sp. 74-75; (Fasc. 1-3) C. CaiUat, IF, LXXI (1966), pp. 306-310. A full account of CPD has been given in an article written by E. Mayeda: "Denmfiku-Gakushiin no 'Hihanteki P~rigo Jiten' sono Enkaku to Genj6 - [CPD of the Danish Academy - its history and present state -]", Tdh~gaku, No. 24 (T6ky6, Sept. 1962), pp. 112-101; No. 25 (March 1963), pp. 121-106. All Indologists are grateful to the Royal Danish Academy for sponsoring this great enterprise. The appearance of each issue is an important event, and is always eagerly welcomed by Phli students. Much important work in the field of Pfli lexicography has been done by scholars such as R. C. Childers, A Dictionary of the Pfili Language (London, 1872-1875, repr. x It is very difficult to know such publications as reports submitted to various academies, e . g . C . C . Berg, "Dictionnaire pfli [Rapport h l'Union acad6mique internationale]", Bulletin de la Classe des Lettres et des Sciences morales et politiques de l'Acaddmie Royale de Belgique (Brnxelles), XLVII (1961), pp. 477-482; etc., etc.

52

REVIEWS

several times) [abbreviated ChiM. in CPD], H. Kern, Toevoegselen op 't Woordenboek van Childers (Amsterdam, 1916) [Kern or Toev.], T. W. Rhys Davids and W. Stede, PTS's P6li-English Dictionary (London, 1921-1925, repr. 1959) [PED], A. P. Buddhadatta, Concise P6li-English Dictionary (Colombo, 1949, 2rid ed. 1957), id., EnglishPali Dictionary (London, 1955), S. Kumoi, Pa-Wa Sh6fiten [Concise P6li-Japanese Dictionary] (KytSto, 1955-1961), Pe Maung Tin, The Student's Pal#English Dictionary (Rangoon, 1920, repr. 1961) [MTD], V. Subhfiti, Abhidh6nappadipika (Colombo, 1865, 5th ed. 1938) [Abh], Widurupola, English-Pali Dictionary (Colombo, 1949); C. Akanuma, Indo Bukky6 Koya Meishi Jiten (Genshikihen) [Dictionary of proper names in Indian Buddhism (primitive period)] (Nagoya, 1930-1931, repr. Ky6to, 1967), G. P. Malalasekera, Dictionary of Pali Proper Names (London, 1937-1938, repr. 1960) [PPN]; Pali Tipit.akafft Concordance, begun by F. L. Woodword (London, 1952-), K. Mizuno, Nanden Daiz~ky6 S6sakuin [General index to the Japanese translation of the Pali Tipit.aka] (T6ky~5, 1959-1961); id., Parigo Jiten [Pfili-Japanese dictionary] (T~3kyS, 1967); C. Akanuma, Kan-Pa Shibu-Shiagon GoshOroku [Concordance to the Chinese .~gamas and Pali Nikayas] (Nagoya, 1929, repr. T6kyO, 1958), and also works by D. Andersen, M. H. Bode, V. Fausb~ll, O. Franke, W. Geiger, E. H. Johnston, S. Konow, R.Morris, H.Smith, etc., etc. After the labour of so many scholars, the Critical Pali Dictionary may well be considered the crown upon the edifice. Canberra

Akira Yuyama

Yutaka Iwamoto, Bukky6 setsuwa kenkys josetsu ( = Bukky6 setsuwa kenkya, I). KySto, HSzSkan, 1967. 300 pp., 1 frontisp. Yen 4200. Professor Iwamoto (2.3.1910-) has been studying the avad~na literature for many years. He has written a series of articles (listed on page 18 of the Bukky~ setsuwa kenky~ josetsu) and two short books: Indo no setsuwa [Indian tales] (TSky6, 1963); Bukky6 setsuwa [Buddhist tales] (T6kyS, 1964).1 Moreover he has published a revised version of the Sanskrit text of the Sumagadhdvadana (T6ky6, 1959), which was first published by Tokiwai Gy6y~ (1872-1951) in 1918, and an edition of the Tibetan version (Acta Asiatica, 7, 1964, pp. 1-19). Bukky6 setsuwa kenkyft josetsu [Introduction to the study of Buddhist tales], which appears as volume I of Bukkya setsuwa kenkyft [Studies in Buddhist tales], constitutes a general introduction to the avadana literature. In the past much work has been done in this field by L6on Feer (1830-1902), Sergej Ol'denburg (1863-1934) and J. S. Speyer (1849-1913). The last comprehensive survey of avadana literature was given in 1909 by Speyer in the preface to his edition of the Avddanadataka (Bibliotheca Buddhica, III, pp. I-CX). The introduction (pp. 9-47) deals with a survey of editions and publications relating to avadana literature, problems in the study of avad6nas, the nature of avadana and the change in the meaning of the word avadana. Speyer (op. cit., p. XIV) divided the avadana texts into three classes. Professor Iwamoto points out that this division does Both volumes contain a wealth of interesting material, lndo no setsuwa contains an introduction, dealing with Indian story-literature, and two main parts. The first studies Indian tales in Japan, special attention being given to the Konjaku monogatari in which many Indian elements are to be found. The second part deals with Indian stories in Europe, and discusses, among others, the story of Barlaam and Josaphat, Shakespeare's Merchant of Venice, and Nathaniel Hawthorne's story "Rappacini's Daughter". In Bukky6 setsuwa Professor Iwamoto studies in detail the different versions of the following ten stories: Afigulim~la, Devadatta, Kot.ikar.na, Kun~.la, King Candraprabha, Maitrakanyaka, Prince Kaly~.nak~rin,Prince Ku~a, l~.sya~r6gaand Si.mhala.

REVIEWS

53

not take into account the distinction between avaddna stories and avaddna literature. A text such as the Avaddna~ataka belongs to the avaddna literature yet contains not only avaddna stories but also vydkara.nas and jdtakas. As characteristic of avaddna stories Iwamoto indicates the fact that they consist of a story of the present and a story of the past. The Buddha identifies the chief actor of the first story with the chief actor of the second story. Avaddna stories contain in general the following five elements: I. Story of the present; 2. Verses praising the power of karman; 3. Story of the past; 4. Appearance on the scene of a Buddha of the past; 5. Teaching relating to karman. With the nature of the avaddna stories thus determined, Iwamoto studies the divisions of the Buddhist teaching in nine and twelve ahgas, and remarks that to itivuttaka, in the group of the nine a~gas, correspond the following four, vydkarana, niddna, avaddna and itiv.rttaka in the group of the twelve a~gas. Iwamoto attempts to determine the original meaning of these at~gas and the classes of texts to which they relate. There are many problems connected with these two divisions into nine and twelve m~gas, and Japanese scholars in particular have studied the relevant passages in Buddhist texts exhaustively. Among the most recent has been Professor Egaku Maeda who has devoted a large section of his Genshi BukkyO seiten no seiritsu kenkyf~ (English title: A History of the Formation of original Buddhist texts) (T6ky6, 1964) to a very thorough study of the nine and twelve m~gas (see pp. 181-547). It seems to me, however, that no definite conclusions can be reached from the available evidence. Most of the materials come from a period when the division into nine and twelve at~gas had lost any significance if it ever had any! The explanations of the meaning of the a~gas given in Buddhist texts are of great interest because they reflect opinions and theories prevalent at the time when these texts were written. But, it is dangerous to accept these explanations as relevant to a much earlier period. It is even doubtful if the Buddhist texts were ever divided into nine or twelve classes corresponding to the nine and twelve a~gas. In these circumstances one can only speculate, and the few pages devoted to the at~gas by Iwamoto (pp. 32-36) must therefore be regarded as largely hypothetical. In order to make a critical examination of the theory advanced by Iwamoto, one would have to consider not only Maeda's book but also the third chapter of Professor Hirakawa Akira's Ritsuz6 no kenkyft [A Study of the Vinaya-pi.taka] (Tokyo, 1960) in which much useful material on avaddna is to be found (see pp. 291-415). Of great interest is the last part of the introduction in which lwamoto studies the change of meaning of the word avaddna which, in the course of time, is applied to parables and examples and, finally, is used as synonymous with the word kath6. In his introduction Iwamoto points out the necessity for careful study of the history of the avaddnas which are to be found in the avad~na literature. The first chapter deals with the history of one story, the Sum6gadh6vad6na, which, according to lwamoto, deserves to be studied for five reasons enumerated on pages 22-23. The Surn6gadhdvad~na is the legend of Sum~gadh~, the daughter of An~thapi.n.dada. She marries a Jain merchant from Pu.n.dravardhana, by the name of V rsabhadatta. She alienates him from the Jain monks, and then invites the Buddha who duly appears from the sky and converts V.rsabhadatta's family and the whole town. The Buddha tells that in a former existence she was K~ficanamfilfi, the daughter of King K.rkin who dreamed ten strange dreams. K~ficanam~lfi advised her father to ask the Buddha K ~ y a p a for the explanation of these dreams. After having told the story of K~ficanam~l~ the Buddha explains that in a former life K~ficanamfil~ was a poor woman who dedicated a wreath of pal6la-ftowers to the eaitya of a Pratyekabuddha. Through the merit of this act she was reborn with a golden garland on her head as a daughter of King K.rkin. This is a brief outline of the story as given in the Sanskrit text edited by lwamoto. In this edition Iwamoto remarks that the story is divided into two parts: an account of Sam~gadh~'s present life and the story of her former existence as a daughter of King K.rkin. However, the legend contains a third part: the story of K~ficanam~l~'s former

54

REVIEWS

existence as a poor woman so that, properly speaking, the Sanskrit text presents a n avad~na in the second degree as it relates the story of Sum~gadhfi and two of her former existences; the first as K~ficanam~lfi, the daughter of King K.rkin, and the second as a poor woman. If one divides the legend in this way into three parts, it becomes evident that parts two and three constitute a regular avadana. Part two relates Kfificanam~l~'s existence as the daughter of King Krkin and part three how she earned in an earlier existence the merit which brought about her rebirth as daughter of King K.rkin. Parts one and two do not constitute an avad~na because, although part two tells how K~ficanam~tl~t believes in the Buddha and provides him with everything he needs, there is no mention of a particular meritorious deed as is usually told in the stories of the past. It should also be noted that in part two K~tficanam~lfi, as daughter of a king, has a higher position than Sumagadha, the daughter of a merchant. For these reasons it is obvious that the Sanskrit text consists of two heterogeneous stories (part 1 ; parts 2 and 3) which have been clumsily put together. Iwamoto has failed to draw attention to this fact which is fully confirmed by the comparison of the different versions of the legend. The Sanskrit text of the legend is transmitted in eight manuscripts of which Iwamoto has been able to use seven. MS B, which dates from the thirteenth century, is, in several respects, different from the six others which descend from a sub-archetype 13. The Sumhgadh~. legend is also to be found in K.semendra's Avad~nakalpalat~ Nr. 93. This text has been re-edited by Iwamoto on the basis of two manuscripts in the Cambridge collection as an appendix to his edition of the Sanskrit text of the SumSgadh~vadana. The Tibetan Kangur contains a version of the legend, and the Chinese canon five different versions, of which the oldest was translated A.D. 230 and the latest after A.D. 980. Iwamoto carefully studies all the versions which he summarizes and compares with each other. In the sixth section of this chapter, the author sketches the development of the legend and distinguishes five different stadia: A, B, C, D and E. Moreover, he tries to date these stadia and to establish their connections with Buddhist schools. Iwamoto's conclusions here are not always acceptable. Regarding the story of the poor woman MS B relates that as recompense for the dedication of a garland ofpalala-flowers she was reborn as Kfifieanam~l~ (w167 265-266 of Iwamoto's edition). The story of the past ends here but the Sanskrit manuscripts add another passage which varies according to the manuscripts. MS B has the following text: pra.nidh~naeittayd guddhay~

guddhavastuy6 ~ryasa .mgh6yadatta .m. ten~dhye.su kules,u praty~j~t~ pran. idh~navas~e ceti. ya Kd~canam616 sa Sum~gadha. In the other manuscripts the following text is to be found: pa.nadvaya .mca tay6 pat.6ntavabaddha .m vimucyaryasahghaya datta .m. ten~d,hyes.u kule.su jata pran.idhdnavasac ceti. There is nothing corresponding to this passage in the Tibetan version or in K.semendra's Sum#gadhavad6na. Only in the Chinese version, translated after A.D. 980, is there a similar passage: "O monks. Therefore, at present, this Sumhgadh~ on account of her past root of merit and the power of her great vow has now obtained to be born in the very rich family of An~.thapi.n.dada. By her excellent guidance and by gifts she has done her duty in regard to the Buddha" (Taish6, II, p. 854a 7-9). It is easy to understand why this passage was added to the Indian text underlying this Chinese version. Once the Sumagadhavadana was formed by joining together two different stories, comprising the three parts mentioned above, it became necessary to explain the relation of part three, dealing with the poor woman, to part one, the story of Sum~gadhh. This passage, which is absent from all the older versions, is differently worded in the three versions in which it is to be found. In the Chinese version it clearly answers its purpose, but in the versions represented by MS B and by fl it contradicts what has been said before. According to MS B she has given gifts to the samgha. The other manuscripts specify that her gift consisted of two small coins. It is evident that in all the Sanskrit manuscripts the text has been corrupted in the course of transmission. MS B, dating from the thirteenth century, has a text which is slightly

REWVWS

55

more satisfactory than that represented by the other manuscripts which probably belong all to a much later period. ~ The gift of the two coins by the poor woman clearly alludes to the famous story 22 of the Satrala~nkara, which, following the translation by Huber in 1908, widely attracted the attention of the scholarly world with its parallel to the story of the "widow's mite" in the New Testament (see the literature quoted by Winternitz, A History of Iudian Literature, II, pp. 408-409). The same story occurs in the Kalpanama.n.ditika of Kum~ral~ta (ed. H.Lfiders, Leipzig, 1926, p. 149). Iwamoto assumes that the Satrala.mkara was written by A~vagho.sa during the reign of Kani.ska in the second half of the second century, and the Kalpandma~jd.itika or D.rs.t6ntapatikti by the Sautrantika patriarch Kum~ral~ta at the end of the second century or in the beginning of the third century. As this story does not seem to occur elsewhere in the Buddhist literature Iwamoto concludes that it must have been borrowed by the author of the sub-archetype [3 shortly after Kum~,ral~ta, probably in the middle of the third century A.D. or at the latest in the second half of the third century. This conclusion seems to me unacceptable. In the first place, the author of 13could very well have found this story somewhere else, for the surviving Buddhist literature represents only a part of the literature which once existed. But a more serious objection lies in the fact that the gift of two coins is not mentioned in the Tibetan version, in K.semendra's version, in the Chinese version, discussed above, and in MS B, but is found only in recent manuscripts. For the same reasons it is impossible to accept Iwamoto's view that the author of 13 belongs to the Sautrhntika school, or that stadium E (which precedes 13)constitutes the transitional stage between the Sarvhstivhdin and Sautrantika schools. Another difference between MS B and 13 is the fact that the former one contains a comparison of the Buddha to gold. This comparison is absent in 13. Iwamoto points out that in early times the Buddha is compared to gold and that later in Mahfiyhna texts the Buddha is said to be ornated by light, or, to consist of light. Iwamoto regards it of great importance that this comparison is omitted from stadium E, but he does not explain why. Possible his idea is that such a Mahhy~na elements was still admissible in stadium D, belonging to the Sarv~stiv~din school, but not anymore in stadium E which marks the transition to the Sautr~ntika school. However this may be, I do not believe that the presence or absence of this comparison carries much weight from a doctrinal point of view. A few pages later Iwamoto quotes a story from the Mahi~8saka vinaya. 8 In it the Buddha is compared to a mountain of gold (p. 96,1. 15). According to the Buddhist traditions the M a h i ~ s a k a descend from the Sarvastivhdin as do the Sautr~ntikas. Using quotations in the texts Iwamoto tries to show that stadia A, B, C and D all belong to the Sarv~stiv~idin school, though in a note he adds that stadium B' belongs to the Dharmaguptaka school because it occurs in the Ekottaragama. This does not conflict with his theory because the Dharmaguptaka school descends from the Sarv~stiv~idin school. However, there is no agreement between scholars that this agama belongs to the Dharmaguptaka school. Some have even maintained that it belongs to the Mahfisfi.mghika school. This would seriously conflict with Iwamoto's view (see Maeda, op. cit., pp. 671-672). Since most of the quotations relate to stereotyped ex2 Iwamoto does not give a detailed palaeographic description of the manuscripts. According to Bendall's catalogue (pp. 129-133) MS C dates from the seventeenth century. Iwamoto shows that MSS C' and P are copies of this manuscript. However, no information is given about the date of the three other manuscripts of which two belong to Kyoto University and one to Tokyo University. a Cf. t~d. Chavannes, Cinq cents contes et apologues, II (Paris, 1910), pp. 343-349; Andr6 Bateau, "La construction et le culte des stfipa d'apr6s les Vinayapit.aka", BEFEO, 50 (1962), pp. 265-267.

56

REVIEWS

pressions, they prove little. The fact that the dreams of King Kg.kin are quoted by Vasubandhu and Ya~omitra in no way proves a doctrinal relation with the Sarv~stiv~tdin school. More important is the invitation of Pi.n.dola which is mentioned in the oldest Chinese translation dating from A.D. 230. The invitation of Pi.n~tola by Sum~gadh~ is mentioned in three versions of the A~oka legend: the Ku.ndldvad~na of the Divydvadana; the A-yii wang chuan and the A-yii wang ching (Iwamoto, pp. 84-85). In La Ldgende de l'empereur A~oka (Paris, 1923) Przyluski has connected the A~oka legend with the Sarv~istivadin school of Mathur~. However, the connection of the oldest Chinese version of the Sumhgadh~ legend with the Sarvftstiv~din school is not proved by a reference in the A~oka legend to the invitation of Pin..dola by Sumfigadh~. The author of the A~oka legend quotes from the Sum~gadh~ legend and not vice versa. It is interesting to note that the author of the A~oka legend has known the Sumagadh~ legend, which guarantees a considerable antiquity for the legend. In the Sum~gadh~i legend the invitation of Pin.d.ola is not an integral part, as has already been shown by S. L6vi and l~d. Chavannes in their detailed study of Pir~.dola (JA, Juillet-Aofit et Septembre-Octobre 1916; see p. 127 of the offprint). In this study, to which Iwamoto does not refer, L6vi and Chavannes proved that Pi.n.dola's flying through the air with a rock is an essential element of the Pin..dola legend and is to be found in several texts. For this reason the mention of Pin..dola in a version of the Sumagadhft legend does not at all prove its connection with a particular Buddhist school. I believe that it is generally very difficult to prove that a story belongs to a particular Buddhist school. There are several Indian stories of which Hindu, Buddhist and Jain versions are known. A story can very well have been adopted by different schools. Probably, at most, one can show that a certain version of a story, but not the story itself, belongs to a particular school. As Iwamoto himself remarks there is a parallel story in the Mahdvastu (vol. 1, pp. 301-307) which certainly has no relations with the Sarvhstiv~din school (cf. Iwamoto, pp. 99-100 and p. 111, n. 42). Already the existence of this parallel makes it difficult to attribute this legend entirely to the Sarvastivftdin school. In the last section of Chapter I Iwamoto draws the attention to three parallel legends in the Dharnmapadat.t.hakatha and the ManorathapFtra~.ff. In these legends, though the names are different the story is quite similar to the story of Sum~gadhh. Much less similar is the story of Vis~khh in the Dhammapadat.t.hakath~. Vis~kh~i is the daughter of Dhanafijaya, a treasurer (not a rich merchant as stated by Iwamoto in his summary) of the city of Bhaddiya. Dhanafijaya settles in Saket~, not far from S~vatthi. Visakh~ marries Pu .n.navaddhana, son of the treasurer Mig~ra in SavatthL Mig~ra is a supporter of the Jain order. Vis~kh~ invites the Buddha who converts Mig~ra and his wife. Apart from the fact that the husband of Vis~khh is called Pu.nn.avaddhana, which in the Sum~gadh~ legend is the name of a king or of a city, only the general theme is similar to that of the Sumfigadh~ legend. A daughter of a wealthy merchant or a treasurer marries an adherent of the Jains in a distant place; she invites the Buddha who converts her parents-in-law. According to Iwamoto this legend is very old because it relates to a period when Sravasti was not yet converted to Buddhism. He believes that this legend is based upon a historical fact because in later times Vis~kh~i was a famous patron of Buddhism in Sr~vasti. Therefore he assumes that the oldest form of the Sum~gadh~ legend is represented by the Visftkh~ story. Later, when Anhthapi.n~ada became famous as a protector of Buddhism in ~rfivasti, Sumfigadh~ was substituted to Visfikhfi. It seems to me difficult to admit these conclusions. The stories of Vis~daa and Sum~gadh~ are variants of the same general theme. In some way or the other the name Pu.n.navaddhana was connected with this theme. Moreover, there is not the slightest proof as to the historical background of the story of Vis~ikhg. Iwamoto has done well in drawing attention to this story, but I am afraid that it is impossible to regard the stories of Vis~kh~ and Sumfigadha as reflecting respectively a

REWEWS

57

historical fact and a later legendary transformation of this fact due to historical developments. Burnouf was the first scholar who studied the legend of Sumfigadh~, comparing the Sanskrit text with the Tibetan version (Introduction & l'histoire du buddhisme indien, Paris, 1844, p. 566; see also L. Feer, Papiers d'Eugdne Burnouf, Paris, 1899, p. 65). Feet had prepared a translation and edition of both the Sanskrit and Tibetan versions but his work has remained unpublished (see Bibliographie bouddhique, II, Paris, 1931, p. 15). Professor Iwamoto has rendered a great service to Buddhist studies by publishing a critical edition of the Sanskrit text (Tokiwai's edition is based on only one manuscript), an edition of the Tibetan version and a detailed study of all the versions of the legend and of parallel stories. Although it is not always possible to accept his conclusions, he has brought to light a lot of interesting material. One can add to these materials a passage of T~ran~tha's History of Buddhism which was first noticed by Minayeff (Recherches sur le Bouddhisme, Paris, 1894, p. 89) de'i-tshe ~ar-phyogs-nas

'phags-pa rTsibs-logs ces bya-ba dgra-bcom-pa ma~-du thos-pa'i mthar-phyin-pa cig byuh-ste / des gnas-brtan mati-du thos-pa "ga'- las gSer-'phreh-can-gyi rtogs-brjod ces bya-ba rgyal-po Kri-ki'i rrni-lam lut~-bstan-pa'i mdo-la sogs-pa redo din-tu dkon-pa dag kya~ 'don-par byed-do // de rgyal-po Ka-ni-s.kas thos nas/ ... : "At that time in the East there was the Arhat P~r~va who was extremely learned. He recited the Kdgcanamal6vadana, in which the dreams of King Krkin are explained, and other very rare sfitras [which he had obtained] from some learned sthaviras. When King Kani.ska heard this ..." (see Schiefner's edition, p. 47.13-19, and his translation, p. 59.16-23). Thran~tha's source for this tradition concerning the relation of P~r~va with the Ka~canamalavadana is not clear, gZon-nu-dpal, who wrote his "Blue Annals" from 1476 to 1478, quotes the Sumagadh6vadana but is silent about Phr~va in this connection (see George N. Roerich, The Blue Annals, I, Calcutta, 1949, pp. 25-27). Perhaps a thorough investigation of the Tibetan literature may lead to the discovery of T~ran~tha's source. In any case, one must be cautious as to how much importance to attach to this tradition, even if it can be proved to have been known in Tibet long before T~rangttha wrote his History of Buddhism in 1608. Chapter Two deals with the Avadanagataka. This text consists of ten groups of ten stories. Later metrical paraphrases as the Kalpadrumavadanamala, the Ratndvaddnamdla and the Agokavadanamala seem not to have known the fourth group, although this already existed in the Chinese translation with a rather different arrangement of the stories. There are a few problems connected with this translation which are not mentioned by Iwamoto. The translation is attributed to Chih Ch'ien who translated many texts between 220 and 253 (see E. Ziircher, The Buddhist Conquest of China, Leiden, 1959, pp. 335-336, n. 136). However, his translation of the Avadanadataka is not mentioned in the oldest catalogues nor in his biography, but seems to be recorded for the first time in the Ta-T'ang nei-tien-lu (Taish6, vol. LV, p. 227c19) which dates from A.D. 664. The same information is to be found in the K'ai-yiian shih-chiao lu (Taisha, vol. LV, p. 488c7; see also Tokiwa Daij6, Yakky6 sOroku, TOkya, 1938, pp. 558 and 564). One would have to compare the translation of the Avadanadataka with other translations which can be attributed to Chih Ch'ien without any doubt, before accepting as a fact that this translation is really his. Another matter which calls for fuller investigation is the indication, given in TaishO edition, that the arrangement of the chapters is different in the Sh6go.za. Iwamoto points out that the construction of stories 1-40 is different from the stories 41-100. One confirmation of this difference, in Iwamoto's view, is the presence of a colophon after story 40 which is absent after other stories. However, a similar colophon is to be found after the stories 50, 60, 70, 80, 90 and 100. Much more striking is the fact that both story 40 and story 100 contain long quotations from the Mahapar#~irva.nasatra. In this connection Iwamoto would have done well to have referred to two

58

REVIEWS

important articles: Jean Przyluski, "Le Parinirvfin.a et les fun6railles du Buddha" (JA, 1918-1920; offprint, Paris, 1920, 216 pp.) and Charlotte Vaudeville, "La L6gende de Sundara et les fun6railles du Buddha darts l'Avad~na~ataka" (BEFEO, LII, 1964, pp. 73-91). Przyluski has pointed out that story 100 can only have been composed in Gandhfira or Ka~mir (p. 7 of the offprint; Iwamoto, p. 134). Iwamoto compares the Chinese version of story I00 with the Sanskrit text and concludes that the original text underlying the Chinese version of the Avadana~ataka is a product of the Sarv~tivfidin school in Mathur~. It seems doubtful whether a conclusion based on comparison of different versions of a single story can legitimately be extended to the whole of the

Avadanagataka. In the third chapter, Iwamoto studies the Divy~vadana and related texts. A section of this chapter is devoted to the Ky6to manuscript of the Divyavad~namal~. Cowell and Neil's edition of the Divyavadana is entirely based on manuscripts which go back to one source, a manuscript of the seventeenth century. It is therefore important to note that there are other materials consisting of two Divyavadana manuscripts in Tokyo, and three Divydvadanamala manuscripts in Paris, Calcutta and Ky6to. Iwamoto gives a very useful synopsis of these manuscripts (pp. 144-147). It is to be hoped that a thorough study of these materials will shed light on the formation of the text of the Divyavadana. In this respect I should like to draw attention to the fact that the text of the Divyavadana contains several verses identical to verses in K~.emendra's Avaddnakalpalata. In the Ku.n~la legend one verse and a half is the same in the Divyavadana and the Avadanakalpalata (Div. p. 417. 22-27; Av. 59.160cd, 161). Bongard-Levin and Volkova assume that later copyists or editors of the Divyavad6na have taken this verse and a half from the Avadanakalpalata (Legenda o Kunale, Moskva, 1963, p. 90; see also llJ, VIII, 1965, p. 238). Recently Dr. P. S. Jaini has found six verses which are identical in the Divyavadana and the Avadanakalpalata ("The story of Sudhana and Manohar~t", BSOAS, XXIX, 1966, p. 541, n. 41 ; p. 545, n. 53). These verses are not to be found in the Gilgit MS of the Bhai.sajyavastu nor in the Tibetan and Chinese versions. It would be of some importance to examine whether these verses are to be found in the manuscripts of the Divyavadana and Divy~vad6nam6l~ mentioned by Iwamoto. The fourth chapter deals with later avadana literature: the Kalpadrumavadanamald (Kv.), the Ratnavadanamala (Rv.), the Agokavad~namala, the Vicitrakar.nikavadana, the Vicitrakarn.ikavadanamala and the Vratavadanamal~. This chapter contains a lot of interesting information which makes it difficult to summarise. A few remarks must suffice. Both Speyer and Takahata have tried to show the influence of Mahfiy~tna concepts on the avadanamfflas. Iwamoto remarks that although Kv. and Rv. contain Mahfiyana expressions, it does not follow that they are products of Mah~tyfina Buddhism. In this context he draws attention to the Kuo-ch'a hsien-tsaiyin-kuo ching (Taish6, hr. 189) and to the fact that there are already Mah~y~na elements in the Sarv~stivfidin school. I do not find these arguments entirely relevant. Speyer writes: "The authors of the avadanam6l~s were adherents of the Mah~yfina. They told the stories over with mahayana colours and imbued them with rnahay6na concepts" (Preface, p. XXVI). The second part of this statement is undoubtedly correct. It is difficult to prove that the authors were really Mahfiyfina Buddhists but it seems even more difficult to prove that they were Hinay~nists. The presence of Mahfiyfina expressions and concepts makes Speyer's conclusion more likely than Iwamoto's opinion to the contrary. In any case, his remarks are hardly sufficient proof that the avadanamalas (or at least Kv. and Rv. to which Speyer refers) belong to Hinay~na Buddhism. Iwamoto quotes a passage from the Subhfiti story in the Kv. which was published by Speyer in his preface (pp. XXXVII-XCI). As was already indicated by the editor, the author of this text quotes many verses from the Vajras~c~which tradition attributes to Aw For Iwamoto this fact provides a motive for arguing that the Vajras~ci

REWEWS

59

can well have been written by Aw and that the Kv. was composed in the third century. The metrical paraphrases of the Avadana~ataka seem not to have known the fourth varga. F r o m this Iwamoto concludes that the prototype of the Kv. is earlier than the Chinese translation which dates from the middle of the third century. Whilst this is quite possible, it does not imply that the Kv. in its present form is very old. The authors of the metrical paraphrases can very well have known, at a much later period, a different version of the Avaddnagataka or simply a manuscript lacking the fourth varga. Even if the Vajrasftci was written by Agvagho.sa, the Kv. could still have been composed many centuries later, and indeed the style points to a much later date than the third century. I must confess that I am unable to understand fully Iwamoto's argumentation concerning the authorship of the VajrasficL It is difficult to follow Iwamoto's reasoning that because, first, the prototype of the Kv. precedes the Chinese version of the Avaddnadataka and, second, because the text underlying this Chinese version was translated by a Tokharian from Central Asia, where from the third century onwards the Sarv~stiv~din school spread itself, therefore it should be possible that the Vajras~ci could have been written by A~vagho.sa. Neither the prototype of the Kv. nor the text translated by Chih Ch'ien (if he is really the translator!) can have contained lengthy extracts from the Vajrashci. In a short concluding chapter Iwamoto studies the classification and the development of the avaddna literature. He classifies the avaddna literature into four groups: 1. Avaddnagataka and its metrical paraphrases: Kalpadrumdvaddnarndld and Ratndvaddnamdld ; 2. Divydvaddna and Divydvaddnamdld ; 3. A~okdvaddnamdl~ and some avad~na texts which are similar in arrangement; 4. Vratdt,addnamdld. For all four groups he establishes connections with the Sarv~stiv~din and Mfilasarv~stiv~din schools. In this connection Iwamoto remarks that some avaddna texts are in the form of conversations between A~oka and Upagupta (Kv. and Rv.) but others are told by Jaya~ri who relates the stories which Upagupta tells to A~oka. He places Jaya~ri as appearing in the transitional period from the Sarv~stiv~din school to the Mfilasarvastiv~din school. Jayagri appears on the stage in the Agokdvaddnamald, the Gu.nakdra.nd.avy~ha, the Vicitrakar.nikdvaddnamdld, the Bhadrakalpdvaddna (see S. Ol'denburg, BuddijskUa legendy, Sanktpeterburg, 1894, p. 1) and in the Suvar.navar.ndvaddna. F r o m the materials available it is difficult to see why these texts would be doctrinally different from texts such as the Kalpadrumdvaddnamdld, the Ratndvaddnamdld and the

Dvavi .mgatyavaddna. In regard to the development of the avaddna literature Iwamoto distinguishes four periods: 1. The period of the formation of avaddna stories; 2. The period of old prose avaddna literature; 3. The period of metrical avaddna literature; 4. The period of late prose avaddna literature. He subdivides the third period into two: the first including the avaddnamalas in the form of conversations between A~oka and Upagupta; the second including the avadanamdlds told by Jaya~ri. In dating these periods he places the second period in the second and third centuries and period 3a from the third century onward. He claims the third century was the great period of the development of the avaddna literature. This is certainly too early. Speyer has drawn attention to the fact that the Kalpadrumdvaddnarndld quotes the Jdtakamald (Preface, p. XXXVI). One would have to study the texts quoted in the avaddnamdlds very closely, and to distinguish the different styles in which they are written, before making any definite statement about their date. Iwamoto relates the four periods to the specimens of styles, given by Professor John Brough (BSOAS, XVI, 1954, pp. 369-374), but I am afraid that much more research is needed to obtain a satisfactory classification of the avaddna literature according to the styles used by their authors. An appendix contains three specimens of avaddna literature: the Bhavagarmdvaddna (A~okdvaddnamdld, ur. 24), the Kavikurndrdvaddna and the Si.mhalasdrthavdhoddhdrm.la (Gu.nakdra.nd. avy~ha, nr. 16).

60

R~vIEws

On p. 61 Iwamoto quotes a passage of the Tibetan translation of the Sumagadhavadana. In lines 2 and 5 ches sruh and pos sruti must be corrected into che sa sru~ and po sa sru~. In the quotation on page 62 of w68 of the Sanskrit text Sumerubh.rgu.naparives.tita.n is incomprehensible. MS B has bhrigune. Perhaps one must read ~ ~ Professor Iwamoto's book contains a wealth of material. His knowledge of Sanskrit, Tibetan and Chinese enables him to make a much more comprehensive study of the avadana literature than his western predecessors. One may expect that his future publications will bring many new results in this field which has been neglected since the days of Feer and Speyer. If Professor Iwamoto can restrain a certain tendency to draw conclusions which are not fully warranted by the available materials, one may look forward to very valuable contributions to the study of the avad6na literature from him. Above all, I hope that he will give u s editions of unpublished texts. Many avadana texts have been known for a long time from catalogues and brief mentions, but very little has been published, especially from the later avadana literature. The most important of these texts must be published first. Only then can we hope to obtain a better insight into the development of this important branch of Buddhist literature. A.N.U., Canberra

J. W. de Jong

Tilmann Vetter, Dharmakirti's Prama.navinigcayaO. 1. Kapitel: Pratya-

k.sam. Einleitung, Text der tibetisehen Obersetzung, Sanskritfragmente, deutsche Obersetzung ( = Osterreiehisehe Akademie der Wissenschaften, phiL-hist. Klasse, Sitzungsberiehte, 250. Band, 3. Abh. ; Verb'ffentliehungen der Kommission fiir Sprachen und Kulturen Sad- und Ostasiens, Heft 3). Wien, 1966. 111 pp, 6S 148,--. Das vorliegende Werk ist eine Ausgabe des ersten Kapitels des nur in tibetischer (Jbersetzung erhaltenen Prama.navinigcaya.h mit einer deutschen Obersetzung und stellt den ersten Schritt zur Erschliel3ung dieses ffir die Erkenntnis der Lehre Dharmakirtis wohl bedeutsamsten Werkes dar. Der Pram&navini~caya.h ist Dharmakirtis Hauptwerk und enth~ilt seine Lehre yon den Erkenntnismitteln (pramatlam) im Gegensatz zum frfiheren Prarn&navarttikam, wo sic in AnschluB an Dign~gas Prama.nasamuccaya.h vorgetragen wird, in Form einer selbstfindigen systematischen Darstellung? Die Darstellung des filteren Prama.navarttikam hat zwar bei dieser Neufassung des Stoffes fiber weite Strecken als Grundlage gedient - - daher werden in den Prama.navinigcayah, auch zahlreiche Verse des Varttikam und groBe Prosaabschnltte der Svav.rtti.h zum ersten Kapitel fibernommen - - , doch ist Dharmakirti in seinem Denken nicht stehen geblieben. Das Werk bietet daher in seiner Beziehung zum Varttikam einen ausgezeichneten Einblick in die philosophische Entwicklung Dharmakirtis. Von den drei Kapiteln des Werkes, welche die Wahrnehmung (pratyak.sam), den SchluB (svarthanumanam) und den Beweis (pararthanumanarn) behandeln, hat nun Vetter das Kapitel fiber die Wahrnehmung vorgelegt. Damit hat er die entscheidende systematische Darstellung der Wahrnehmungslehre Dharmakirtis zug~inglich gemacht, so dab es nun auch dem nicht auf Dharmakirti spezialisierten Indologen m6glich sein wird, sich start - - wie bisher fiblich - - auf die wenigen S~itze des ersten Ny6yabinduKapitels und einiger g~ingiger Verse des Pram6.nav6rttikam auf die eigentliche Darstellung der Lehre selbst zu beziehen. Es versteht sich, dab Dharmakirti bier, bei der t ~ b e r das literarische Verh~iltnis der beiden Werke vgl. E. Frauwallner, "'Die Reihenfolge und Entstehung der Werke Dharmakirti's", /lsiatica, Festschrift F. Weller (Leipzig, 1954), pp. 142-154.

REVIEWS

61

Behandlung der Wahrnehmung, die wesentlichen erkenntnistheoretischen Probleme bearbeitet, insbesondere das Problem der Erkenntnis eines "/iul~eren" Gegenstandes und des Subjekts der Erkermtnis. Diese Probleme 16st Dharmakirti rnit dem Begriff des Selbstbewul3tseins. Die eine Erkenntnis hat drei Teile - - Objekt, Funktion des Erkennens und BewuStsein des Erkennens - - und aus dieser Vielfalt in der Einheit erscheint ihm das Zustandekommen von Erkenntnis allein erkl~irbar. Die Einleitung, in der Vetter eine knappe historische, die systematische Position des Werkes erkl~irende Einf/Jhrung in die Problematik gibt und den Fortgang des Werkes analytisch nachzeichnet, vermittelt einen klaren ersten Eindruck. Die Obersetzung selbst ist dann durchwegs verl~iBlich, dabei flflssig und verst~indlich geschrieben, was bei der manchmal sehr komprimierten Ausdrucksweise Dharmakirtis eine Leistung f/Jr sich darstellt. Die Lektfire wird zudem sehr erleichtert durch die synoptische Anordnung: Der l]bersetzung stehen der tibetische Text mit den Varianten-Apparat und die Sanskrit-Fragmente, soweit sie bisher bekannt sind, gegenfiber. Die Anmerkungen hat Vetter auf das Notwendigste beschr~inkt, das zum Text und zur Interpretation gesagt werden muBte. Dem tibetischen Text liegt die Obersetzung von Parahitabhadra und Blo ldan ges tab in den Ausgaben von Derge, Narthang und Peking zugrunde. Ebenso wurde zur Textherstellung die Ubersetzung der .Tika Dharmottaras - - in der Ausgabe yon Peking herangezogen, da sie von denselben Obersetzern stammt. Die Pratikas aus der Obersetzung von J~nagdbhadras T.tk~ wurden, da diese von anderen Obersetzern bearbeitet wurde, nicht herangezogen. Die iiblichen dialektischen Unterschiede von Derge gegenfiber Narthang und Peking (z.B. ltos : bltos etc.) wurden in den Variantenapparat nicht aufgenommen. Alles in allem ein grol3er Fortschritt in der so notwendigen ErschlieBung eines der bedeutendsten Philosophen, die Indien hervorgebracht hat. Daher w/ire zu wOnschen, wenn dieser Band auch in nichtindologischen, an indischer Philosophie interessierten Kreisen ein Publikum finden k6nnte. -

-

Wien

Ernst Steinkellner

Ernst Steinkellner, Dharmakirti's Hetubindu.h, Teil I: Tibetischer Text und rekonstruierter Sanskrit Text; Teil II: Obersetzung und Anrnerkungen ( ~ Sitzungsberichte der Osterr. Akademie der Wissenschaften, Phil. Hist. Klasse, Band 252, Abhandlung 1 und 2; Ver6ffentlichungen der Kommission fiir Sprachen und Kalturen Siid- and Ostasiens, Heft 4 und 5). Wien, 1967. GroBoktav, brosch., 115 Seiten, iSS 132,-- und 220 Seiten, 5S 212,--. The two volumes under review represent another important publication which is an outcome of the profound studies of Dharmakirti's works and theories done by the Vienna indological school. By editing the Tibetan translation of the Hetubindu.h, reconstructing its Sanskrit text (vol. I), adding a German translation and many exhaustive notes (vol. II) Steinkellner makes accessible probably the first and, at the same time, most important work in the history of Indian philosophy which exclusively deals with logic. The first volume is mainly devoted to the reconstruction of the Sanskrit text by help of the quotations in Arcat.a's Hetubindut.ika (Gaekwad's Oriental Series, CXIII) (abbr. HBT) and the Tibetan translation which is added in a critical edition. These together with the quotations in the epistemological literature of the Jainas serve as an excellent basis for such an undertaking. There has already been a reconstruction done by R. Safik.rtyfiyana (printed in HBT,

62

REVIEWS

parigi.st.dni, pp. 52-72), but Sfifikrtyfiyana has in many places not taken advantage of the excellent material available working carelessly and therefore producing an unreliable text. This may be illustrated by two characteristic examples taken at random. He reconstructed, e.g., the following sentence of the Hetubindub on p. 65,7 (HBT):

uktam atra yathd paryudasavrttyd apeksdta.h tadvivikto 'rthas tajj~dna.m vd 'bhdvo 'nupalabdhig cocyata iti. The underlined words are not to be found in the Tibetan translation (vol. I, p. 72,18). If we look for a possible source of these words there is a quotation in HBTp.175,21-22 that is explicitly ascribed to Iw (cf. vol. II, p. 164 for the interpretation), which S~fik.rtyfiyana obviously failed to notice. On p. 68,12 (HBT) Sfifikrty~yana has the following sentence: tatra kdra.navyrpakayor api abhdvavyaval~ra.h tadanyabhdvasiddhyaiva sidhyati. Here Sfifilff.tyfiyana keeps to the Tibetan translation (vol. I,p. 84,9). What is typical for his procedure, however, is that he omits a word that must have seemed unimportant to him, namely rio bo ~id. Furthermore he makes no use of the Sanskrit text ofHBT, p. 202,24-25 and so arrives at a text, that might be an equivalent to the Tibetan, but hardly represents the original. These two examples may have shown perhaps how many of the most elementary philological principles have been violated by this first reconstruction. Steinkellner, o n the contrary, establishes a truly reliable text. He never looses sight of the Tibetan translation and also makes much better use of the words that are interspersed into the text of the HBT. He justifies the smallest deviation of his reconstruction from the Tibetan in the foot-notes and marks by italics every passage and wordformation that is not supported by a Sanskrit text. This is an exceedingly diligent and fine study, and the attempt to reconstruct the original Sanskrit text of the Hetubinduh. must be called successful as far it is possible at present. The second volume gives a German translation which is meant to prove that the reconstruction makes sense (vol. I,p. 23) - something that should go without saying, yet is rarely observed. A t the same time the translation renders accessible the contents of the Hetubindu.h to non-indologists as well. The same purpose is served by the extensive notes (sometimes assuming even the form of small essays), which hardly leave a n ~ h i n g unexplained and, above all, place the sentences of the text in their proper historical context. Steinkelhier has deliberately desisted from giving a metalinguistic description of Dharmakirti's logic for - as the author rightly remarks (vol. II, p. 5) - it is the primary task of any work on Dharmakirti to make accessible the material. When this is done in such an objective way as in Steinkellner's book also the logician who does not know Sanskrit can get a clear idea of the text without having to be afraid that the material has already been distorted by being interpreted according to a presupposed system. Finally it should be noted that the whole text of the translation is accurately divided into paragraphs and well analyzed. Digressions of a more factual kind are set off. Among these especially digression b. "The cognition of concomitance in the inference of the momentariness of things" might be of a more general interest. Utrecht

T. Vetter