Economics specialist Bibek Debroy, who has translated the Mahabharata and the Ramayana into English from the Sanskrit, was in conversation with writer Avik Chanda at the CANTO Poetry Festival 2023, an international multilingual multicultural travelling poetry festival curated by Chanda, which was held across New Delhi and Kolkata. Excerpts from the conversation:

The first question that strikes one, from a purely layman’s standpoint, is when one is talking about the Mahabharata from a literary standpoint and is contrasting or juxtaposing that against the other epics of the canon – what are some of the literary nuances that stand out? Either in the way the narrative is shaped or the way the drama is built up or the non-linear style of story-telling dating back two and a half millennia? Are there significant literary nuances that stand out, distinguishing it not just from other epics in the milieu but across the entire span of world literature?
I translate from Sanskrit to English. The title of our session says “Fording the Epics” – the word “Epic” is normally used for Itihasa which is the Valmiki Ramayana and the Mahabharata. I prefer to think of it as the Itihasa Purana corpus, because the Puranas are connected certainly to the Mahabharata. Again, just purely for the record, my intention is to translate all the 18 Mahapuranas. By translate, I mean unabridged translations. To give some idea of the rough size that we’re talking about, I’d like to mention that the Mahabharata is believed to have 1,00,000 shlokas.

My translation ran into 10 volumes, which is about 2.25 million words. The 18 Puranas together account for 4,00,000 shlokas, which is about 12 million words. Translations of the Bhagavata Purana, the Markandeya Purana, the Brahma Purana, and the Vishnu Purana, have already been published. The Shiva Purana will be published next month. The Brahmanda Purana is pending with Penguin and I am on the Matsya Purana now.

Now, before we talk about the Mahabharata and since this festival is on poetry, let me first speak a bit about poetry. All of these texts we just mentioned were all poetry. Prose, as we understand it, came into Sanskrit literature much later. Whenever we think of poetry, often but not invariably, we tend to think of rhyming. Sanskrit poetry was not like that. Rhyming came to Sanskrit poetry much later. The rhyming, essentially, started with Jayadeva and the Vaishnava poets which was much later.

Most of us have heard of the Dashavatar Stotram:

“pralayapayodhijale dhṛtavānasi vedama
vihita vahitra caritram akhedam” 

“Victory to Keshava, the lord of the universe,
Who assumed the form of a fish”

There is a rhyme here, but that is due to Jayadeva’s poetry.

Sanskrit poetry was defined in terms of chhanda (metre). When we were in school, teachers would tell us about “badi ee” or “chhoti e”. Now, that is because we don’t pronounce correctly. If we pronounce it correctly, there is no need to qualify these sounds as short or long. For all Indian languages including Sanskrit, you must open your mouth. When we speak English, we learn to speak with the mouth relatively closed. The first thing you need to understand about Sanskrit is that you need to open your mouth while speaking it.

Akshara(s), which are syllables with a single vowel sound, were divided into hrasva (short) and deergha (long), depending on whether the vowel sound was short or long. Then, you would begin to classify chhanda(s) or metre, which depended on how many sounds or akshara(s) were there in a line (pada). Typically, there were four lines in a shloka. Depending on the number of akshara(s) in a line and the permutation and combination of the short vowel sounds and the long vowel sounds, there were more than 1,200 different chanda(s).

The Valmiki Ramayana tells us that Valmiki was the “Adikavi” (First Poet) and that he composed the first shloka. As everyone knows, he wrote the first shloka when he saw the hunter shoot a pair of birds:

“mā nisada pratistham tvamagamah sasvatih samah
yat krauncamithunadekamavadhih kamamohitam” 

“Oh hunter, may you repent for life and suffer, find no rest or fame, for you have killed one of the unsuspecting, devoted and loving krauncha couple.”

The Valmiki Ramayana tells us that this is the reason this came to be called a “shloka”, which comes from “shoka” or sorrow.

Valmiki, as a poet, was a class apart. Valmiki’s work was beautiful poetry in terms of describing nature, just like Kalidasa. When we talk of the Mahabharata, as we are today, and the Puranas, these were believed to have been composed by Veda Vyasa. He was a completely different kind of poet. He didn’t waste his time in describing nature. So-and-so came here and did this, Avik came and drank coffee, Bibek Debroy refused to drink coffee…very matter-of-fact. Nonetheless, it was beautiful poetry.

One of the most common chhanda(s) is anushtup. We don’t even realise it but anushtup is the structure of the simplest chhanda. The structure of anushtup is eight akshara(s) per line. I won’t get into the permutation and combination that goes along with this. The first shloka of the Bhagavada Gita:

“dharmakshetre kurukshetre samaveta yuyutsavah
mamakah pandavascaiva kimakurvata samjaya”

“O Sanjay, after gathering on the holy field of Kurukshetra, and desiring to fight, what did my sons and the sons of Pandu do?”

You must have heard this hundreds of times without realising it is anushtup. In our mantras, we constantly use anushtup without realising it. For example:

“vakratunda mahakaya suryakoti samaprabha”

“O lord with the curved trunk, large body and the splendour of a million suns.”

Or even,

“ya devi sarvabhutesu”

“O deity, who abides in all beings.”

All of these are anushtup. It has influenced non-Sanskrit poems and songs greatly, even though we don’t realise it.

“Dhono dhanno pushpe bhora…” 

“Fields of golden grain, meadows of fragrant flowers…”

This is also anushtup.

They varied the chhanda depending on the theme. For example, this is also one of the most famous shlokas from the Bhagvada Gita:

“yada yada hi dharmasya glanirbhavati bharata
abhyutthanamadharmasya tadatmanam srjamyaham” 

“Whenever there is a decay of righteousness, o Bharata, and a rise of unrighteousness, then I manifest myself.”

As you can see, this is different. The poet chose it depending on the theme. For example, in Meghadutam, Kalidasa wanted to mirror the cloud. He alternated between short bursts and long bursts which gave rise to the Mandakranta metre.

“kascit‍ kan‍tavirahaguruna s‍vadhikarat‍pramata”

“Where Ramagiri’s cool dark woods extend.”

You can see that the chhanda is different.

The Mandakranta metre with which most people are familiar is:

“santakaram bhujagasayanam padmanabham suresam”

“A prayer for the one who has a calm appearance, who rests on a snake, who has a lotus on his navel and who is the god of the deities.”

You can see that this is visibly different from anushtup. One of the great things about the Mahabharata, which no translation can bring out, is the superlative poetry.

Just a quick question on the anushtup, Bibek. Are the other Puranas also in anushtup chhanda, and what chhanda does the Ramayana which preceded the Mahabharata, follow?
In the Itihasa Purana corpus, 80 per cent of the shlokas will be in anushtup, which constitutes eight aksharas. Sometimes, trishtup is used, which constitutes 11 aksharas. Essentially, anushtup, and variants of anushtup or trishtup – and, as I said, if you wanted to introduce a discontinuity you varied it. For instance, the Vishwaroop Darshan in the Bhagavada Gita is a completely different kind of chapter and, thus, the chhanda is completely different.

When we are talking about an unabridged translation, shloka for shloka, it is exceptionally important to talk about fidelity. But, even before we go there, let us maybe talk a little about authenticity. If I’m not mistaken, you have based your translation on the critical edition from the Bhandarkar Society. Now, again please correct me if I’m wrong, but right at the end of the Mahabharata, it says that whatever there is in the world is in the Mahabharata, and whatever isn’t there in the Mahabharata is not in the world. Similarly, I think the very next page or the preceding one says that there are 1,00,000 shlokas in the Mahabharata. But, the Bhandarkar edition has fewer than that, maybe 75,000 or 80,000 shlokas. Are there possibly other extant versions of the Mahabharata which are even more “authentic” and can account for these missing 20,000 shlokas?
I think I need to explain what the critical edition did. The Valmiki Ramayana is, in a sense, a single manuscript. There is a Northern version and a Southern version but, essentially, it is a single manuscript. Since someone was unhappy with what Valmiki depicted, he or she composed another Ramayana like the Yugavashistha Ramayana or Adhyatma Ramayana. Those of us who are sufficiently old and used to travel on trains will remember hold-alls. The Mahabharata is a bit like a hold-all. It subsumed everything. Consequently, there were about 1,215 different versions of the Mahabharata floating around the country. All of these are in Sanskrit. I’m not talking about non-Sanskrit versions, those are different. Each of these 1,215 versions varied from the others in minor details.

Did Veda Vyasa actually dictate to Ganesha or not? When Draupadi was being disrobed in the sabha, did she pray to an anonymous divinity or was it Krishna by name? These weren’t substantive differences, but subtle differences.

In 1916, the Bhandarkar Oriental Research Institute was set up in Pune with the specific task of sifting through these 1,215 versions to see if they could, in some sense, determine what might have been the original. The Bhandarkar Oriental Research Institute never claims that anyone is able to ever identify the original Mahabharata. That is impossible as there is subjectivity. What they did was very similar to what we did in school when we studied maths – the highest common factor. If a shloka occurred in a large number of manuscripts, it was probably original.

This kind of thing is also done in genetics. Thus, it is pretty standard. Having done this, they produced the critical edition, which took them 50 years. The last volume of the critical edition was published in 1966. There is a text called the Harivamsa, which is not exactly a Purana but more of an appendix to the Mahabharata. They took another four years to complete the Harivamsa, which was published in 1970. Bhandarkar never claims this is the original, because it is impossible to make this claim. Furthermore, Bhandarkar has not only published the critical edition but has also published all the other variants. If you want the other variants, you will get them courtesy of Bhandarkar.

It is a popular belief, as you said and I myself said earlier, that the Mahabharata has 1,00,000 shlokas. The critical edition has just over 75,000 shlokas. If you add the Harivamsa, it comes to about 80,000. Thus, it has removed some shlokas and there are people who have criticised the critical edition for omitting certain shlokas. But, as I said, if you want all of them, Bhandarkar has published all of them.

This is a question about language. The genesis of this question is pure happenstance. We were doing a presentation at the EACPM last year and we were talking about behavioural economics, behavioural sciences, resilience, stress management, alienation, and workplace well-being. You had mentioned in passing that there is no word for boredom in Sanskrit and I asked the obvious question – why not? Presumably, back in the day, people did not feel bored and there was no emotion around boredom. Now, if you talk about the entire spectrum of emotions and feelings that characterise the human condition today, be it stress, boredom, alienation, loneliness or depression, some of these may have direct equivalents in Sanskrit and some may not. Would it be fair to say that we can look at language per se and use the words extant in a vocabulary at a particular period of time as an index of the human condition in that period?
We certainly can. Since you’ve mentioned language, there are some words that are simply untranslatable because whatever you do you can’t really translate them satisfactorily into English. That is obvious enough. A word like “dharma”, for example, is something I can’t translate adequately. Or perhaps a word like “varna”. I am asked this question sometimes and this is a simple question to answer. But there is a more complicated response and let me mention this as I don’t think it is adequately appreciated.

The structure of English or the nature of English is completely different from the nature and structure of any Indian language. I don’t mean it in the sense of Sanskrit being a loose language and, thus, the subject can be placed anywhere or the verb can be anywhere – the fact that the structure doesn’t have to be subject-verb-object in that order. I don’t mean that, I mean something a little more profound.

In Sanskrit, the verb is the most important thing. The nouns or the proper nouns are secondary. The meaning of the noun or the proper noun is derived from the meaning of the verb. For example, take a word like “ajagara”. Anyone would say that “ajagara” is a python or a boa constrictor, as it really is. But, anyone who knows Sanskrit would say that “gara” means to swallow and “aja” is a goat. Thus, anything that swallows a goat is “ajagara”. It doesn’t necessarily have to be a python or a boa constrictor. For anyone who knows Sanskrit, the verb is always there at the back of the mind.

Out there, I see a tree and one of the words for a tree is “padapaha”. Any Westerner or a Western-educated person will say that “padapaha” means tree. But, what is “padapaha”? It is something that drinks water with its feet. So it is conjuring up not just a vision of a tree but the act of drinking water with the feet.

One of the words for chair is “asandaha”. When I hear “sandaha”, it is no longer just a chair but something that is giving me a posture. If you ask a person who is not very westernised their name then that person will not say, “I am Avik”. The I in the capital is always at the back in our language system. They will say “I am known as Avik” or “My name is Avik”. Often Westerners make fun of this. This is because Sanskrit words have multiple meanings. Of course they do. You have not understood the basics unless you understand that it’s the verb that is important.

Let me give you an example of how this causes slight problems in translation. There is a famous shloka, but I’m not concerned with that one. I’m concerned with the one that comes afterwards.

“vasamsi jirnani yatha vihaya
navani grhnati naro’parani
tathā sarirani vihaya jirna
nyanyani samyati navani dehi”

“Just as a man casts off his worn out clothes and puts on new ones, so also the embodied-self casts off its worn out bodies and enters others which are new.”

This is from the Bhagavada Gita, everyone has heard this. This is the next shloka:

“nainam chindanti sastrani nainam dahati pavakah
na cainam kledayantyapo na sosayati marutah”

Let us translate this: “Weapons cannot cut it, fire cannot burn it, water cannot wet it and wind cannot dry it.” This would be an immaculate translation. But what this translation has not told me is what I have…I have “shastra”, which means weapons. What were they made from? Metal. Where did the metal come from? Earth. So I have earth, fire, water, and wind. It is four of the Panchabhuta with the fifth one (akasha – space/atmosphere) which is everywhere. What is the word for fire? As I said, there are multiple words.

It’ll be difficult for me to think of synonyms for friend in English. In Sanskrit, I’ll be able to think of 12 different synonyms for friend; all of which mean something different, based, as I said, on the verb. What is the word for fire, if I ask you? It is “agni”. But, what is being used here is not “agni”. It is “pavakaha”.

“Pavakaha” is something that purifies. If I translate it as “the fire cannot burn it”, of course I’m correct. But, I will never be able to convey these hidden nuances. This is a much more serious problem of translating from Sanskrit than the simple one of how to translate “dharma”, “karma”, or “varna”.

I was just spellbound, took me some time to process what you were saying. I’ll come to something very specific and then perhaps we can talk about the choices that were available to the poet of the times and the decisions that he took. This is again something that you and I have talked about before. There are various perspectives in the Mahabharata. Most perspectives are from the viewpoints of the Pandavas or maybe from Draupadi and so on. But in the abridged versions or the serial versions, we don’t have the so-called negative archetypes. One of the predominant things before the battle begins is not just the predicament of Arjuna but it is also about what Duryodhana is doing and who he goes to for counsel. One of the interesting things that you have pointed out in your notes is that he doesn’t go to Bhishma or his father or anyone else, but chooses to go to Suyodhana or Dronacharya, the guru, to seek counsel. Various semantic and philosophical hair-splitting analyses have asked, debated and tried to answer this question – why Dronacharya? The answer that you’ve suggested was extremely innovative. You said it is entirely possible that the nuances of the plot or the very complex and tenuous relationship that Dronacharya may have had with Suyodhana is far less important or perhaps had nothing to do with it. It just may have fit the metre of that shloka. Let us assume for a second that this was indeed the reason why it was Dronacharya. It could have been somebody else. Then my question would be that as something which claims to be itihasa, which is something that has occurred, that is, if expressed in a sort of Von Rankeian formulation, this claims to be authentic, empirical history which can be verified – would it be fair to say, just using this one simple example, there would have been conflicting demands of recording exactly what happened as Vyasa or any other subsequent poet knew it to be, and the demands of kavya, which made Dronacharya a better fit than Bhishma.
So you’re asking two different questions at the same time. As far as the first part of the question is concerned, people who are listening in may not have understood the context. What we have today is the texts as they have finally been crystallised in some form.

There is some dating, independent of the text, which has been done of the Kurukshetra war that includes archaeology, geology, and genetic evidence to validate this proposition about itihasa. We also know that the Mahabharata, in its present form, became a firm text by about 500 CE. Thus, 1000 years had passed between the occurrence of these events and the text getting some kind of final shape. Obviously, over a period of 1000 years, there will have been embellishments and additions to the core story. This happens with everything else.

For example, if we look at the Rigveda, in the earlier Mandalas, the river Saraswati is a thriving overflowing river. If we come to the later Mandalas, it is no longer a mighty river. It hasn’t dried up like it has by the time of the Mahabharata, but it is no longer the mighty river it once was. The incident that you are talking about is at the beginning of the Kurukshetra War. At this stage, Bhishma has not died. He is the commander of the army. At the beginning of the Bhagavada Gita… I have already cited the first shloka earlier. It is the only shloka that is actually spoken by Dhritarashtra in the Bhagavada Gita. Dhritarashtra asks Sanjaya:

“dharmakshetre kurukshetre samaveta yuyutsavah
mamakah pandavascaiva kimakurvata samjaya”

“O Sanjay, after gathering on the holy field of Kurukshetra, and desiring to fight, what did my sons and the sons of Pandu do?”

Sanjaya answers:

“drstva tu pandavanikam vyuḍham duryodhanastada
acaryamupasangamya raja vacanamabravit”

“Having seen the army of the Pandavas drawn up in battle array, King Duryodhana then approached his teacher (Drona) and spoke these words.”

Duryodhana goes to Acharya Dronacharya, not to Bhishma. The question is why did he go to Dronacharya, shouldn’t he have gone to Bhishma? All kinds of people have speculated about this. Dronacharya was a bit like a salaried employee. He was paid, Bhishma was not paid. Maybe that’s the reason. Dronacharya was a Brahman, so he may decide not to fight. Various kinds of people have speculated on this.

What you mentioned is just a speculative thought. Let me give you a different approach. I’m going to cite a shloka and you tell me where this is from.

“yada yada hi dharmasya glanirbhavati parantapah
abhyutthānamadharmasya tadatmanam srjamyaham”

“Whenever there is a decay of righteousness, o conqueror of enemies, and a rise of unrighteousness, then I manifest myself.”

Where is this from? Don’t be hasty, the shloka in the Bhagavada Gita says:

“yada yada hi dharmasya glanirbhavati bharata
abhyutthanamadharmasya tadatmanam srjamyaham” 

“Whenever there is a decay of righteousness, o Bharata, and a rise of unrighteousness, then I manifest myself.”

What is the difference? In the latter, the anushtup chhanda is intact while in the former it breaks down. It occurs elsewhere in the Mahabharata, not in the Bhagavada Gita. This means that the person who composed this bit was not as skilled a poet as Veda Vyasa. It was a later interpolation. This is speculation, but that’s what I think.

Similarly, if we are doing anustup, which is eight aksharas, it is easier to handle names that have four aksharas or even two aksharas. Thus, I’ll have shlokas in the Mahabharata on Satyavati or Draupadi. I will, of course, call her Yagnaseni and not Draupadi. These things will occur naturally to a poet. Then, thousands of years later, we will speculate about the powerful women characters in the Mahabharata. Why is Ambika not there as much as Amba? Maybe it was just a poetic thing.

Similarly, Duryodhana has been depicted as seeking counsel from Dronacharya because the word “acharya” would help you stick to anushtup. Bhishma wouldn’t have done so. This is pure trivial speculation. I think all that happened was that the poet was trying to fit the verse in that metre, but later on people have written academic papers on why Duryodhana went to Dronacharya.

When you’re talking about translations as subtle, nuanced and complex as the epics, how difficult is it to subsume or, if I may use the Buddhist or Hindu concept of “atma-vinash” (self-destruction), to kill the ego of the self when immersing oneself in the text? Very often, we see the unmistakable stamp of the translator in British or other 20th century poets who are translating from Russian, French ,and other languages. Does that present a challenge to you? If so, what is the way around it?
I don’t think so. I am reminded of what Chapman said in his translation of Homer. He said that my job as a translator is to depict accurately what Homer said. I, as a translator, should not intrude. I think that is the essential job of any translator. In my case, to depict or state exactly what Veda Vyasa said. Since I wanted to conform to the structure of the shloka, as the introduction of my English translation says, the English verses are not necessarily as smooth. If the text says Kaunteya, I will keep it as Kaunteya and not take the liberty of changing it to Arjuna.

But, there are things that require explanation. Let me give an example. Not from the Mahabharata but the Markandeya Purana. Towards the beginning of this text, it says that Duryodhana was Balarama’s “jamatri”. “Jamata” or “jamatri” – everyone would say it means son-in-law. The Markandeya Purana may not have been translated too many times into English, but it has been translated into other languages. I looked up all the translations and, lo and behold, everyone says son-in-law. But there is no record of Duryodhana ever having been Balarama’s son-in-law. Then I looked up multiple dictionaries and there was a minor meaning of “jamatri”, which is someone who is loved. That is all that was meant and, of course, Duryodhana was loved by Balarama as a favoured disciple.

To take a completely different example again, Yudhistira wants to do the shraddha ceremony of Karna and Jayadratha and goes to seek permission from Bhishma. This isn’t the shraddha that takes place immediately after the war, it’s the ceremony to be held one year later. So he asks for Bhishma’s permission to do Jayadratha’s sraddha. Apart from translations in English, the Mahabharata has been translated before in other languages. Many translators claimed that Yudhistira needed permission because Jayadratha was evil. That doesn’t sound right because so were a lot of other people. To me, the permission was needed because Jayadratha was Dushala’s husband. He belonged to a different family. The Pandavas did not have the right to perform the shraddha of someone belonging to another family.

Now, how do you ensure that these things are brought to the attention of the reader? I do that through my footnotes. All of these examples are given through footnotes. As far as the text is concerned, my intention is to keep it as authentic as possible to what Veda Vyasa said. Everything else is relegated to the footnotes.

The first volume of the Mahabharata was the most difficult. After all, I am translating the same author and I’ve got used to Veda Vyasa’s style. I can sort of almost mechanically or blindly translate a Veda Vyasa shloka now. Except that this is also “dharma” or “addhyatmadharma” (metaphysical). Eighty-five per cent of the time, I can type it out straight. But sometimes I have to research something such as – this particular river, where is it located today? I want to make it clear to the reader. Sometimes I get stuck on the meaning. After all, as I said, this is “addhyatmaadharma”. So I keep asking what exactly does this mean. There might be a word like “yoga” and we may believe it to mean something. But it could have all kinds of different meanings depending on the context. So this is how it works.

We probably can’t get away entirely from economics and statistics. I wanted to pose a statistical question on a ridiculously small sample size of three. But, correct me if I’m wrong, thus far, there have only been three complete unabridged translations of the Mahabharata – shloka for shloka, verse for verse. The first one, back in the 1880s, was by Kishori Mohan Ganguly. The second was done just a decade later by Manmatha Nath Dutt. The third person is you. I don’t know if it is a coincidence or otherwise, but all three translators of the complete unabridged Mahabharata are Bengalis. Are we to infer something from this?
Well, there are others who have tried to translate it but they died before it was completed. But, yes, I think we should infer something from it. By the way, these are translations into English. In the mid-19th century, Bengal was one of the first provinces to be exposed to Western education. It generated a lot of interest in English along with a lot of interest in our heritage as defined by this corpus. The first translation of any Purana, the Vishnu Purana, was done by Horace Wilson published by the Asiatic Society under Bibliotheca Indica. Haraprasad Shastri went off to the Royal Library of Nepal because it was the only one which had maintained some texts of the Puranas. He brought them back. They were then published in Sanskrit by the Asiatic Society under the Bibliotheca Indica series.

The first unabridged Bengali translation of the Mahabharata was done by Kaliprasanna Singha. Michael Madhusudan Dutt wrote his own version of the Ramayana. A lesser known fact is that Ashutosh Mukhopadhyay’s father did his own composition of the Ramayana. Today if you go to Kolkata and want to find a text of the Puranas, you will find one translated by Panchanan Tarkaratna, actually published in the 1890s. I’m trying to make a double point here. First, the exposure to English education and, secondly, the exposure to education in general, leading to an attempt to rediscover all of this.

Also, Bengal is a rather odd state in the sense that the Varna system, as we know it elsewhere, doesn’t quite exist in Bengal. First, we must remember that Greater Bengal was ruled by the Palas for a long period of time. The Palas, not the kingdom but the kings, were Buddhists. The Chaitanya movement led to great change as well. What am I trying to drive home? There is some truth in it but I obviously don’t want to exaggerate the point. The bulk of these translations were done by non-Brahmins. In many other parts of the country, these texts are regarded as the exclusive preserve of Brahmins. What was characteristic about Bengal was that whether it was Kaliprasanna Singha, Manmatha Nath Dutt, or Bibek Debroy – all of them are non-Brahmins.