The remarkable and large anthology of epigrammatic single verses known as “Hāla's 700” epitomizes all the grave and gay aspects of love in a rural Indian environment. The ancient and rustic Maharashtri Prakrit language in which they are couched was made accessible through the sterling work of Albrecht Weber between 1870 and 1883, but many philological problems remain. There has been no previous full-scale attempt to render them into English verse, but a new translation by Peter Khoroche and Herman TiekenFootnote ¹ has often succeeded in producing accurate and attractive renderings in elegant free verse, rearranged thematically, and competently and succinctly explained in prefatory notes.

The translation has, however, its shortcomings. In the following example of a typically colourful twist on one of the more hackneyed themes (v. 305 = AW [i.e., A. Weber] v. 496),Footnote ²

the translators' rendering makes the point more forcefully than any strictly literal prose presentation; but their wilful transposition of the verse into the past tensed, clashes awkwardly with the retained imperative.

In another verse (54, AW 318):Footnote ³

It may not matter that we lose something of the direct comparison between an unfindable rendezvous and a forgotten hiding place. Again, however, the notion of arranging a meeting in “all those reed-beds” is odd. In fact, she is searching “here and there among the canes” (vāṇīravaṇammi).

For the vocative māmi, where the maiden is confiding her romantic adventures, neither “Aunt” (passim), “O auntie!” (178, AW 344), nor “Mother” (43, AW 246 and 465, AW 124: tacitly reading *māĕ?) seems appropriate. Even if a youngish relative is implied, the contexts suggest that māmi is no more than a prosodically conditioned substitute for sahi, and might be translated as “dear friend” or the like (PSM: māmi “sakhī ke āmantraṇ mẽ prayukt avyay”).Footnote ⁴

Another linguistic problem underlies 160, AW 271:Footnote ⁵

The number of cows and eyes might have been made to correspond.

In all these, more attention ought to have been paid to existing translations. The explanation given for another verse (53, AW 365)Footnote ⁶ is that “the boy does not show up or is late”, but the intimation is that he arrives on cue: “(she) awaits” may adequately render diṇṇasaṃkeā “she has made an assignation with”, but “Listening for the rustle of dead leaves / Stirred by his footfall” does not do justice to āaṇṇei and aggapaa- (she hears his cautious approach, tiptoeing through the leaves). Here, as elsewhere, one misses Weber's succinct explanatory caption: “Da kommt er!”.

The minimal annotation does not cater for a case like 372, AW 26,Footnote ⁷ where the translation:

tends rather to imply a woman reproaching an errant and impenitent male, in keeping with the masculine participles of the Vulgate; but the feminine readings of the Jaina and Kerala recensions are listed as having been adopted: ṇa kuṇanti ccia (māṇaṃ) and jāṇantī. This, and the discussion in Tieken's Hāla's Sattasaī, 1983 (HS), 230 ff., show that we are meant to envisage a third party's advice that the woman must make the best of things lest worse befall. In fact, past conditional “you wouldn't have, if you had known” makes rather better sense: it serves her right, he's gone (as in AW 129).

A duly corrected version of Weber's translation of 310, AW 909Footnote ⁸ also requires some elucidation:

We are presumably supposed to understand from this that he merely teases the girl thus, those unmistakable big eyes being hers (since an equivalent phrase eddahamettehi acchivattehiṃ in AW 973 refers, as one would expect, to a girl's outstanding beauty). Nothing is gained by transposing the habitual present (gottakkhaliehi … kiṃ va ṇa pecchai aṇṇaha “he can't really be making such mistakes, can he?”) into an ambiguous “could”, or indeed by separating the verse from its pendant 306, AW 908,Footnote ⁹ where the notion of deliberate teasing is explicit.

The maverick Telinga-Kerala readings listed in HS, 35–7, have generally, and rightly, been passed over.Footnote ¹⁰ Surprisingly, however, the clear instances listed in HS, 37f., of correct Bh. readings that are retained largely or only in Kerala manuscripts have been ignored in favour of Weber's Vulgate distortions.Footnote ¹¹ A dozen examples of alleged “common errors” in Bh. and Kerala (HS, 43) could rather be discounted as mere perseverance of orthographic vagaries (e.g. Bh. garuyaṇa “parents” versus AW 367 guruaṇa), except for Bh. sirīsatala, a better reading than AW 49 sarīratala. Neglect of the evidence of Bh. can lead to a struggle to read meaning into a text to which the key has been lost, as in 408, AW 219:Footnote ¹²

The need to provide a non-existent “but” indicates that the sense is rather “I assure you, foolish boy, that spring can do anything: the scent of amaranth has made me quite unfaithful”: irrespective of whether the Vulgate reading is maṇammi / manasi or maṇaṃ / manāk, its commentators (with śīlaṃ khaṇḍitam, etc.) variously accept an admission of infidelity. The explanation may lie in an Apabhraṃśa pun, whereby maṇammi can convey, besides “only mentally” (the deprecatory Leśa figure of the Bh. commentary), a positive incitement maṇaṃ pi. Only thus does the provocative address vālaa “foolish boy” (T: jāraṃ d ḍhānurāgāya protsāhayati) make obvious sense.

A set of readings that were dismissed in HS, 42, as innovations peculiar to the Jaina recension (Bh. and R) have also been ignored, although they are defensible and even preferable. The 26 “variant readings” that have been adopted for the translation are listed without apparatus or comment (p. 205), and they too seem justified only to the extent that they are supported by Bh. and R. An inverted bahuvrīhi reading hiaaṃ māṇapautthaṃ “pride has left her heart” that has been preferred to prevailing hiaaṃ pautthamāṇaṃ “h dayaṃ proṣitamānam” (343, AW 188) is convincing only because, as was shown in HS, 239, it is based on the combined evidence of Bh., Kerala and Daśarūpa. For 202, AW 759,Footnote ¹³ the opaque Jaina reading paṃḍiaittaṃ is tentatively adopted and paraphrased (following Bh. “paṇḍitamanyam”) as “If you are the scholar / You say you are”; but the simple emendation paṃḍia itt[h]aṃ would still yield the appropriate irony and better explain the variant readings (R paṇḍiya ittaṃ, S paṇḍiaṃ ṇiccaṃ, T paṇḍia iddhiĕ/itthiĕ). Patwardhan's edition has confirmed the Jaina readings that have been adopted in the case of AW 114 sirigoviāĕ, 226 -suttāe 692 ahilijjai, 748 ne[v]a tāṇã khayamaggo, and 788 vaṇa-. Though absent from Bh., other confirmation exists for 2 taṃta- and 651 jāi (misprinted on p. 205 as jai); and 949 kaaveṇiāṇa is fully justified by both sense and metre.

Other proposed readings are far from convincing. For 334, AW 488,Footnote ¹⁴āṇāvarāhĕ kuvio is no improvement on *āṇāvarāhakuvio, which was Tieken's conjecture in HS, 84. The verse is translated as:

The proposed reading pesattaṇāvarāha- “one's being servile” for vesattaṇāvarāha- stems from a palpable misreading in the Bh. commentary (not retained in Patwardhan's edition), and *āṇāvarāha- “one's being bossy” was Tieken's consequent conjecture in lieu of aṇṇāvarāha-; but the attested forms (equivalent to anyāparādha “the faults of others” and dveṣyatvāparādha “the fault of being unloved”) are as convincing in the context as the proposed forms (preṣyatva “servility” and ājñā “bossiness”) are implausible.

For 353, AW 420, the Telinga reading paravasaṇa-ṇaccirī-, adopted and rendered as “laughing at someone else's misfortune”, is likely to be condoning a lectio facilior, since an attractive pun “dancing to another's sins/tunes” is implied elsewhere (par'avajja- in Bh., para-vādya- in G). The same may be said of redundant āligasi “you are embracing” for 369, AW 33:Footnote ¹⁵

The “embracing from behind” ousts the preferable explanatory phrase in Bh. sayaṇaddhe kīsa me parāhuttiṃ “why, when I am turned away on my side of the bed”; and, engendered by the Kerala-Telinga finite reading palīviaṃ (tacitly adopted instead of gerund palīviuṃ), it breaks the verse uncharacteristically into three sentences.

A (Kerala?) reading ṇa daṃsase piṭṭhaṃ for 384, AW 943, rendered as [most politely, you] “never turn your back on me”, is banal compared with the irony of Weber's ṇa jampase piṭṭhaṃ “sprichst nicht einmal von dem, wonach du gefragt wirst!” In any case, “back” is otherwise paṭṭhiṃ, puṭṭhiṃ in the text, and not piṭṭhaṃ. Even support from Kerala alone would hardly justify preference for sirivandīṇaṃ as “royal hostages” over AW 54 sarivaṃdīṇaṃ (“fellow-prisoners”), oraṃtamuhīe “with tears streaming down her face” over 539 oruṇṇamuhīe, or uppakka “clot” over normal uppaṃka, the Bh. reading for 586. The graph āsaṃdhia- “intended” (as read by Weber also in Bh. 389: Ind. St., XVI, 135) seems preferable to āsaṃghia- understood as “reached” (76 and 707). An awkwardly placed negation kae ṇa has been introduced into 473 (Jaina kaeṇa, not kae na). It is not clear why a (Kerala?) reading phalahī is recommended for 550, since the rendering remains “bolt”, in keeping with Jaina phalihaṃ. The Jaina readings hayasuṇhāĕ taha kao “(my son) has been so ruined by his atrocious wife” and aṃgacchittaṃ miva vammaheṇa “(given) by Eros as a personal gift” (Patwardhan, pp. 282, 277) seem preferable to the Vulgate variant selected in the former case, 632, and to the violent and improbable alteration imposed upon the latter, 782.

A brief introduction gives a good account of the background, but it represents (p. 10) Weber's edition as containing “no fewer than 964 poems, of which only 430 are common to all versions”. These are indeed Weber's figures (AW, xlviii): Tieken's own study of the Kerala recension brought the total to 989, however, of which (judging by HS, App. 2) hardly more than 400 are ubiquitous. Conveniently, for the sake of presentation, the 964 are listed (p. 207 ff.) as 9⅔ separate “centuries”, but with no indication that they represent the seven arbitrarily separated “centuries” of Weber's Madhyadeśa Vulgate, followed by miscellaneous sets of progressively less “well” attested verses. By making their own personal selection of 701 verses, the translators have created yet another recension, one that corresponds to a degree with Weber's Vulgate: it admits only 17 of the 52 stanzas that are peculiar to the two Telinga texts and none of those that are peculiar to the Kerala.

It seems, however, highly improbable that any one single 700-verse archetype ever existed, although the recensions have variously striven to reach this total. In his evaluation of the Kerala recension, Tieken did not question Weber's assumption of an original 700-verse compilation, emending it only to the extent of granting equal authority to the constitution of the Vulgate and Jaina versions (HS, 41). His explanation, however, that the Kerala text (and ultimately the Telinga) was forged from some full-scale Vulgate-Jaina compromise version was hardly plausible. It is not likely that the compiler of the Kerala would go through this text picking out less than half its verses, more or less in sequence; go through it again for a second collection, drastically re-shuffled this time; and finally restore the total by strewing in other and apocryphal verses throughout the whole. The initial trawl, thus postulated, would have dismissed many verses that subsequent compilations, including the present one, preferred to retain.

Rather more credible is a much smaller archetype, comprising the bulk of those that are in consistent sequence, compactly within the first 425 verses in the Kerala, but dissipated throughout the 700 in Vulgate and Jaina, and further jumbled up in Telinga. The factitious title “Hāla's 700” is after all in competition with non-committal labels like Gāthākośa, a name that applies also to the 156-verse collection published under the title Chappaṇṇaya-Gāhāo by A. N. Upadhye (Kolhapur, 1970). The latter (literally “56 verses”) is a name so closely associated with the genre, especially via its palpable Sanskritization Ṣaṭprajñādi-gāthāḥ, as to suggest that a 56-verse compilation of similar material has long since been subsumed within the longer version.

The Gagādhara Vulgate verses AW 188–95 occur earlier and more widely scattered in the Jaina recension (in Bh. as 29, 34, 45, 104–5, 123, 30, 140; in R among its vv. 31–127) and in the Kerala (among its vv. 69–84 and 173–5). This seems to be not so much a clear indication of arbitrary “transposition in the Vulgata” (cf. HS, 161), but rather a symptom of deliberate culling of supplementary Jaina material, on the part of the “Madhyadeśa” Vulgate. The intention would be to complete a further century, for AW 188–95 correspond to 191–9 in another Vulgate manuscript and most of them correspond to 192–9 in a third. The manuscript χ, however, had achieved a total of 196 before drawing on any of the verses in question, and so used only three of them to reach 199, three that form a similar compact group only in R. Comparable inflation of the Kerala offshoot is to be observed: most of the thirty-odd verses that in Kerala intervene between AW 19 and 20 (Bh. 17 and 18) recur elsewhere in the seventh and final century, somewhat similarly arranged in the Jaina recension (R), but randomly dispersed in the Vulgate. Expansion using supplementary Jaina Maharashtrian material seems again the likely hypothesis.

The ubiquitous near-initial verse (1, AW 3):

is rendered tendentiously as “Among countless elegant poems,/King Hāla, patron of poets,/Has selected seven hundred”.Footnote ¹⁶ “King” is indeed the implication of the term “patron of poets” and of the Telinga reading Sālaiṇa, this being a compressed attempt to introduce the Sālāhaṇa-ṇarindo mentioned in AW 467.Footnote ¹⁷ The rendering “selected among countless” recalls Weber's implication of merely editorial activity: “mitten aus einer Unzahl (hier) zurecht gestellt”.Footnote ¹⁸ Since, however, the compound virac- is always at most perfective, like vinirmā-, the preverb would not affect its regular sense of original composition. Bhuvanapāla's gloss on it, prakhyāpitāḥ, was not intended to imply “published by Hāla”: he presented the qualification “patron of poets” as confirmation that the stanzas could rightly be ascribed to Hāla (tadīyā eva gāthāḥ prakhyāpitāḥ).Footnote ¹⁹ The notion of selection is not justified even by the explicit ablative majjhaārāo that is read by Bh., Telinga, and Kerala and was adopted in HS, 187, as implying “collected from among”. The contrast between Kalhaṇa's ablatives tebhyo d ṣṭaṃ catuṣṭayam … abravīt tān dvāpañcāśato madhyāt “four of them are seen … he mentioned those (five) of the 52” (Rājat. 1.16, 19) and Kālidāsa's locative jahāra tayor madhye “he took (her) from between the two” (Ragh. 12.29) suggests that the ablative has no semantic significance, and that the Vulgate locative majjhaārammi may reflect a preference for Mahākāvya elegance.

Especially since it was prefixed also to a Telinga text comprising only one century of such verses, there is no reason to believe that this Hāla verse was meant to be a description of the contents of the work in question. It is not an unnumbered final colophon, but forms part of a prefixed apologia for vernacular poetry: it is classified as jayakāra (in Telinga) and namaskāra (in S). The sense is that the vast legacy of Prakrit literature had been enhanced by Hāla's contribution. It is, like Bāṇa's similar verse, not a compiler's recommendation of his own work, but a reference to the established fame of a predecessor.

The manuscript colophons have been slow to adapt to the notion of Hāla as sole author. Only the Gagādhara version of the Vulgate has an actual final colophon verse that introduces the fiction of Hāla's authorship (AW 698 ia siriHālaviraie …). It is included in the numbering, presumably because Vulgate versions do not always manage to reach a complete seventh “century”. The R recension evidently reached the 700 mark independently, since most of its non-Vulgate makeweight verses are clustered in the final century, just as are most Vulgate-only verses. After its v. 705, R has a colophon verse, unnumbered according to Weber, which differs in substance from AW 698 only in attributing the verses, not to siriHāla, but to unidentified poet(s), sukai. Being ambiguous as to number, the R colophon is compatible with the view of several recensors who have tried to provide different authors' names for each verse (some palpably guessed from the context). It runs:

The curious wording tends to suggest that plural authorship is indeed the intention of the colophon. Its application of the regal epithet “patron of poets” to unnamed sukai shows that it has adapted the expression kaivacchaleṇa Hāleṇa of the integral verse AW 3, pluralizing it in keeping with that verse's actual intimation that Hāla was not the only eminent Prakrit poet (koḍīe majjhaārammi). An equally non-committal colophon verse (ettha cautthaṃ viramai gāhāṇa saaṃ …), anomalously occurring, largely unnumbered, after the fourth century of the Vulgate presumably marks the original ending of an as yet anonymous 400-verse recension. Since, unlike all the other colophon verses, it involves no false inference from AW 3, it can have some claim to authenticity.

The verse R 705 that immediately precedes its colophon happens to close the sixth century in Gagādhara's Vulgate (8, AW 599). This must be why the R colophon has been incorporated, and again numbered, in most Vulgate manuscripts, not after the seventh century, but mechanically after the sixth (suitably modified: saṭṭhaṃ gāhāsaaṃ). It recurs sporadically (after the first century in R and also elsewhere in the Vulgate), but it is not found at all in available centuries of the Kulanātha Vulgate or in any other non-Vulgate text. The fact that the Gagādhara Vulgate has its own distinct final colophon seems to be further confirmation that R and the Vulgate texts independently achieved the total of 700 verses that was thought to be promised at the outset in AW 3.

Even if there is thus no compelling reason why Khoroche and Tieken should have made, specifically from Weber's 964, yet another, and quite arbitrary, selection of 701 verses “following our own judgment and interpretation”, their anthology is a most necessary addition to the scanty materials available in English. Arvind Krishna Mehrotra's work, The Absent Traveller (London: Sangam Books, 1991), not mentioned in their bibliography, was a poetic English paraphrase of over 200 of the Sattasaī verses; but a complete English version that would match the lucid accuracy of Weber's German prose has yet to be achieved. Although the volume leaves the Prakrit original texts “buried in the learned publications in which they first appeared”, its publication should prove a welcome stimulus for further philological investigation of a significant and enthralling component of world literature.

Abbreviations

AW:

Albrecht Weber, Das Saptaśatakam des Hāla (AKM, VII, 4), Leipzig, 1881.

Ind. St.:

Albrecht Weber, “Über Bhuvanapāla's Commentar”, Indische Studien, XVI, 1883, 1–204.

Patwardhan:

M.V. Patwardhan, Hāla's Gāthākosa (Gāthāsaptaśatī) with the Sanskrit Commentary of Bhuvanapāla (Prakrit Text Series, 21.), Ahmedabad, 1980.

HS:

Herman Tieken, Hāla's Sattasaī: Stemma and Edition (Gāthās 1–50), with Translation and Notes, Leiden, 1983.

Bh., R:

Bhuvanapāla, Raivāsā: the Jaina recension.

K, γ, ψ, χ:

Kulanātha and other versions akin to Gaṅgādhara's Vulgate recension.

S:

Sādhāraṇadeva's recension.

T:

the “first” Telinga recension.