Part II. The Author in Context
3. Clues in the Devīcandragupta
The Date of the Devīcandragupta
The age bracket open for the Devīcandragupta is slightly narrower than that avail-able for the Mudrārākṣasa. It certainly does not predate Candragupta II whose exploits against the Śakas and contention for the throne of his brother Rāmagupta provided its plot. Its terminus ante quem cannot be established so clearly, but a sentence in the Harṣacarita of Bāṇa (first half of the 7th century) is very likely to refer to its plot. Here Harṣa’s commander of elephant troops warns the king to be vigilant all the time, and rat-tles off a long list of rulers killed in unlikely circumstances or by unlikely assailants. These include a chieftain of the Śakas, who was “desirous of another’s wife” and whom “Candra-gupta, hidden in the guise of a woman, slew.”266
Now it is of course possible that Bāṇa had access to a tradition independent of (and possibly previous to) Viśākhadatta, from which he derived this snippet. However, among extant records the detail that Candragupta dressed as a woman to assassinate the Śaka king is only preserved in the Devīcandragupta and the Mojmal al-tawāriḵ, the relevant part of which is probably based on this play.267 Beside the literary tradition, there are two Rāṣṭrakūṭa inscriptions of the ninth and tenth centuries CE that refer to the episode of Candragupta killing his brother and marrying his wife. One268 teasingly mocks, the other269 sharply scorns the deed. Both, of course, conclude that the current Rāṣṭrakūṭa ruler is a better man than the famous Gupta monarch, yet neither mentions the saucy bit about cross-dressing,270 which sounds more like a playwright’s fancy than actual history. The im-plication of one inscription that Candragupta “willingly did an abominable thing contra-dictory to [the accepted notions of] purity and impurity”271 might be taken as a reference to his transvestitism, but it is much more likely to be a hint at another tradition (also found
266 Harṣacarita p. 51 (end of ucchvāsa 6), aripure ca parakalatrakāmukaṃ kāminīveṣaguptaś ca candraguptaḥ śakapatim aśātayad iti.
267 SOHONI (1981:171n10) argues that the story of Ravvāl and Barkamārīs must have been based on independent historical materials rather than the play, because it says at the end of the story that “The power of
Barkamárís and his kingdom spread, until at length all India submitted to him” (ELLIOT 1867:112), which could not have been there in a play written and performed soon after the events dramatised in it. In my opinion this argument holds no water: even if the Devīcandragupta was produced in the early period of Candragupta’s reign, the “epilogue” could have been added by the author of the Mojmal al-tawāriḵ or of the history from which he borrowed the account.
268 The Sanjan plates of Amoghavarṣa I, dated Śaka 793 (BHANDARKAR 1926:248, verse 48 in lines 51–53), hatvā bhrātaram eva rājyam aharad devīṃ ca dīnas tato lakṣaṃ koṭim alekhayat kila kalau dātā sa guptānvayaḥ. (Text cited with Bhandarkar’s emendations.)
269 The Cambay plates of Govinda IV, dated Śaka 852 (BHANDARKAR 1903:38, verse 22 in lines 26–27), sāmarthye sati ninditā pravihitā naivāgraje krūratā bandhustrīgamanādibhiḥ kucaritair āvarjjitaṃ nāyaśaḥ| śaucāśauca-parāṅmukhaṃ na ca bhiyā paiśācyam aṅgīkṛtaṃ tyāgenāsamasāhasaiś ca bhuvane yaḥ sāhasāṅkobhavat||
270 The idea was dismissed as a mere “scandalous tradition” by Vincent SMITH (1914:292) when it was only known from the Harṣacarita.
271 śaucāśauca-parāṅmukhaṃ … paiśācyam aṅgīkṛtaṃ, in the Cambay plates quoted in note 269 above.
in the Devīcandragupta) according to which Candragupta used black magic to enlist the aid of a zombie (vetāla), or alternatively at his pretence of (ritually impure) madness to kill his brother.272 It is thus possible that Bāṇa referred to the Devīcandragupta, while the Rāṣṭra-kūṭa copywriters knew only a more factual (or, if the cross-dressing did happen, a more demure) history.
Śaṅkarārya, who wrote a commentary on the Harṣacarita in the early 18th cen-tury,273 may also have had access to the Devīcandragupta or to a quasi-historical work based on it, since he adds a snippet which Bāṇa does not mention, but which is found in the story of Ravvāl and Barkamārīs (and thus probably in the Devīcandragupta too), namely that Candragupta did not perform his gambit alone, but had associates who also dressed up as women.274 Another detail found in this commentary and the Devīcandragupta but not in the Harṣacarita is the name of Dhruvadevī. This agrees with actual history as far as we know it, so it does not directly prove that Śaṅkarārya knew the Devīcandragupta, but it does lend some strength to the hypothesis based on the reference to the other men dressed up as women.
Within the range between 400 CE or so and 600 CE or thereabouts, it is at present impossible to determine a more precise date for the Devīcandragupta. Some scholars have attempted to date the play relative to the reign of Candragupta II, but they reached widely varying conclusions starting out from pretty much the same point: the dramatization of Candragupta’s scandalous actions. Thus JAYASWAL (1932a:35–36)—while reasserting his ear-lier theory that Viśākhadatta was a contemporary of Candragupta II275—opined that “the drama must have been written in the time of [Candragupta’s] son and might not have been published by the author in his own life-time,” primarily because it has a love scene be-tween Candragupta and a courtesan. Quite to the contrary, WINTERNITZ (1936:360) re-marked that before the discovery of the Devīcandragupta he had been “inclined to agree with those who would assign the Mudrārākṣasa to the period of Candragupta II. But it is not likely that Viśākhadatta would have written the Dēvīcandragupta, a drama in which
272 See DEZSŐ 2010:403 for more details about Candragupta’s vetālasādhana. RAGHAVAN (1978:864) notes that as far as the relevant fragment of the Devīcandragupta indicates, Candragupta did not actually perform this rite. SOHONI (1981:178) vehemently opposes any such slander about Candragupta, whom he seems to regard a national hero, and says that “there has been a good deal of consciously and unconsciously held malicious and erroneous association of ideas in regard to the alleged vetāla sādhana of Candragupta Vikramāditya,”
claiming it was “unthinkable that a hard headed and carefully trained prince like Kumāra Candragupta should have thought of experimenting in black magic or spirit control,” and the idea of doing so is
“mentioned only to be given up immediately” in the Devīcandragupta, a drama “hewn by Viśākhadatta from the rock of pure patriotism” (ibid. 187).
273 GUPTA 1974:148.
274 The commentary to the sentence of the Harṣacarita cited above in note 266 says, śakānām ācāryaḥ śakādhipatiḥ. candragupta-bhrātṛ-jāyāṃ dhruvadevīṃ prārthayamānaś candraguptena dhruvadevī-veśa-dhāriṇā strī-veśa-jana-parivṛtena vyāpādita iti (as quoted in BHANDARKAR 1932:206). In the Mojmal al-tawāriḵ,
Barkamārīs says, “let an order be given for me to be dressed like a woman, and let all the officers dress their sons in like manner as damsels, and let us each conceal a knife in or hair … All the officers of the [enemy] army will thus be slain” (ELLIOT 1867:111).
275 See page 33.
Candragupta marries the wife of his elder brother murdered by him, at the lifetime of this king, or even of Kumāragupta … Thus, the Dēvīcandragupta, as far as we know it at present, would support the sixth century as the date of Viśākhadatta.”
On the other hand, DAS GUPTA (1938:217) theorised that Candragupta’s murder of his brother and his subsequent marrying of the latter’s widow would have been “neither morally wholly indefensible, nor socially illegal,” in which case “there exists no cogent reason why the poet could not write the drama during the lifetime of the king whom it celebrates.” He bases this claim on a hypothesis originally put forth by BHANDARKAR (1932) in an article about historical data that may be gleaned from the Devīcandragupta and the Mojmal al-tawāriḵ. According to this, Rāmagupta must have been seen as impotent because of his inability to resist the Śaka ruler’s demands for the handing over of his wife. This
“impotence” rendered his queen suitable for remarriage in accordance with the Nāradasmṛti.276 DAS GUPTA’s additional arguments include a statement in the Mojmal al-tawāriḵ that the queen (that is, Dhruvadevī), when still a maiden, had actually chosen Barkamārīs (i.e. Candragupta) for her husband, but the elder Ravvāl (Rāmagupta) took her for himself.277
Apparently neither Das Gupta, nor Jayaswal find a problem with Candragupta’s murder of his reigning elder brother. Such a deed, though possibly standard practice in royal circles, was in my opinion unlikely to have been a feat to be flaunted in a play in-tended to flatter Candragupta. We do not, however, know what justification (excuse) the Candragupta of the Devīcandragupta has for fratricide.278 The drama may after all have been written precisely with the aim of showing that the new king had every reason for this act and indeed, did the realm a favour by eliminating his brother.
Is It by the Same Author?
The author of the Devīcandragupta is declared in only one of all the known cita-tions—a passage of the Nāṭyadarpaṇa treating on the heroine of a prakaraṇa—where his name is said to be Viśākhadeva.279 Nevertheless the scholarly consensus is that this drama
276 BHANDARKAR (1932:201–204). The basis of Bhandarkar’s reasoning is a verse of the DCG (see note 280 on page 84 for text) which says Rāmagupta “behaved like a eunuch (klība) even though he was a man” and also uses in connection with the queen the word kṣetra, which may be a technical term for a wife as the “field” in which a man sows his seed. While Rāmagupta may or may not have been deemed impotent, this interpretation of the verse is in my opinion forced: the mental states of lajjā, “shame,” kopa, “anger,”
viṣāda, “despair,” bhīti, “fear” and arati, “distress”—prominently featured in the verse next to kṣetrīkṛtā (for which see also page 84)—have nothing to do with the legal conditions for a woman’s remarriage.
277 See ELLIOT 1867:110 for this part of the narrative. The story sounds suspiciously like post-hoc legitimation for marrying the widowed queen, invented by the author of the DCG or of the Mojmal al-tawāriḵ.
278 The Mojmal al-tawāriḵ, or at least Elliot’s report of its account, is strangely silent on this.
279 ND p. 118 (near beginning of viveka 2), yathā viśākhadevakṛte devīcandragupte. The same stanza that follows this ascription is also cited in a similar context in the Abhinavabhāratī and the Śṛṅgāraprakāśa, but in those works the author is not named, and the names of the characters involved in the cited verse are different (GUPTA 1974:165n55). Note also that although the edition of the Nāṭyadarpaṇa (SHRIGONDEKAR &GANDHI 1929) reads viśākhadeva here and RAGHAVAN (1978:846) also clearly says that “the author of the Devīcandragupta …
was indeed written by the same person who wrote the Mudrārākṣasa. While it is entirely reasonable to presume that Viśākhadatta and Viśākhadeva were one and the same person, a single ascription in a fair-sized corpus of anonymous citations of the Devīcandragupta is not entirely convincing.
There are a few thematic and idiomatic similarities between the two plays that may be viewed as corroboration of the identity of their authors, though such details can by no means be viewed as solid proof. RAGHAVAN (1978:862) pointed out that the phrase lajjā-kopa-viṣāda-bhīty-aratibhiḥ kṣetrīkṛtā280 in one of the verses cited from the Devīcandra-gupta is reminiscent of vañcanā-paribhava-kṣetrīkṛto in the Mudrārākṣasa.281 The noun-verb compound kṣetrī-√kṛ is sufficiently rare in Sanskrit literature282 to believe it may be an id-iosyncrasy of our poet, though calling it “a favourite expression of Viśākhadatta”
(RAGHAVAN, ibid.) on the basis of one proven and one suspected occurrence is perhaps going too far.
BHANDARKAR (1932:203) noticed the word in the Devīcandragupta for another rea-son, interpreting it to mean that Queen Dhruvadevī, whom it qualifies in the stanza, has been rendered “fit for being used by a stranger.”283 In my opinion kṣetrīkṛtā in this verse simply means “become subject to” the list of negative emotions (in a dvandva compound with an instrumental ending) that precede the word, and that is precisely the meaning in which it is used for Rākṣasa’s person in the Mudrārākṣasa stanza, there also preceded by a list of negative emotions (in a dvandva compound, in this case also compounded to kṣetrīkṛto). Notice also that both occurrences of the word are in the same metrical position of a śārdūlavikrīḍita verse.
Another uncommon usage found in both the Mudrārākṣasa and the Devīcandra-gupta is the word praṇayin, literally “fond of,” but applied to jewellery in the sense “worn on.” The former play describes Rākṣasa’s seal-ring as “attached to” his finger, while the latter has a chain girdle “attached to” the hips of the courtesan Mādhavasenā, again in identical metrical positions of a śārdūlavikrīḍita stanza.284 The use of astronomical imagery
has invariably been cited as Viśākhadeva”(emphasis in original), some scholars (e.g. DHRUVA 1930:271 and BHANDARKAR 1932:209) quote this locus of the ND as yathā Viśākhadattakṛte devīcandragupte. Trivedi’s translation of the ND also says “in the play Devīcandragupta of Viśākhadatta” (TRIVEDI 1966:82). It is not clear if any of these scholars have seen manuscripts of the ND that support this reading, or have simply made a mistake.
280 ND p. 86 (example of the saṃdhyaṅga called krama in viveka 1), ramyāṃ cāratikāriṇīṃ ca karuṇāśokena nītā daśāṃ tatkālopagatena rāhuśirasā gupteva cāndrī kalā| patyuḥ klībajanocitena caritenānena puṃsaḥ sato lajjākopaviṣādabhītyaratibhiḥ kṣetrīkṛtā tāmyate||
281 MR 7.5(158). yena svāmikulaṃ ripor iva kulaṃ dṛṣṭaṃ vinaśyat purā mitrāṇāṃ vyasane mahotsava iva svasthena yena sthitam| ātmā yasya ca vañcanāparibhavakṣetrīkṛto ’pi priyas tasyeyaṃ mama mṛtyulokapadavī vadhyasrag ābadhyatām||
282 I personally have no recollection of ever coming across it elsewhere. A hasty search of various electronic corpora yields one occurrence of the word in the Agnipurāṇa, one in a subhāṣita in the Saduktikarṇāmṛta ascribed to Bhikṣu, and several occurrences in alchemical texts where it seems to be a technical term.
Monier-Williams notes its attestation in the Agnipurāṇa and in the Kādambarī.
283 See also note 276 above.
284 MR 5.15(123), mudrā tasya karāṅgulipraṇayiṇī siddhārthakas tatsuhṛt; ND p. 84 (example of the saṃdhyaṅga
may be a further significant parallel between the two dramas, though arguably the name of Candragupta, “sheltered by the Moon,” may naturally tempt any poet to use the meta-phor of an eclipse of the Moon for situations in which the king is threatened. The Mudrā-rākṣasa commences with Cāṇakya’s wrathful entry on stage after misunderstanding the stage manager’s stanza about an eclipse to refer to an attack on Candragupta’s person,285 while Act 5 of the Devīcandragupta has a song accompanying the entrance of Candragupta286 which says that the Moon, having overcome darkness, now enters the sky to leap over the planet287 that eclipsed it.
Even before the discovery of the Devīcandragupta, JAYASWAL (1913:265) had pro-posed that the killing of Parvataka in the Mudrārākṣasa by means of a poison damsel may be a “a veiled defence of the scandalous murder of the Śaka Satrap,” accomplished by Candragupta in the guise of a woman.288 The poison maid, however, is featured in other versions of the legend of Candragupta, independent of Viśākhadatta,289 and the Candra-gupta of the Mudrārākṣasa was not directly involved in (and probably not even aware of) the murder of Parvataka and was at worst guilty by inaction. If Viśākhadatta was indeed a contemporary of Candragupta II, then Candragupta Maurya’s complicity in the killing of Parvataka by a deadly lady may indeed have been seen by the contemporary audience as a reference to the slaying of the Śaka kṣatrapa, but the mere similarity of the two dramatic episodes cannot qualify as evidence either for the date of the Mudrārākṣasa or for the iden-tity of its author with that of the Devīcandragupta.
For want of any evidence to the contrary, I nonetheless believe we should con-tinue to grant Viśākhadatta the benefit of doubt and accept the Devīcandragupta as another of his plays, but any grand hypotheses built on this postulation must be consumed with a grain of salt. The following section describes some further circumstantial evidence that may contribute to establishing both the identity of the authors of the two dramas and the date of the plays.
The Minister Safar
We learn from the Mojmal al-tawāriḵ that Ravvāl had a minister (wazir) blind in both eyes, named Safar. It had been he who advised the king to accommodate the Śaka ruler’s demand and hand over Queen Dhruvadevī (ELLIOT 1867:110), while later on he was
called prārthanā in viveka 1), pādau tvajjaghanasthalapraṇayinī sandānayen mekhalā.
285 MR 1.6, krūragrahaḥ saketuś candraṃ saṃpūrṇamaṇḍalam idānīm| abhibhavitum icchati balāt…
286 ND p. 194 (example of prāveśikī dhruvā), eso siya-kara-vitthara-paṇāsiyāsesa-veri-timiroho| niya-vihi-vaseṇa caṃdo gayaṇaṃ gahaṃ laṃghiuṃ visaï|| (Note that vihivaseṇa is an emendation of vihavareṇa by the editor of the ND. RAGHAVAN 1978:848 gives the following chāyā for this stanza: sita-kara-vistara-praṇāśitāśeṣa-vairi-timiraughaḥ| nija-vidhi-vaśena candro gaganaṃ grahaṃ laṅghayituṃ viśati||)
287 I.e. Rāhu; see page 75ff. for an explanation the appearance of this virtual planet in the Mudrārākṣasa. Just as the prologue of that play, both these verses cited from the Devīcandragupta employ the generic term graha rather than the specific name of Rāhu.
288 Known at that time only from the Harṣacarita, see note 266 above.
289 See Part III about the legend and page 154 about this femme fatale.
instrumental in forcing Barkamārīs to hide in the guise of a madman, fomenting the king’s suspicions against his younger brother (ibid. 111). At the end of the story, when Barkamārīs had killed the king and assumed the throne, he summoned Safar and said to him:
I know that it was you who counselled my brother in his dealings with me, but this was no fault nor is it blameable. It was God’s will that I should be king, so continue to govern the kingdom as you did for my brother.” Safar replied, “You have spoken the truth, all that I did was for the good and advantage of your brother, not out of enmity to you. But I have now resolved upon burning myself, and cannot do as you desire. I was with your brother in life, and I will be with him in death.”
Barkamárís told him that he wanted him to write a book on the duties of kings, on government and justice. Safar consented, and wrote the book, which is called
“Adabu-l Mulúl,” “Instruction of Kings.” … When it was finished he took it to Barkamárís and read it, and all the nobles admired and praised it. Then he burnt himself.290
Provided that the narrative of the Mojmal al-tawāriḵ is indeed based on the Devī-candragupta, we have here a very interesting parallel to the storyline of the Mudrārākṣasa:
a minister of the old regime faithful to the deposed and killed ruler, whose services are wanted and obtained, at least temporarily, by the victorious pretender. There is also a no-table subtext shared with the Mudrārākṣasa here: while loyalty to the king is paramount, politics is all business, nothing personal. Safar was never called to task about his actions against Barkamārīs, because they had been “not out of enmity.” Nor is Rākṣasa held ac-countable for his manoeuvring against Candragupta, and indeed, according to the Mudrā-rākṣasa it is the mark of an unsophisticated mind to take personal offence at what a politi-cal opponent does.291
JAYASWAL (1932a:21–22 and 1932b) links Safar to Viśākhadatta by a completely dif-ferent thread. He cites epigraphic evidence that Candragupta II had a minister named Śikharasvāmin, and argues that the name of Safar (
رفس
) can be derived with negligible alteration (i.e. assumed corruption) in Arabic script from Saqar or Siqar (رقس
), which in turn may have been an adoption of Sanskrit Śikhara. The identification is intriguing but goes far out on a limb in several respects and has no strength as evidence. Safar, after all, only consented to serve Barkamārīs to the extent of writing a manual on politics, not in a ministerial capacity. Furthermore, given the degree of dissimilarity between the name290 ELLIOT 1867:112.
291 See Bhāgurāyaṇa’s lecture to Malayaketu when the latter receives the (false) information that Rākṣasa had been the agent of Parvataka’s death, after MR 5.7(115): kumāra iha khalv arthaśāstravyavahāriṇām arthavaśād arimitrodāsīnavyavasthā, na laukikānām iva svecchāvaśāt. yatas tatra kāle … rākṣasasya … devaḥ parvateśvara … mahān arātir āsīt tasmiṃś ca rākṣasenedam anuṣṭhitam iti na doṣam ivātra paśyāmi. “Surely, Prince, in such matters those who deal in politics act in a hostile, friendly or indifferent manner as dictated by utility, not by personal choice like commoners. Since at that time his majesty Parvateśvara was a great enemy to Rākṣasa, and Rākṣasa proceeded in this way against him in that capacity, therefore I see no fault at all in this.”
pairs Ravvāl—Rāma(gupta) and Barkamārīs—Vikramāditya, the name Safar may have been derived from a wide gamut of Indic names.
BHANDARKAR (1932:200) and JAYASWAL (1932b) both raise the idea that the “book on the duties of kings, on government and justice” written by Safar may have been the Kāmandakīya Nītisāra, a textbook on polity probably composed in the reign of Candragupta II,292 and DIKSHITAR (1993:13–15) endorses the suggestion. SOHONI (1955) goes even further and equates Safar/Śikhara to Rākṣasa through a marvellous linguistic feat. The name of Rākṣasa, he begins, would of course be incised backwards on his signet ring, so that the seal impression would show the correct sequence of akṣaras. The string रा-ᭃ-स, rā-kṣa-sa would thus be स-ᭃ-रा, sa-kṣa-rā on the signet, which shows considerable similarity to the name Śikhara.293 Sohoni sees a clue to this pun in the first nāndī verse of the play, where Śiva asks Pārvatī, “‘How is that you do not know this name which is so familiar to you,’ a query which any contemporary or knowledgeable audience could have deeply appreci-ated.”294 The idea is ingenious and may even be correct, but as evidence, its weight is even more infinitesimal than that of the identification of Safar as Śikhara.
292 See also page 102 about this book.
293 At least in the consonants and especially on the provision that speakers made little or no distinction between kṣa and kha on the one hand and śa and sa on the other. SOHONI (1955:492) mentions the issue with kṣa but is silent about śa; both provisions are possible but in my opinion not likely to have been the case for educated speakers. Sohoni also fails to mention that the text as incised on the signet would of course be a mirror image of रास rather than the straight letters in an inverted sequence. (See the cover image, disregarding the first and last glyphs of the inscription, for what the straight string may have looked like in a contemporary script.) Note also that Sohoni later became a staunch adherent of the theory that Viśākhadatta lived in the Maukhari court, saying (SOHONI 1981:186) that “There are some scholars who hold that Viśākhadatta was a contemporary of Candragupta … But this is unthinkable. For Viśākhadatta was a contemporary of Avantivarmā Maukhari of Kānyakubja.” An audience of that age would of course have been unable to appreciate this alleged pun.
294 Translation and remark by SOHONI 1955:496. A more accurate translation would be, “Of course it’s her name. What makes you forget even though you know it?” (See page 176 for the text and discussion of this stanza.) Sohoni omits from his translation that the quote is about the name of a woman, and that it speaks about “forgetting” rather than “not knowing.”