Part II. The Author in Context
5. Śakaṭāla and Cāṇakya
Unlike Śakaṭāla in Bṛhatkathākośa 157, Kavi here does take revenge on his king, and achieves it by provoking Cāṇakya’s anger. The present Bṛhatkathākośa chapter intro-duces Cāṇakya as the son of a man called Kapila, whose sister was married to Kavi. The two may thus have known one another, but this is not mentioned in the tale. Cāṇakya is intro-duced at the start of the chapter, and reappears when Kavi, after his restoration, meets him by chance. At this moment Cāṇakya is digging some grass out of the earth, and when Kavi asks why he is labouring, the former explains (at length) that the grass has pricked his foot, so now he is destroying it down to the roots.
Kavi reckons that a person who is ready to go to such trouble over such a trifling matter would be just the right man to destroy Nanda, provided that the king insults him.
Therefore (after subjecting Cāṇakya to some tests) Kavi suggests to Nanda that he should donate a thousand cows to brāhmaṇs, to which the king replies, “sure, bring on the brāhmaṇs.”133 Kavi then leads Cāṇakya to the palace and bids him take the seat of hon-our.134 The brāhmaṇ takes the first seat and puts his things down on others, then Kavi tells him that Nanda wants him to sit elsewhere. The minister moves him through a succession of seats,135 and finally—still saying it’s Nanda who wants all this—tells Cāṇakya to wait out-side. The version in the Vaḍḍārādhane has no family connection between the minister and the brāhmaṇ; instead, Subandhu hears an astrologer’s prediction that Cāṇakya (son of Somaśarman and Kapilā) would destroy the Nandas, and resolves to befriend him. He per-suades Mahāpadma to donate some villages to Cāṇakya, then revoke the grant.
In either case, Cāṇakya does not tolerate the insult (in the Vaḍḍārādhane he swears to kill Nanda and Subandhu) and asks if there is anyone who wants the kingdom, where-upon “a man” volunteers.136 They depart together from the capital, take refuge in an island fortress137 and foment yet another rebellion among the rulers of the provinces. They have no named ally, but there is one particular provincial ruler (pratyantavāsibhūpa) who first approaches Cāṇakya. This part of the text is not entirely intelligible,138 but it seems that Nanda manages to turn the provincial rulers against one another, while Cāṇakya in turn
133 Bṛhatkathākośa 143.48, gosahasraṃ dadāmy eva brāhmaṇān ānaya drutam.
134 Bṛhatkathākośa 143.49, kavir niveśayām āsa pradhānāgrāsane tadā.
135 This seems like an inferior description of the episode in Pariśiṣṭaparvan 8 above.
136 Bṛhatkathākośa 143.60, cāṇakyavacanaṃ śrutvā naraḥ ko ’pi jagāv idam| aham icchāmi bho rājyaṃ dīyatāṃ me drutaṃ prabho|| The man’s name is Candrabhukta in the Vaḍḍārādhane and Caṃdaguttu (i.e. Candragupta) in the Kahakosu of Śrīcandra, which apparently also includes this story (WARDER 1992:247). The variant Candrabhukta may have arisen because of a sort of popular etymology rather than mere ignorance: in an episode described in the Jaina commentarial tradition (omitted from my summary above) and also included in the Vaḍḍārādhane (KHADABADI 1979:85) Candragupta’s pregnant mother “drinks the Moon” to satisfy her craving (dohaḍa). The episode is clearly intended to serve as an explanation of Candragupta’s name (though it also has another function in the plot), and actually works better with Candrabhukta, (“one who has consumed the Moon”) than with Candragupta (“one protected by the Moon”).
137 Bṛhatkathākośa 143.62, jaladurgaṃ praviśyāsau vārdhimadhye sudhīradhīḥ| rājyam anveṣayaṃs tasthau cāṇakyaḥ kṛtaniścayaḥ||
138 Verses 66–70 of this chapter seem partially corrupt; the word parvata occurs in 65 and may be a name. The summaries of RAGHAVAN (1973:63) and WARDER (1992:247) both gloss over this part.
incites distrust among Nanda’s ministers and ends up winning the war. Victorious, they kill Nanda and anoint “the man”139 as the new king.
The story of the campaign is slightly different in the Vaḍḍārādhane and contains more details that seem to be derived from the Jaina commentarial tradition. Here Cāṇakya recruits a Buddhist monk who makes gold by alchemical processes, but fails to conquer the capital. There is also a hot gruel episode, but here it is Cāṇakya himself who burns his fingers and is rebuked by the old woman (KHADABADI 1979:86). They raise a rebellion of provincial rulers, and at last conquer the capital. Nanda survives the attack and flees, and Cāṇakya first imprisons Subandhu, then releases him and restores him to ministership under Candrabhukta, though earlier he had sworn to kill him.
The story is not quite over yet in either text: Cāṇakya subsequently becomes a Jaina mendicant, while the ex-minister Subandhu takes refuge with Sumitra, the governor of a place called Krauñcapura in the western part of the southern region, and continues to harbour anger at Cāṇakya.140 When Cāṇakya in his wanderings comes to Krauñcapura, he contrives to burn the mendicant to death.141 He goes to hell for his deed, while Cāṇakya willingly chooses not to escape the fire and attains siddhi.
The Bṛhatkathā Tradition
A version of the fable that brings together Śakaṭāla and Nanda with Cāṇakya and Candragupta is incorporated into the framework story of some versions of the Bṛhatkathā.
The original Bṛhatkathā, “The Great Story”—supposedly composed in the Paiśācī language by a poet named Guṇāḍhya—may itself be legendary. If there ever was such an original, it almost certainly did not include this framework story, one of the heroes of which is Guṇāḍhya himself. The two retellings of the Great Story that incorporate the story about the end of the Nandas both originate from 11th-century Kashmir. The Bṛhatkathāmañjarī of Kṣemendra is slightly the earlier of the two; the Kathāsaritsāgara of Somadeva was com-posed a few decades later. These two epitomes in Sanskrit verse probably go back to an earlier Kashmiri (or in any case, northern or north-western) version of the Great Story, perhaps written in Prakrit. The Kathāsaritsāgara probably retains most of the content of this archetype, while the Bṛhatkathāmañjarī compresses it heavily, but does include some material not found in the former (TRAUTMANN 1971:33).
The framework story of Guṇāḍhya is absent from the Bṛhatkathāślokasaṃgraha, a (probably) Nepali retelling of the Bṛhatkathā written in the 8th or 9th century, as well as the Tamil transcreation of the Bṛhatkathā, the Peruṅkatai of Koṅkuvēḷir142 and the Prakrit
139 Again, simply taṃ naram in Bṛhatkathākośa 143.72.
140 For killing Nanda in the Bṛhatkathākośa, but his motive is unclear in the Vaḍḍārādhane.
141 Warder’s summary of the Vaḍḍārādhane gives no details of where this happens, but does include the incident.
142 Also 8th or 9th century, ZVELEBIL 1974:135.
Vasudevahiṃḍī.143 On the other hand, versions of the story of Guṇāḍhya do appear, but without the Nanda episode, in the Nepālamāhātmya.144 Therefore the inclusion of this par-ticular episode in the Kashmiri version(s) may be a late and/or localised feature (LACÔTE 1908:35–36).145 Nonetheless, as demonstrated above, stories very similar to this episode are found in old and geographically widely spread sources, so whether or not it was incorpo-rated into an early Bṛhatkathā incarnation is immaterial as far as the present study is con-cerned.146
Śakaṭāla’s Revenge by Cāṇakya
The story told in the Kathāsaritsāgara (and, more concisely, in the Bṛhatkathā-mañjarī) begins with Vyāḍi, Indradatta and Vararuci. Upon the completion of their studies, their teacher requires ten million gold coins. Since Nanda is known to possess nine hun-dred and ninety million,147 they go to him to beg for a donation. However, as in the previ-ously discussed versions, they find Nanda dead (again, for no particular reason148), so Indradatta uses his magical expertise (yogasiddhi) to take possession of his body and grant the gold to Vararuci.
The king’s clever minister, Śakaṭāla, realises what has happened, but reasons that since Nanda’s son is still a child, it’s safer for the realm to keep an adult, even if an impos-tor, on the throne. He orders a search to find and burn any lifeless bodies in the vicinity, forcing the brāhmaṇ’s soul to remain in the king’s body. Vyāḍi reports to Indradatta (posing as king, referred to as Yogananda as opposed to Pūrvananda or Satyananda, the
“original” or the “true” Nanda) that his original body has been burned, and cautions him that Śakaṭāla is in the know and may sooner or later depose Yogananda to make the gen-uine Nanda’s son, Candragupta, king,149 so Indradatta should appoint Vararuci as his chief minister to foil any attempts Śakaṭāla might make. On Vararuci’s suggestion Yogananda
143 Pre-6th century, ALSDORF 1935:278.
144 A section of the Skandapurāṇa; see LACÔTE 1908:291–304.
145 The Haracaritacintāmaṇi—a poetic collection of legends about Śiva written by Jaya(d)ratha in the 12th century, also in Kashmir (LIENHARD 1984:203)—also tells the legend of Guṇāḍhya including a brief version of the Śakaṭāla story, but omitting the part about the fall of the Nandas.
146 WARDER (1992:60) does not exclude the possibility that the Nanda cycle might have been part of Guṇāḍhya’s original, but considers it unlikely. TRAUTMANN (1971:33) says it is very probable that this particular yarn is a late addition to the Guṇāḍhya legend. CHATTERJEE (1935) constructs an elaborate theory about successive recensions of the Bṛhatkathā in an analysis of whether Viśākhadatta could after all have based the Mudrā-rākṣasa on (some version of) the Bṛhatkathā (as alleged in the Daśarūpāvaloka, see page 92). He proposes (unconvincingly) that the episode in question was introduced before the creation Kashmiri archetype, but rejects (convincingly) the possibility that any Bṛhatkathā version could have contained a story that shares more with the Mudrārākṣasa than those told in the Kathāsaritsāgara and Bṛhatkathāmañjarī.
147 Kathāsaritsāgara 1.4.93, svarṇakoṭir me dīyatām and 1.4.95 navādhikāyā navateḥ koṭīnām adhipo hi saḥ.
148 Kathāsaritsāgara 1.4.98 simply says the king died; in Bṛhatkathāmañjarī 1.2.114 he dies “because of fate,”
daivāt.
149 Kathāsaritsāgara 1.4.116, pūrvanandasutaṃ kuryāc candraguptaṃ hi bhūmipam.
casts Śakaṭāla and his hundred sons into an oubliette,150 each day giving them food suffi-cient for just one man. As ministers and their sons are wont to do in such situations, they decide that Śakaṭāla should eat the food and live to take revenge on Yogananda.
This coherent and detailed account is probably very similar to the archetype that was summarised in a nutshell in the Avantisundarī and told with some variation in Bṛhat-kathākośa 157. Subsequently Yogananda, corrupted by wealth and power, becomes whim-sical and spends his life in the pursuit of pleasure instead of governing. Vararuci decides to restore Śakaṭāla to help control the unruly king and the two become friendly rivals.
Shortly afterward, the king orders Vararuci’s execution on the suspicion that he has been cuckolding him. Unlike Bṛhatkathākośa 157, it is not Śakaṭāla who casts this suspicion on Vararuci;151 indeed, Śakaṭāla actually saves Vararuci’s life, hiding him in his house and kill-ing another man in his stead. Later on, Vararuci is in turn recalled because his help is needed in solving a riddle concerning (as in Bṛhatkathākośa 157) the false Nanda’s son (called Hiraṇyagupta in the Kathāsaritsāgara and Harigupta in the Bṛhatkathāmañjarī), who takes to a tree to escape from a lion, is joined by a bear, and at night throws the sleeping bear down to the lion.
After solving this riddle, Vararuci decides he has had enough of court life and retires to a hermitage. It is there, much later, that he hears the rest of the story of Yogananda and Śakaṭāla from a wandering ascetic. Śakaṭāla once met Cāṇakya (of whose origin there is no information) in the street, digging up a clump of grass because it had pricked his foot. Like Kavi in Bṛhatkathākośa 143, Śakaṭāla reasons that such a man is just what he needs to implement his revenge. He promises to arrange for Cāṇakya to preside at a śrāddha ceremony in the king’s house, and even presents him to the king. Yogananda approves of Cāṇakya, but when it comes to the actual ceremony and Cāṇakya takes the first seat, another brāhmaṇ called Subandhu objects. The king (in his usual whimsical character) assigns priority to him, while Śakaṭāla washes his hands.152 The Bṛhatkathā-mañjarī narrates this event in a rather different form: here Śakaṭāla simply brings Cāṇakya to the śrāddha, shows him to a seat in the lowest row, and tells him (lying, apparently) that the king has humiliated him (by allotting this seat).153
Cāṇakya has a fit of anger, unties his topknot and vows not to tie it again until he has killed Nanda, which he proposes to accomplish in seven days.154 Unlike the tales in the
150 Kathāsaritsāgara 1.4.120 and passim, Bṛhatkathāmañjarī 1.2.124 and passim: andhakūpa.
151 Instead, Vararuci sees a painting of the queen in the nude. By his superhuman intuition he divines from the arrangement of auspicious marks on her body that there must be a mole somewhere on her private parts (which, obviously, even the painter had not seen exposed), and paints the mole in to complete the picture. When the king sees in the painting what no one but he ought to have ever seen in life, he comes to the conclusion that Vararuci has been having an affair with the queen.
152 Kathāsaritsāgara 1.5.117, na me ’parādha ity uktvā.
153 Bṛhatkathāmañjarī 1.2.215-216, upaviṣṭam adhaḥpaṅktyāṃ śakaṭālas tam abravīt|| rājñāvamānito ’sīti.
154 Kathāsaritsāgara 1.5.119, avaśyaṃ hanta nando ’yaṃ saptabhir divasair mayā| vināśyo bandhanīyā ca tato nirmanyunā śikhā|| The vow is missing in the Bṛhatkathāmañjarī, but interestingly, there Cāṇakya’s hair is said to be already loosened when Śakaṭāla leads him to the palace (1.2.215, muktaśikhaṃ cāṇakyaṃ).
Pariśiṣṭaparvan, Bṛhatkathākośa 143 and the Vaḍḍārādhane, here Cāṇakya does not need a lengthy peregrination and a war to achieve his aim. He simply hides somewhere (according to the Bṛhatkathāmañjarī, in Śakaṭāla’s house), and there performs some conjur-ation (kṛtyā) as a result of which Nanda contracts a fever and dies in a week.155 Śakaṭāla anoints the original Nanda’s son Candragupta (in the Kathāsaritsāgara also killing the false Nanda’s son and making Cāṇakya minister), then retires to the forest.
Second Cross-Fertilisation
It is my impression that the fully fledged legend of Cāṇakya’s revenge—which had already absorbed some details from the Śakaṭāla stories—had a second encounter with the tale of Śakaṭāla’s revenge. There were after all, at this stage of evolution, two distinct types of tale about brāhmaṇs taking redress on kings called Nanda. In one of them, Nanda had, rightly or wrongly, imprisoned a brāhmaṇ minister and his family. Either the minister or his son survived the punishment, was returned to favour, and avenged the destruction of his family. In the other type, a brāhmaṇ (and a future minister) called Cāṇakya came to Nanda to seek a donation, but was refused and disgraced, for good or bad reasons. He left to enlist allies (or puppets) to his cause, and returned to avenge his humiliation.
As the two stories above show, these tales became linked to form a new story, in which the wronged minister, rather than killing the king himself, found an ally (or puppet) to accomplish his aim—none other than Cāṇakya. The context of their meeting—Cāṇakya’s revenge on grass and Śakaṭāla’s conclusion that this is the right man for the job—may have a prototype in the way the man who burns ant heaps is enlisted by Cāṇakya to root out robbers. This episode is related in the Pariśiṣṭaparvan (see page 109), in which the story of Cāṇakya’s revenge (page 107) starts immediately after (and within the same chapter as) the story of Śakaṭāla (114), but with no connection between the two apart from contiguity.
This shows that the second cross-fertilisation of the two tales was probably a gradual pro-cess. The Avantisundarī may in fact also contain a clue to this merging. The tale of Nanda and Āryaka there (page 112) ends with the cremation of Indradatta’s body, but a summary account of the future follows, according to which Mahāpadma lived for (altogether) eighty-eight years;156 thereafter Cāṇakya destroyed all Mahāpadma’s sons and established Candragupta Maurya in their place.157 If I am right in emending and interpreting the text,
155 Kathāsaritsāgara 1.5.121-122, sa cāṇakyo dvijaḥ kvāpi gatvā kṛtyām asādhayat|| tadvaśād yoganando ’tha dāhajvaram avāpya saḥ| saptame divase prāpte pañcatvaṃ samupāgamat|| and Bṛhatkathāmañjarī 1.2.216, cāṇakyanāmnā tenātha śakaṭālagṛhe rahaḥ| kṛtyāṃ vidhāya saptāhāt saputro nihato nṛpaḥ|| (This is one of the verses cited in the Daśarūpāvaloka to show that the Mudrārākṣasa is based on the Bṛhatkathā, see page 93 and note 9 there.)
156 As in the Purāṇic list, see note 229 on page 136.
157 The Avantisundarī is corrupt here, but it does seem to say that Mahāpadma’s sons were eight in number and may also mention the total length of their reign: eṣa khalv idānīm aṣṭāśītim abdān atītya tiṣṭhate mahāpadme tatputrā(va?na)ṣṭāvapyalābhir daśabhir vatsarair ekaikam uddhṛtyāryākāpyovāpakopitena manasvinā cāṇakyena mauryacandraguptaḥ pratiṣṭhāpitaḥ… The parentheses and question mark in the citation are Pillai’s; bold
then it was the minister who infuriated Cāṇakya, causing him to embark on the destruc-tion of the Nandas. In the light of the more detailed versions of the story presented in this chapter, this probably does not mean that Cāṇakya was furious at the minister, but rather that the minister provoked Cāṇakya’s fury, directed at the Nandas. Daṇḍin may have thought this so obvious as to need no further elaboration; on the other hand, the composer of the Avantisundarīkathāsāra may have omitted this detail precisely because it was not transparent to him.
To venture a few steps in another direction, actually the title “Śakaṭāla’s Re-venge” (which I borrow from Trautmann 1971:35) may not be the best choice for the tales that influenced the early Cāṇakya stories. According to WARDER (1992:74-75) there is a tale known widely (and with much variation), originally perhaps called Nandaprakaraṇa, about a minister called Virocana or Vairocana.158 King Nanda sends Virocana on a mission to the frontiers, and attempts to seduce his wife while he is away. Upon his return, Virocana concludes that Nanda is corrupt, and kills him while out hunting, throwing his body in a well. He crowns Nanda’s son, who later becomes suspicious and starts an investigation, eventually discovering a brāhmaṇ boy159 who had witnessed the murder. He imprisons Virocana and his family,160 but the minister’s youngest son survives and eventually kills the king. Possibly the original Śakaṭāla was a problem solver who saw through the false Nanda and killed him, and was later conflated with Virocana, giving rise to stories in which the motifs of the false Nanda, the imprisonment and the minister’s (or his son’s) revenge on Nanda (or his son) mix in various and sometimes self-contradictory manners and the minister goes by diverse names.
To return to the fable of “Śakaṭāla’s Revenge by Cāṇakya,” this story accounts (more or less in harmony with the Mudrārākṣasa) for most of the events and characters that the drama presupposes. There remain only two major points of divergence: why is there no Rākṣasa in these tales, and why is there no Śakaṭāla (or an analogue) in the Mudrā-rākṣasa?
emphasis is mine, marking corrupt text that I am inclined to emend to …tatputrān aṣtāv api dvādaśabhir vatsarair ekaikam uddhṛtyāryakakopitena… (possibly with another word, such as udvāpa or āvāpa between āryaka and kopitena). The Avantisundarīkathāsāra (4.62) summarises the essentials but omits the details:
mahāpadmasutān sarvān uddhṛtyāmātyakopitaḥ| tatpade khalu cāṇakyaś candraguptam atiṣṭhipat||
158 Compare Vairocana (a Śakaṭāla-clone) in the anonymous Bikaneri preamble (page 128), and possibly Vairodhaka, Parvataka’s brother in the Mudrārākṣasa, whose name is also known in MSS as Vairocaka.
159 A baṭu, like Cāṇakya in the Mudrārākṣasa and its epitomes (page 130).
160 Presumably in a well, which would be as fitting here as it would have been in the story of Kavi (page 116).