Balarāma Drags the River Yamunā


By Dr. Peter Singer

The Sanskrit Reading Room welcomed Dr Simon Brodbeck from Cardiff University on 17 January 2018. It was the second time we had hosted a scholar from that institution and, as on the previous occasion (with Dr James Hegarty), we were treated to a fascinating and illuminating presentation based on a work-in-progress translation of a classical Sanskrit text – this time the Harivaśa of Vyāsa.

In strikingly lucid and thoughtful style, Dr Brodbeck, after first outlining the nature and structure of the Harivaśa (and drawing our attention to the basic issues in the text’s manuscript tradition), presented us with a detailed analysis of one particular episode, from section 83 of the work – the dragging of the river Yamunā by Balarāma – which he also contextualized within the broader narrative structure in order to highlight its salient themes. Three major strands in recent interpretation of the text were identified, and Dr Brodbeck’s own analysis – focussing especially on sexual violence as well as on psychological or psychosocial themes in the text – placed in relation to those.

The Harivaśa is a tailpiece to the Mahābhārata; it can be characterized in a way as a supplement containing the episodes of the legend not found elsewhere in the great epic (those which had nothing to do with the war between the Kauravas and the Pāņḍavas), and offering a range of other stories concerning Kṛṣṇa.

Our session had one particular episode from the Harivaśa at its centre, namely that in which Kṛṣṇa’s brother Balarāma drags or ‘ploughs’ the river Yamunā. At one level, the story is clearly an aetiological myth explaining the present course of the river and the fact that it fertilizes the plains of Mathurā and Vṛndāvana. But the narrative is replete with sexual imagery – Yamunā becomes a woman whom Balarāma violently assaults – as well as raising questions of the role and function of Balarāma in the narrative, especially in relation to his brother Kṛṣṇa; and it is these aspects especially that were explored.

Here, we considered Balarāma as an elder brother whose deeds and status are problematically ‘overstepped’ by the younger brother, Kṛṣṇa (with reference to parallel cases of such overstepping, e.g. that of Yudhiṣṭhira by Arjuna, which is central to the Mahābhārata). There is a psychological dimension here – how does Balarāma react to his younger brother’s exploits? – and it is relevant to observe how differently their exploits and experiences are presented in the narrative.

The Yamunā episode is immediately preceded by fight scenes against Jarāsandha, in which Balarāma’s plough plays an important role, and by Balarāma’s return to Vṛndāvana, which is presented as a sort of coming-of-age.

We may, however, contrast the depiction of Balarāma ‘enjoying himself’ in the forest with that of Kṛṣṇa. In an extraordinarily sensual scene, the latter is shown revelling in pleasures, including sexual ones, in a way that Balarāma is not. There is significance, too, in the way that the consensual nature of the activity is highlighted in this scene of Kṛṣṇa’s sexual abandon – he is presented, in quite explicit terms, as the object of sexual attention of a large number of cowgirls – and that this is thrown into relief by a scene immediately following (with bulls and heifers) highlighting the possibility and nature of non-consensual sexual activity.

When we come to the episode involving Balarāma and Yamunā itself, the sexual as well as the non-consensual nature of the encounter is clear. The plough itself seems clearly to be represented at some level as a phallic image. On some levels it is an incongruous choice of weapon, but one constantly associated with Balarāma in literature and iconography; indeed the name Halāyudha, one of several used for him in the narrative, derives from this. It is also conceived (from as early as the Ṛgveda) as distinctively the weapon against the wilderness, that which tames – again with specifically sexual connotations.

In this context it was interesting to consider the myth of Sītā emerging from a furrow ploughed by the king, her father Janaka, and the various interpretations that can be placed upon it. Should such a mythical account be taken ‘at face value’ – or perhaps as somehow co-existing with a natural birth account? Or should it be taken as purely metaphorical? In discussion some pointed to the prevalence of such non-sexual accounts in divine births throughout Indian mythology, and it was also suggested that there are kings who plough and sow in other Indian myths; Dr Brodbeck pointed to the sexual imagery in play in this use of ‘ploughing’, as well as to the standard linguistic usage associating ‘field’ with ‘wife’, and suggested that we should take the myth of Sītā’s birth as metaphorically representing natural, sexual reproduction.

To return to the episode itself: Balarāma’s importunate address to Yamunā, her intriguing response to it – where, in her reluctance to comply, she is described as ‘confused by her nature as a woman’ (strīsvabhāvena mohitā) – and the further verbal exchange between them at the end of the episode all rewarded further exploration.

Dr Brodbeck drew attention especially to the interpretive approaches of Couture, Vemsani and Sanford. Couture emphasizes the nature of Balarāma and Kṛṣṇa as aspects of Viṣṇu, and, in a deep theological interpretation, suggests that the destabilized behaviour of Balarāma in our scene – his tendency to invert the natural cosmic order – is due to his temporary separation from his brother. Vemsani meanwhile stresses Balarāma’s association with fertility cults, suggesting (what is not explicit in the text) that his main purpose was to reroute the river through Braj. Dr Brodbeck’s own interpretation was closer to that of Sanford, who pays attention to the psychosocial aspect of the episode, pointing out its violent dissonance with the ‘Braj pastoral’, and highlighting the themes of rape and of the plough as phallic image.

What, though, are we to make of that puzzling phrase strīsvabhāvena mohitā, with its implication that things could have been different if Yamunā’s response had been different – or, indeed, of her impassioned plea to be returned to her previous course after the assault (something dissonant with the sexual metaphor, since sexual acts in their nature cannot be undone)? Problematic, too, is Balarāma’s response to this, which seems to suggest that she should be happy, or at least at peace, because his fame will endure through her.

The first of these problems is particularly challenging. It seems to be implied that some response was possible whereby Yamunā was neither taken by force nor complied voluntarily – but what response? And is it that she does not see this possibility because she is a woman, or that she does see it, but because she is a woman, decides not to pursue it? The text is mysterious. Whatever the precise interpretation, it seems unavoidable to conclude that some deficiency is being attributed to the female Yamunā – a ‘deficiency’ which can, for the contemporary reader, serve at best as the mote which throws into violent relief the beam of Balarāma’s transgression.

Those present were grateful to Dr Brodbeck for a highly engaging textual analysis which at the same time provoked thought in relation to a number of areas – brother-brother relations, sexual imagery and sexual violence, metaphorical or literal interpretations of birth narratives, patrilinearity, the religious and psychosocial function of mythic characters – far beyond the confines of this particular text.

Dr Peter Singer is Research Fellow in Classics at Birkbeck, University of London.


For further work by Dr Brodbeck please see:

Brodbeck, Simon Pearse. The Mahabharata Patriline: Gender, Culture, and the Royal Hereditary. Routledge, 2017.

Brodbeck, Simon, and Black, Brian, eds. Gender and Narrative in the Mahabharata. Routledge, 2007.



Refutation of the Existence of a Demiurge God


This Sanskrit Reading Room focused on a text by Ratnakīrti, the 11th-century Buddhist philosopher. He wrote the Īśvarasādhanadūṣaṇa as a refutation of a proof of the existence of a creator god propounded by the Hindu tradition. Dr Gorisse, a specialist in Jaina logic, set up the reading by firstly presenting some basics of the debating hall in Indian philosophy in order to prepare us for the vocabulary and discourse of the text. Then she presented the proof under discussion. Finally, she lead us through a line by line reading of one attack on this proof, in which we were able to get into the nitty-gritty of the form and semantics of the argumentative discourse.

Logical conventions

Firstly, Dr Gorisse explained that the basics of pan-Indian epistemology were very much influenced by the Nyāya school. It is the Nyāya Sūtra (c. 2nd-century CE) that is the reference manual for the definition of the pramāṇas, the types of cognition, amongst which inference (anumāna), the means to acquire new knowledge by reasoning, also constitutes the core of any philosophical argument. The Nyāya Sūtra teaches the only good way to state an inference:

pratijñā-hetu-udāharaṇa-upanaya-nigamanāny avayavāḥ (NS 1.1.32)

[Statement of] the thesis, the evidence, the account, the application and
the conclusion are the members [of reasoning]

This reasoned process is illustrated by the canonical five-step proof, sometimes called the “Indian syllogism”.

  • There is fire on the hill (pratijñā or thesis)
  • Because there is smoke on the hill (hetu or reason, evidence)
  • Wherever there is smoke, there is fire e.g. in a kitchen (udāharaṇa or justification, which consists in the statement of the vyāpti, the relation of necessary concomitance, between the two inferential properties ‘being endowed with fire’ and ‘being endowed with smoke’, along with an example)
  • So too in the case of the hill there is smoke (upanaya or application)
  • Therefore there is fire on the hill (nigamana or conclusion)

A basic familiarity with Naiyāyika theory is necessary to read the Īśvarasādhanadūṣaṇa, because Ratnakīrti engages with the terminology of Nyāya logic. However, Buddhist epistemology was fundamentally different to Naiyāyika in that the number of pramāṇas differed, as did the interpretation of those pramāṇas. And so Ratnakīrti’s complex argument, although built on the foundations of Nyāya logic, critiques those very foundations.

Furthermore, to read an Indian logic text, one needs to be alert to the features of literary style unique to this genre. Ablatives are used everywhere to indicate causal relationship in the sense of a logical cause. Copycat syntactic structures are employed to construct effective refutations. And philosophical texts have their own stock phrases, such as ‘nanu iti cet’, which introduces an objection.

The basis of the overall argument

Ratnakīrti’s text is a dūṣaṇa, a refutation. It addresses the Nyāya proof that there is a buddhimat, or conscious being, who has made the world and Ratnakīrti rejects the said proof as defective. The vivādādhyāsita, the subject under debate, in this proof consists of natural effects such as mountains, grass, or trees. The sādhya, or property to be inferred, is “buddhimaddhetuka”, namely the fact that there is a conscious cause even of these natural effects. Following the model of the above fivefold proof, the Nyāya proof goes more precisely as follows:

Thesis: Mountains etc. are produced by a conscious creator (vivādādhyāsitaṃ buddhimaddhetukam)

Evidence: Because these are effects (kāryatvāt)

Explanation: Whatever is an effect is produced by a conscious one, like a pot (yat kāryaṃ tadbuddhimaddhetukam, yathā ghaṭaḥ)

Application: And this [subject under discussion] is an effect (kāryaṃ cedam)

Conclusion: Therefore it is produced by a conscious one (tasmād buddhimaddhetukam)

Dr Gorisse pointed out that as Ratnakīrti’s subsequent reasoning unfolds, we can see the accretion of layers of argument like a geologist dates rocks – from Nyāya (2nd century) to Dignāga (5th-6th century) to Dharmakīrti (7th century) to Ratnakīrti himself (11th century).

Defective arguments

(A) First, the Naiyāyikas have at heart to show that their proof suffers from none of the five defects presented in the Nyāyasūtra, namely inconclusive (savyabhicāra), contradictory (viruddha), unestablished (sādhyasama), neutralized (prakaraṇasama) and inopportune (ātītakāla). During the session, we translated the section devoted to showing that the proof is not inconclusive, that it has no counter-example. Concretely, it is not possible that something which is an effect is not produced by a conscious being. There, Ratnakīrti asserts that there is such a counterexample, namely the fact that the grass – which is an effect, since it is constituted of parts arranged together – is growing, while no activity of a conscious maker is to be observed.

(B) At this point, Buddhists and Naiyāyikas engage in a discussion over the necessary conditions to legitimate an inference. The Naiyāyikas argue that it is not necessary to observe the co-presence of the effect ‘grass’ with its conscious maker since, in the standard example, the presence of a pot is sufficient for us to know that there has been a potter even if we have not seen this potter. In response to this, the Buddhists point that there is an important difference between the two cases, since we have never seen grass growing with the activity of a maker. And the Naiyāyikas hold their position, arguing that also in cases of specific types of pots that we have never seen produced by a potter, we would still agree that there was a potter at some point in their process of production. In the end, both participants agree on the fact that no certainty can be reached by mere observation in similar cases (sapakṣa) and non-observation in dissimilar cases (vipakṣa), which is a reference to Dignāga’s theory of the triple nature of the hetu. Instead, only from a reflexion from the very concept of pot, respectively of grass, can we learn about their origin.

(C) From this point onwards, it is from Dharmakīrti’s framework that Ratnakīrti challenges the Nyāya proof. According to Dharmakīrti, it is by virtue of being the natural property (svabhāva) or the natural effect (kārya) of the property to be inferred, that an evidence is probative. Concretely, only if “being an effect” is proved to be the natural property or the natural effect of “having a conscious producer”, can it be considered efficient in this proof. But here too, the disagreements persist, since Ratnakīrti rejects the fact that the evidence (being an effect) is a natural effect of the property being inferred (having a conscious creator) – on the basis that there are different class distinctions to ‘being an effect’ (a mountain, a pot, a temple, grass) and so a property such as ‘having a conscious creator’ cannot be extended to them all equally and that only a specific type of effect proves the presence of “having a conscious maker”. In contrast with this, Naiyāyikas defend the view that “effects in general” are the probative mark.

This illuminating Sanskrit Reading Room underlined the technically dense nature of theories of inferential reasoning and the necessity of training in specific reading strategies to decode a text of Indian logic.

For more work by Dr Gorisse, see:

Gorisse, M. 2016. ‘Logic in the tradition of Prabhācandra’ in the Oxford Handbook of Indian philosophy, Ganeri, J. (ed.). Oxford: Oxford University Press

Pimps and Beggars: Poets through the lens of Sanskrit satirical verse

By Rohini Bakshi

IMG_4126SOAS Sanskrit Reading Room session by Dr. Lidia Wojtczak, Senior Lector in Sanskrit

In this provocatively titled session, building on an image from the Subhāṣitaratnakośa of the poet as a pitiful cātaka bird and the king as cloud, Dr. Wojtczak prompted us to re-examine our understanding of the poet-patron relationship in pre-modern India. The image posits the king as a generous patron who rains down rewards and the poet as part of an obsequious retinue dependent on the wealth that is showered. Dr. Wojtczak explained that the image gains traction with the establishment of kāvya, effectively overturning a previous relationship of ‘bard legitimising king’ which held sway in Vedic times. Through dhīḥ and manīṣā, Vedic poets are believed to have created the most exquisite compositions (brahman) extolling the devas (gods) who then bestowed sought-after gifts like victory in battle, wealth and progeny. Narāśaṃsa (the praise of men) was a marginal activity.

The gradual decline in Vedic ways of worship and the rise of kings and courts in epic times changed this balance. The praise of men, especially kings became important, but was left to sūta-māgadhas (charioteers and bards), vandins (praisers) and vāgmins (eloquent of speech).1 While these composers were essential to the courtly process, they most certainly did not share the status of Vedic kavis. Where did this leave royal poets in subsequent centuries? We see some continuity with epic times in Kālidāsa’s Raghuvaṃśa (17.15) which describes King Atithi at his coronation: “The moment he was praised by panegyrists (bandhibhiḥ)*, he looked as if [he had] grown in power, like a rain-cloud when it is saluted by cātakas.

Dr. Wojtczak quoted 6th-century alaṅkāra śāstras (systematic treatments of poetics) which register a balance of power between king and kavi favouring the poet as both a legitimising power and an immortalising one. In short, kings come and go but are only remembered through the works of an eloquent poet. Daṇḍin’s Kāvyadarśa (1.5) says “See here! Once it has obtained a mirror made of words, the reflection of the fame of the first kings does not perish, even when they themselves are gone” (trs. Wojtczak). Rudraṭa’s Kāvyālaṃkāra (1.4,5) adds, “When creating kāvya endowed with rasa, a great poet who is eloquent with blazing clarity will surely spread the fame of others the world over and till the end of the world.  Indeed, [the fame of all] that which has been made beginning with the abodes of gods, is destroyed over time. And so, not even the names of kings would survive, if there were no good poets” (trs. Wojtczak). Dr. Wojtczak is quick to clarify that these are normative texts about what poetry and poets should be like, and do not necessarily reflect reality.

At the other end of the spectrum are satirical works like those of Kṣemendra and Nīlakaṇṭha Dīkṣita which present the poet in a far more derogatory manner; as a grovelling pathetic sycophant; a fraud who praises what should not be praised, shamefully prostituting himself and trading in the sacred word. The ‘poet as cātaka, patron as cloud’ image has two ramifications. One of the poet as noble and proud – after all, the bird will not drink just any water, but only water from the rain-cloud directly from the sky. The other, of the poet’s tragic fate. He is totally dependent on that source because if it does not rain, the bird will die. Dr. Wojtczak cites images from Vidyākara’s Subhāṣitaratnakoṣa illustrating this:

Oh cloud! The cātaka, for long has fainted, 
stretching toward you his upturned beak 
in the unsupporting sky. 
And yet, so far from finding rain, 
he has not even heard a kindly word of thunder.2

The cātaka flies at the cloud; he hears it roar,
endures its hail, endures its waves of lightning. 
He shakes his wings and cries out, piteous; 
all for how small a drop of water.3

Harsh as these images of poet as a pathetic beggar are, they are less damning than his portrayal as a pimp who prostitutes the sacred word. Really contemptuous images are to be found for example in Nīlakaṇṭha Dīkṣita’s Kaliviḍambana. The work is a satire in 102 anuṣṭubh verses portraying the degeneration of various social groups in kaliyuga, including, of course, poets. According to Nīlakaṇṭha, kavis do not comprehend the essence of poetry; they do not understand guṇa (quality) and they abuse Vāṇī, the personification of speech. The following extract demonstrates:

Cowardice, vulgarity, parsimony, rudeness, want of acumen – 
poets, who are slaves for a handful of grain, can erase all of these [faults]. 
Poets who are eager to praise don’t need an impulse. 
Their tongues quiver if they are not eulogising something. 
Poets praise what has already been praised, 
by nature not understanding poetic quality. 
There is a certain insect called ‘bee’ – what is the use of additional description? 
Just one poem grants men a village, a horse, an elephant 
or, ultimately, food, clothes or betel. 
Some sell and others use frivolously the secondary form of Brahma known as ‘Speech’
which is embellished by being woven into a composition. 
Those men who use the Goddess of Speech 
to extol frauds once they have obtained her, 
are likely to yoke the wish-giving cow to a plough upon receiving her. 
Praising pretenders, distorting one thing into another, 
how can poets who claim the excellence of kāma even live? 4

Some of the most critical views on poets are to be found in the works of Kṣemendra, Dr. Wojtczak tells us. Quoting from his Kalāvilāsa, she says the criticism of men of letters is acute, their trade leading to dire consequences. Here one sees condemnation of poets who alter creation (sṛṣṭi) itself, much like Kṣemendra’s condemnation of poets who manipulate one of the primordial and most powerful forces of the cosmic order, Vāc, just for material riches.

By inspiring faith in others through eulogising excellence where there are defects, 
these boors, thieves of morality, masters of speech create a new world-order.5

One of Kṣemendra’s less known satires, Darpadalana goes one step further, describing the kavi as a pimp who prostitutes the goddess of speech for material gain.

Speech becomes a whore – decked with glorious ornaments and made into the instrument of others by poets motivated by profit who are in the service of kings.6

The session successfully pushed us to review our understanding of the poet-patron relationship in pre-modern India. The relationship as depicted in normative treatises on poetics was problematised. Far from the poet having the upper hand, or even a relationship of parity with his patron, it became clear that there were other dynamics in play. The financial dependence of the poet towards his patron could and did easily tip the equilibrium of the relationship. Dr. Wojtczak calls the satirical works a ‘rare case of textual counterbalance to normative treatises.’ She concludes ‘These satires were necessarily topical, crude, colloquial and possessed biting humour which may inspire wry smiles but had a tendency to leave a bitter aftertaste. The effect was exposing hypocrisies and punishing evils with the help of ridicule.’

(This blog is based on Dr. Wojtczak’s Sanskrit Reading Room session and an unpublished forthcoming paper.)
1 See Rāmāyaṇa 2.26.12 (Vulgate)
* bandin also written vandin
2 Subhāṣitaratnakoṣa 1080, trs. Ingalls 1965: 313, 314
3 Subhāṣitaratnakoṣa 1083 trs. Ingalls 1965: 314
4 Kaliviḍambana 33-39, trs. Wojtczak
5 Kalāvilāsa 9.34, trs. Wojtczak
6 Darpadalana 3.10, trs. Wojtczak

Working with Sanskrit sources: theory and method

By Karen O’Brien-Kop

This Sanskrit Reading Room was a session with a difference. Titled the Graduate Sanskrit Reading Room, the session shared the work of three members of the SOAS research community, who each work on Sanskrit sources and who reflected on theory, method, and methodology. The presentations were each very different, but together they gave a broad sense of possible approaches to and discourses on Sanskrit texts in current scholarship.


Olivelle’s translation of the Upaniṣads

First up was Graham Burns, a PhD candidate in Religions and Philosophies, who spoke on ‘Formulation, Substitution and Veneration: Three Controversial Upaniṣadic terms’. Using a philological approach he contextualised brahman, ādeśa and upāsanā, contrasting their later Vedānta-inflected interpretations with less clearly defined semantic fields for their early usage. Brahman, a prominent term in the Upaniṣads, is widely used in Indian philosophy to designate a  ‘supreme universal force’. It was commonly believed to have derived from the root bṛh (to ‘expand’ or ‘burst forth’), until Thieme’s suggestion in the 1950s that the word was in fact derived from the root brah (‘to formulate’), reflecting the meaning of brahman in earlier Vedic texts as a ‘hymn’, or ‘verbal formulation’. This prompted re-readings of the meaning of brahman in the Upaniṣads and the recognition that brahman in Vedic contexts was polyvalent, often denoting the sacred verbal formulation or its power rather than the ‘ground of being’. Next, Burns moved on to the term ādeśa, examining how it was translated by Olivelle, probably the best known translator of the Upaniṣads at the current time. Burns suggested that Olivelle’s translation of ādeśa as ‘the rule of substitution’, in line with Pāṇīni’s grammatical formulation, may be anachronistic and gave us an example from the Chanḍogya Upaniṣad.

ekena mṛtpiṇḍena sarvaṃ mṛṇmayaṃ vijñātaṃ syāt […] sa ādeśa bhavati (CU 6.1.4-6)

‘by means of just one lump of clay one would perceive everything made of clay […] That […] is how this rule of substitution works’ (trs. Olivelle   1996: 148)

A common interpretation of this passage is that clay and the items made of clay are ontologically identical, a view which can perhaps be supported by the ‘rule of substitution’ translation of ādeśa.  However, Burns suggested that this interpretation might be over-dependent on the later Pāṇīnian understanding; in fact, here the term might be making the simple point that, by observing the properties of clay in one form (e.g. clay earth), one can understand how it would behave in other forms (e.g. clay pot). So rather than indicating an ontological identity, ādeśa in this instance has its more generic meaning of ‘teaching’ or ‘rule’. The final term was upāsanā, often translated in Advaita Vedānta circles as ‘meditation’ but which, in its more literal sense of taking one thing as another, is closer to the idea of veneration or worship. This again suggests a less non-dualist interpretation than is often supposed. This meaning is also connected to one of the basic strategies of the Upaniṣads, which was to identify correlations or patterns in the universe. The bottom line of the talk was how later non-dualist Vedānta interpretations can often elide both the semantic liquidity and the technical specificity of these terms in the Upaniṣads. The best interpretive strategy is to read such terms carefully in their own context, and, of course, contexts vary across the Upaniṣads.


A traditional Vedic chanting class in India

Next, Rohini Bakshi – SOAS graduate, Sanskrit teacher and writer – led an interactive session on Vedic chanting (mantroccāraṇa or mantraparāyaṇa). She explained the slow progress and perseverance required to learn Vedic chanting from her guru in South India; indeed, it can take a full-time student up to sixteen years to complete Vedic studies (veda-adhyayana), so long as one is dedicated to no other pursuit. Bakshi’s access to this type of training is rare for a woman. Focusing on the saṃhitā method of chanting – i.e. recitation of verses in their entirety as they were ‘heard’, or śruti – she guided the group through two verses from the Taittirīya Āraṇyaka in the style of the Taittirīya Śakhā (branch) of the Yajurveda. This entailed the teacher leading the chant, followed by the group reciting in unison and then singly. The focus of the session was on the unique Vedic notational rules for correct pronunciation. This included much fine-tuning of our retroflex pronunciations (, , , and ṭh). At other points, she laid into our timid efforts with warm gusto: ‘C’mon, spit it out’ she urged when it came to the aspirates. We learned of special rules for Vedic chanting. For example, the anusvāra when followed by a consonant must be pronounced as its class nasal – thus ūrdhvaṃ jigātu is pronounced ūrdhvañ jigātu, while mukhaṃ kim will be mukhaṅ kim. Another interesting rule was that, when followed by two consonants, the anusvāra is pronounced as ‘g’ – and so oṃ svāhā becomes og-svāhā. We also learned how Vedic accents denote tonal changes up (udātta) and down (anudātta). Although it was just a brief introduction, by the end of the session many of us were eager for more teaching from Rohini Bakshi– and fortunately for us she is starting a Vedic chanting class in London at the City Lit in 2018:


Canti d’amore indiani (“Indian love songs”, 1930), a typical miscellaneous anthology of second-hand translations of Sanskrit erotic poetry into Italian verses

The third speaker, Maddalena Italia, who is completing her PhD in the School of Languages, Cultures & Linguistics, turned our attention to a selection of early 20th-century French, Italian and English translations, re-translations and transcreations of the 7th-century Sanskrit poetry anthology, Amaruśataka, or ‘The Hundred Verses of Amaru’.  Her readings included an English free-verse translation by Powys Mathers, French renderings by Chézy and Toussaint, and Italian versions by Ghiron and Norsa. Although some of these European translations conformed to type in their projections of ‘Eastern exotic erotica’ and in their rather free and creative approach to ‘translation’ (most of these translators did not have any knowledge of Sanskrit), Italia’s talk focused on other specific features of these collections – the deliberate omissions and fabrications of verses. She traced how one particular passage in the Amaruśataka that described a sexual posture was omitted by almost all European translators, in what appeared to be an attempt to censor the poetry for prudish European audiences. This is an example of the ‘vociferous silences’ of her talk’s title. On the other hand, several other translators fabricated and added verses that reified the notion of the ‘exotic Orient’. Carefully tracing the genealogy of influence between this group of European translators, she began to unravel the complex weave of creative translational layers that removed the poetic material ever further from the original Sanskrit text. Indeed, the creative embroidery of these European translations crossed so far into transcreation that they almost took on a separate life from the Amaruśataka. Italia also noted the tonal difference between Italian and French translators, who were willing to include and elaborate erotic details, and English translators, who tended to downplay and erase what they saw as explicit detail. Because the philological method prioritises the establishment of  ‘pure’ and ‘spurious’ readings of a text, applying the genealogical-reconstructive philological method to these European translations would be a futile task. Such is the degree of ‘contamination’ of the original ‘pure’ Sanskrit material that it renders these European translations as ‘pure contaminations’, for which there can be no philological remedy. Rather, Italia concluded, it is ‘inventive muddling’ that is the very defining feature of these literary translations.

For more on these scholars’ work, see:

Graham Burns:

‘Stories of the Self: The Exploration of Reality Through Narrative in the Vedic Upaniṣads (‘Three Narratives in Search of a Principle’)’ at

Rohini Bakshi:

Maddalena Italia:

(Forthcoming doctoral dissertation 2018) ‘The Erotic Untranslatable: The Modern Reception of Sanskrit Love Poetry in the West and in India’

A 16th-century Nibandha on Vaiṣṇava Devotional Procedure

By Avni Chag

For the fourth Sanskrit Reading Room session of this year we were joined by our first specialist in Vaiṣṇava studies, Dr. Måns Broo, all the way from Åbo Akademi University, Finland. The session took place on Wednesday 15th November at SOAS. Dr Broo presented on a 16th-century Sanskrit Vaiṣṇava ritual text, the Haribhaktivilāsa, also known as the Bhagavadbhaktivilāsa. The text is one of the earliest texts of the Gauḍīya Vaiṣṇava texts and historically important for this tradition. The verses are said to have been written by Gopāla Bhaṭṭa Gosvāmin, a devotee of Caitanya, and the Digdarśinī commentary by Sanātana Gosvāmin. It is a voluminous nibandha, and a tantric text, which as Broo explained, not only describes ritual practice but also the fruits of that practice, and has played a significant role in developing ritual procedures within the sampradāya. Generally the text dedicates 3 to 4 verses to a particular procedure, followed by a selection of verses on the procedure’s māhātmya, i.e. what great fruits the practitioner can expect from the particular ritual. We learned, to our surprise, that most procedures were directed to also members that fell outside the 3 twice-born varṇas as well as women. Like most texts written in the early stages of a developing sampradāya the Haribhaktivilāsa is orthopractic, referencing authoritative treatises, especially the Veda, whilst also incorporating something new, something specifically Vaiṣṇava.

The devotional and ritual procedures the text covers are varied. With a total of nearly 10,000 verses the text describes in various degrees of detail how a Vaiṣṇava devotee ought to:

Honour one’s guru

Self-purification in preparation of worship

Offer worship to Kṛṣṇa

Reciting mantras

Reciting the 1000 names

Read the scriptures and narrate Kṛṣṇa’s līlās

Perform rituals at various junctures of the day

Shun bad company

Make offerings

Perform arcana

Perform vratas

Installing images in a temple


Partake of offerings made to the deity

Prostrations and circumambulations etc. etc.

Broo’s work in putting together a critical edition of the text is the first comprehensive one of its kind as most earlier editions are not critical editions. These earlier editions do not allow the scholar to understand what the text might have originally looked like or how it might have changed during its dissemination throughout India. His efforts in producing a critical edition involved collating as many manuscripts as are available, a process that he lamented had not been straightforward. The catalogus catalogorum records around 100 manuscripts, many of these unfortunately are not available or are not in fact manuscripts of the Haribhaktivilāsa. He has also been able to locate many manuscripts that do not appear in the catalogue. Getting copies of manuscripts has not always been an easy process; even the selected archives in India that have digitised collections are not always so interested in sharing them.

Broo has been collecting manuscripts for quite some time now and has collected near to 100. Manuscripts are located all over India, as far west as Gujarat, or north as Panjab and Assam, down to Orissa etc. This was the product of the rise in manuscript culture during the Mughal period, when paper was comparatively cheap and readily available. Most of these are not dated, but those that are were copied in the 19th or early 20th centuries. The oldest dated manuscript he was able to secure is from Jaipur, dated 1743 CE:


Other early manuscripts include one from 1770, from Patna:


One from the same year, from Orissa:


Broo interestingly brought us back to Dr. Formigatti’s session from last year on manuscript marginalia with this particular manuscript. While the marginalia, he explained, were often glosses of particular words in the manuscript, they also referred to various other treatises and well known interpretations, something that might help us to understand more about the scribe or the particular manuscript’s use.

And another manuscript from Calcutta, dated 1790:


We do not know much about the scribes because they generally did not include their names. Broo has found that for those manuscripts in which scribes did sign off their work, though many were brāhmins, who probably copied manuscripts for a living, some were Vaiṣṇava devotees, recognised by their dikṣā name ‘dāsa’.

The edition we read from in the session used 11 manuscripts, which were chosen on the basis of age, geographical spread and interesting particularities that have affected further discoveries to do with the text. We began right at the beginning of the text from the first Vilāsa, here meaning ‘types of bhakti or devotional past times’. Although the first vilāsa didn’t quite give us a taste of the codified ritual aspects of the Gauḍiya Sampradāya, it showed us how the Digdarśinī gloss reports on the mūla text. We were looking for how, despite the understanding that the gloss is an auto-commentary or was written by a close companion of the author, it sometimes reports curious, alternative readings of the mūla text or uses a different text than the received one. In the gloss of verse 2, for example, the commentator leaves out the names of two of the gurus honoured in the verse, only mentioning Ragunāthadāsa:

bhakter vilāsāṃś cinute probodhānandasya śiṣyo bhagavatpriyasya |

gopālabhaṭṭo ragunāthadāsaṃ santoṣayan rūpasanātanau ca ||2||

…śrīraghunāthadāso nāma gauḍakāyasthakulābjabhāskaraḥ paramabhāgavataḥ śrīmathu-

rāśritas tadādīn nijasaṅginaḥ santoṣayitum ity arthaḥ ||2||

This might tell us something to do with the author of the commentary – did he perhaps consider Ragunāthadāsa the most important of the three, or did he feel that his non-brahminical background needed to be explained?

Another interesting aspect of the text, or what we learned is a typical feature of many Vaiṣṇava texts, is the common use of the ablative, such as the suffix –taḥ. For example, verse 1 ends with samastaśāstrataḥ (from all the śāstras).

The first volume of Dr Broo’s work on the Haribhaktivilāsa will hopefully be available next year, the first of what will probably be a series of volumes. Here is a link to Dr Broo’s most recent text critical work on the Rādhā Tantra.

Graduate Sanskrit Reading Room

Wednesday 29th November, 2017, 3-5pm, Room T102, 22 Russell Sq

Session 5: Graduate Sanskrit Reading Room


Translation immoral? Contamination, hybridity, and vociferous silences in

early twentieth-century translations of Sanskrit erotic poetry

Maddalena Italia (SOAS)

This paper focuses on a selection of early twentieth-century French, Italian, and English translations, re-translations, and trans-creations of the 7th century Sanskrit anthology titled Amaruśataka, or “The hundred verses of Amaru”. The practice of comparing, collating and inventively muddling variants from different sources is the defining feature of these literary translations. Indeed, most of their authors lacked the linguistic skills to access the original Sanskrit text, which was only available to them through earlier translations; what is more, each translator creatively ‘contaminated’ the text with his own notion of what the ‘exotic erotic’ should be like, going so far as to insert forged passages and excise what was not perceived as aesthetically or – more often – morally appropriate. My aim is to follow and unpack the selective, comparative and combinatory process that produced each modern re-translation or trans-creation. In this perspective, what I call ‘translational contamination’ reveals its creative rather than polluting force: paradoxically, it is often the hybrid, if not outright spurious, details of these modern versions that act as a litmus paper of the aesthetic and moral preoccupations of their authors and projected readers, thus turning such texts into new and newly translatable originals.

Bio: Maddalena started working on her doctoral thesis in 2013, after completing her MA in Languages and Cultures of South Asia at SOAS, preceded by a BA and an MA in Classics. In her PhD thesis (“The erotic untranslatable: the modern reception of Sanskrit love poetry in the West and in India”), Maddalena analyses commentaries, translations, and rewritings of Sanskrit erotic poetry authored by Western and Indian poets and philologists.

Vedic Chanting and Demonstration

Rohini Bakhshi (Independent Scholar)

The saṃhitās of the four Vedas have been transmitted faithfully for millennia not through the written word, but orally through rigorous and varied recitation styles. Even today, traditional paṇḍits reject printed copies as potentially faulty, and rely exclusively on the unbroken oral transmission of their own school.  Each Veda has its own method of chanting. In modern India it is the taittirīya śākhā of the Yajur veda which has reached lay audiences more than other schools. The sound of Veda being chanted is considered mystical and sacred by most Hindus even if they do not understand a word of what is being said. The method of even the most basic chant is extremely challenging. It calls for multiple and simultaneous application of attention (avadhāna). It can take a full-time student dedicated to no other pursuit up to sixteen years to complete Vedic studies. Traditionally this has been and continues to be exclusively a male domain. While there have been reformers who allow women to learn/chant the Vedas (adhyayana) Hindu orthodoxy prohibits it even today. There is no provision for girls to do veda-adhyayana in traditional gurukulas. Even ‘part-time’ engagement is discouraged actively. As a female, finding a traditionally trained brāhmin willing to teach Vedic chanting is no easy task. Having said all that, technology has facilitated access to the few who are willing to teach. In this session of SRR, I would like to share my lived experience of learning how to chant Veda from a teacher from rural Karnataka – over Skype. The challenges, the disappointments, but most of all the exhilaration.

Bio: Rohini Bakshi is a Sanskrit teacher and columnist with an M.A. in Hindu Studies from SOAS. She has been popularising the study of Sanskrit using social media for the last five years. Her platform #SanskritAppreciationHour reaches over 14,000 followers on Twitter. Having authored an independent reader for Sanskrit students, she is currently working on a book about the early history of Rudra-Śiva. A devout Hindu, she campaigns for women’s rights based on her study of ancient Hindu texts. Oxford alumna, she returned to academia after a successful career in marketing communications spanning twenty years.

Formulation, Substitution, Veneration: three controversial Upanisadic terms

Graham Burns (SOAS)

Later philosophical interpretations of the Vedic Upaniads (compiled c. 700 BCE to the very early years of the Common Era) often proceed on the basis of specific technical understandings of certain important terms. But are those understandings necessarily justified? In this session, Graham Burns will look at three important Upaniṣadic terms, exploring their etymology and background, and considering whether the way in which they were later interpreted helps or hinders understanding of what the Upaniads themselves were trying to say.


Bio: Graham Burns holds a BA in Law from Durham University and an MA in Religions from SOAS. He is currently in the final year of work in the Department of Religions and Philosophies at SOAS on his doctoral thesis: ‘Neti neti: the Search for the Ultimate Principle in the Vedic Upaniṣads’. He has taught yoga professionally since 2001 and helped train yoga teachers since 2004, and has also lectured at SOAS on the ‘Origins and Development of Yoga in Ancient India’ and ‘Hinduism: Foundation’ modules.


To book your space, please contact Ruth Westoby:


















Room T102, 22 Russell Square, SOAS


An audio-recording of the session will be made.



To book your space, please contact Ruth Westoby:




Don’t cross the line

Critiquing a Critical Edition: The Anāthapiṇḍadāvadāna with Dr Camillo Formigatti

On a crisp November afternoon the Sanskrit Reading Room was hosted by the Bodleian Library in Oxford. Under the careful guidance of Dr Camillo Formigatti, the John Clay Sanskrit Librarian at the Bodleian, we read the critical edition of the Anāthapiṇḍadāvadāna, deconstructed the text with the use of additional manuscripts, and were treated to a fascinating manuscript viewing.


Manuscript 1: Kāraṇḍavyūha (13th or 14th century CE) on palm leaf and Svayambhū Pūrana (18th century CE) on paper.

The genre of the Anāthapiṇḍadāvadāna from the Avadānaśataka is the jātaka and avadāna tales – the stories of the previous lives of the Buddha. The Avadānaśataka is one of the most ancient of its kind, the earliest forms dating to the early centuries of the common era. The form of these tales present three levels: a story of the present, a story of the past, and the conjunction of the two which reveals the moral.

During the reading Camillo engaged precisely what is presented in the manuscript rather than making automatic adjustments in conformity with our expectations. He drew attention to how previous studies – of Pāṇini, of kāvya – might influence the choices we make in reading manuscripts. We saw the use of virāma in the manuscript for the purpose of punctuating the text where a danda would have been too decisive. J. Speyer, in his critical edition of 1902, chose to use a danda.

A lively discussion sparked about the interpretation of an apparently simple passage, in which Anāthapiṇḍada fills a basket with… what? Dr Formigatti provided this blog with further detail on the debate: ‘The Sanskrit goes as sa hiraṇyasuvarṇasya helāṃ pūrayitvā, and several suggestions were made as to how to interpret the compound hiraṇyasuvarṇasya. The first one is that Anāthapiṇḍada filled a basket (BHS helā) with hiraṇyasuvarṇa, the second one is that the basket was made of hiraṇyasuvarṇa. But what exactly does the compound hiraṇyasuvarṇa mean? Other suggestions were made: does it mean ‘coins and gold’ or ‘gold and coins’? Or maybe ‘gold and silver’? Or does it even denote a specific type of gold? Apparently, the most plausible answer is that hiraṇya means ‘coins’ and suvarṇa simply ‘refined (but unminted) gold,’ and that Anāthapiṇḍada filled a basket with them. According to the Nachtragswörterbuch des Sanskrit (NWS), the word hiraṇyasuvarṇa (corresponding to Pāli hirañña-suvaṇṇaṃ) can mean both ‘coins and unminted gold’ (‘Goldmünzen und ungeprägtes Gold‘) as well as ‘silver and gold’ (‘Silber und Gold’). Unfortunately, the Tibetan translation of hiraṇyasuvarṇa with gser dngul shares the same ambiguity of meaning with the NWS, and is therefore of no help. (Moreover, in order not to make any mistake, the Tibetan translators decided to render the Sanskrit helā with gser gyi gzhong pa, ‘golden bowl’!) In order to reach a decisive interpretation it is necessary to consult also Oliver Hellwig’s rasavidyā dictionary, where the second meaning provided in the entry hiraṇya is ‘a gold piece or coin (generally with suvarṇa as opp. to base metal).’ However, since both hiraṇya as well as suvarṇa can mean both simply ‘gold,’ but also ‘coin,’ it is also possible that the compound hiraṇyasuvarṇa simply means ‘gold coin’, as suggested by the entry in the rasavidyā dictionary. Bottom line, without a careful perusal of all passages in which this compound occurs, we cannot be sure of its precise meaning.’

Familiarity with Classical Sanskrit was not enough to unravel the clues in the manuscripts. As well as unpicking the (perhaps less familiar) forms of Buddhist Hybrid Sanskrit, Camillo’s expertise in Tibetan enabled him to elucidate the texts. The 9th-century Tibetan translation adds significant detail to Speyer’s critical edition. As well as including stock phrases used also in other avadānas, the Tibetan includes a key segment upon which the plot hinges (and not included in CUL MS Add.1611 used by Speyer as the basis for his edition): that in the previous life-time recounted by the Buddha he reneges on an agreement to provide security for a debt incurred by his friend whom he vouched for during a gambling match. This is the evidence for the samavadhāna – the comparison the Buddha makes between the past and the present – and the reason he cannot cross the line, the lekhā.

To conclude the session Dr Formigatti showed us an array of manuscripts to indicate different writing technologies across Asia. The manuscripts appear delectable to any bookworm and I was astonished to learn that the yellow of some was due to treatment with arsenic, intended to dissuade insects from wreaking destruction. Even more delightful was the revelation (at least to me) that sometimes turmeric was used by scribes to trick patrons into believing arsenic had been applied.


Manuscript 2: Prajñapāramitā Sūtra (1880) on arsenic-treated yellow paper.

MSS 3.jpg

Manuscript 3: Muva Jātakaya (1864), Sinhalese manuscript with a sublimely illuminated incarnation of the Buddha as a gazelle. This Sinhala versification of the Nigrodhamiga Jātaka is incised with a stylus on palm leaf, which is passed over with soot, then cleaned.


Manuscript 4: Kāraṇḍavyūha (1050 CE) on palm leaf.

MSS 5.jpg

Manuscript 5: Japanese facsimile of the Horiuzi (Hōryū-ji) manuscript (beginning of the 7th century CE), produced in the 19th century for Max Müller’s edition of the Prajñāpāramitāhṛdayasūtra.


Manuscript 6: Collection of containing Shaiva Stotras and texts of Sāhib Kaul and his pupils, written by Dīlārāma (18th century) from Kashmir, in śāradā script. The second part of the manuscript contains notes as in a personal notebook. The layout of this form of annotations may be influenced by the Persian culture in which manuscripts may be heavily annotated in diagonal lines.

MSS 7.jpgManuscript 7: A Vasudhārā dhāraṇī manuscript in Jain nāgarī [1663 CE]: a Buddhist dhāraṇī transformed by Jainas to fit their purpose. It has the typical Jaina ornamented stringhole in a geometric pattern. The decoration has now faded but it would once have been bright red.

In his first session for the Sanskrit Reading Room in May 2017 Dr Formigatti steered participants through the meticulous study of marginalia in Sanskrit manuscripts. For a detailed write up of this session please see Rohini Bakshi’s blogpost at

To explore Dr Formigatti’s work further please see:

‘Walking the Deckle Edge: Scribe or Author? Jayamuni and the Creation of the Nepalese Avadanamala Literature’, in Buddhist Studies Review33(1-2), 2016, p.101-140.

‘Towards a Cultural History of Nepal, 14th – 17th century. A Nepalese Renaissance?,’ in Studies in Honour of Luciano Petech. A Commemoration volume, 1914-2014. Rivista degli studi orientali, vol. LXXXIX, Roma 2016, p. 51–66.6.

‘A Forgotten Chapter in South Asian Book History? A Bird’s Eye View of Sanskrit Print Culture,’ in Printing as an Agent of Change in Tibet and Beyond, Leiden: Brill, 2016, p.72–134.8.

‘Travelling Books,’ in Buddha’s Word. The Life of Books in Tibet and Beyond, Museum of Archaeology and Anthropology, Cambridge 2014, p. 19–24.

An audio recording of Dr Formigatti’s session will be available soon.

Ruth Westoby is a first-year doctoral researcher at SOAS and a co-organiser of the Sanskrit Reading Room.