Wednesday, September 4, 2019

Nabarun Bhattacharya’s Harbart - Bengali innovation becomes a translation theory case study

A review-like bit first.  The book at hand is Nabarun Bhattacharya’s short, crazy 1994 (or 1993?) Bengali novel Hārabārta in the New Directions edition, translated by Sunandini Banerjee.  Harbart is what they’re titling it.  Harbart is the hero.  He kills himself five pages into the time-scrambled first chapter.  The rest of the book is more conventionally structured: childhood, adolescence, a surprising career as a “Conversations with the Dead” mystic.  Suicide, eventually.  There are real ghosts, although his medium business is a con; there’s a blue fairy, like in Pinocchio; there are some other unusual things.

Harbart has some resemblance in tone and attitude to the novelinos of César Aira which I enjoy so much, so the book is a good fit at New Directions.  Aira would love the ending.

Bhattacharya’s conceptual purpose has nothing in common with Aira’s.  Bhattacharya was political, a left radical.  He was blowing up Bengali literature.  I get the idea that one of his innovations, one reason Hārabārta became a cult novel, is that it is full of swears.  He introduced the street’s everyday crudity and idiom” – full of swears, I think that means.  That 2013 Shamik Bag article goes through the politics in much more detail than the novel itself.

The exciting thing in the novel is the language.  The Banerjee translation is the third English version – confusingly, also the fourth, but let’s save that idea.  I know nothing of the first version, but looked at the second (2011), by Arunava Sinha, on Indian Amazon.  Here he is, on the first page:

The mad old woman in rags was sitting there, her legs, splayed, splashing water.  Dogs with rotted flea-ridden skins were trembling in their sleep at the sound of screeching owls.  (Sinha)

Vivid, yes?  And here is Sunandini Banerjee, the new version:

A mumble-mad woman sitting there in rags, legs wide open, splashing the water.  Every now and then an owl letting out a screech and the skin-rotten dogs whimpering in their dreams. (Banerjee, 3)

The first page of the Banerjee also has “frothy-foaming light-dust,” “piss-cross games,” and “a Star-TV-signal-sucking satellite dish,” more compound neologisms.  The inventive compound words are a feature of the entire novel, a central device.  Great fun.  The novel is also full of languages – Hindi and English especially – which are translated or absorbed in the older translation but highlighted in Banerjee’s, the Hindi annotated and the English italicized.  This a noisy novel.

Looking back to that 2013 article, Sinha says that he “wants to retranslate Herbert (Sinha translated it as Harbart) in more creative ways.”  Banerjee has done it.

Harbart is now an outstanding case study in the domesticated (Sinha) versus foreignized (Banerjee) translation.  Better than I have suggested, even, because next week in England and elsewhere, yet another translation of Hārabārta will be published – also by Sunandini Banerjee!  Two translations appearing at roughly the same time by the same translator.  This cannot be common.  As Joe at roughghosts describes the books, the non-American version is even more foreignized, with more compound words and with Bengali in the translated text.  No reviewer of the American translation besides Joe has any idea that there are two versions.  I hope, once the texts are all available, that someone – ideally someone with Bengali – will knock the translations together and see what happens.

So interesting.  I want to pursue this idea tomorrow.


  1. The two editions of Harbart/Herbert are the same translation—Seagull and New Directions split the rights to share the cost of the translation (which is not unheard of). The differences between the books lie in editorial decisions. Seagull who is publishing the edition which is about to come out (which I also read) is a publisher of primarily European translations based in Calcutta for the past 37 years. Sunandini Banerjee is the senior editor and designer there. She is also a translator. So if it comes to a question of which edit is closer to the translator's intentions...
    And by the way, I'm Canadian, not American. :)

  2. Fascinating! But how can there be so many translations? When I was griping about the quality of the existing translation of Grossman's great Life and Fate and saying there should be a competing one, I was told "There isn't actually any question of another publisher bringing out another translation because all of Grossman is still in copyright." Surely a 1994 novel is still in copyright!

  3. If I had written what I meant, what would I have written? Not “No American reviewer” but “No reviewer of the American translation besides etc.” Not too late to change that. Thanks! I obviously know you are Canadian – I helped you get borrowing privileges at the University of Calgary library.

    I think I do not understand, Joe, how you are distinguishing between “editorial decisions” and “translation.” Here is how I think: New Directions says “We want the Bengali slang translated” (an editorial decision) and then the translator – a translator – translates it. The result is a different translation.

    Do you mean something else? The changes in verbs in your post look like a big clue, to something. A translator has an opinion on “whirled” versus “reeled,” not an editor. An editor can – should – also have Bengali and worry about accuracy. But then the editor is also translating.

    Isn’t the editing collaborative with the original translator, too? Versions sent back and forth, STET, that sort of thing?

    The number of translations of this book, within 15 years, not a century, is a phenomenon. It is an important book, sure, but the constant return to it is of interest. New Directions could have published the Sinha version, presumably, rather than helping pay for a new one. It seems perfectly readable, if less fun. The article that mentions the two Indian English versions (2004 & 2011) does not explain anything.

    1. If the editor had made major structural edits or demanded significant re-translation you might have a case for seeing the two editions as different translations. That is not the case here. The impression if you read the two is that the distinctively energetic translation has been toned down in the USS version. I have not spoken to Sunandini about her response to the edit, but remember, she is a highly respected and experienced editor herself. I do know what name she thought the main character should have. If I get back up to Calcutta when I'm in India later this year I'll be sure to have a chat with her. Off the record.

      I disagree that an editor can and should know Bengali. That could be a nightmare for a translator leading them to wonder why they even bother to put all the work into their translation. Did you see this recent essay (also involving New Directions sadly) which demonstrates just how much damage an editor can do in such a case?

      My point with discussing differing edits for translations published in different markets was simply to call attention to the fact that this occurs and normally amounts to little more than a denuding of some of the local flavour and character of the original translation. The ND edition is excellent. Seagull and Bhattacharya's family have been very supportive of it.

      I knew there were two editions because I have a close association with Seagull, but had not intended to read both until it chanced to come up after I worked with Bhattacharya's daughter-in-law on an essay and some translations of his poetry. I tried to be careful in my review to draw a balance between both editions as much as possible. The differences are interesting, but not, in my mind, an egregious misrepresentation of the text.

  4. Off the record, like that does me any good. Someone should interview her, like the Leon de Kock interview in the post after this one. Fascinating.

    Until the book is published, you are the holder of evidence, and I only have this passage to go on:
    “Herbert could sense it. He would have to charge-barrage now. Binu had had his time. It was Herbert’s time now. He would have to produce panic-pandemonium. Rip apart everything, torment-turmoil everything until the entire universe whirled in the dance of devastation.” (Seagull)

    “Harbart could sense it. He would have to charge-barrage now. Binu had had his time. It was Harbart’s time now. He would have to produce pandemonium – rip apart everything, turn everything upside down until the entire universe reeled in the dance of devastation.” (ND, Ch. 4, 45-6)

    Half of it is the same, and half is really different. Maybe this is a definitional disagreement again. I call “Moncrieff” and “revised Moncrieff” different translations of Proust. These passages are “Banerjee” and “revised Banerjee.” Of course they are mostly the same.

    If the changes have no reference to the Bengali, then, yikes! Those are big changes in that last sentence. Someone who can’t read the original changed the verb? Oh no. ND must have someone with Bengali involved somewhere. I mean, what if, in theory, a submitted translation is full of errors?

    There was no demonstration – no hint – in the article about Moser about any damage done to the text of the translation. Seemed to kind of dodge that issue, even. I did not understand that piece.

    It’s the differences that are interesting, yes. I am not sure where “egregious” and “misrepresentation” come from. This is the seminar room. We are working on the texts, seeing what we can learn from them.

  5. Sunandini is a friend of mine. I cannot imagine her agreeing to be interviewed. That would be very unprofessional given that this is a joint project. My guess is that Seagull approached ND with the proposal, but I don't know. My copy of Herbert has quotes from reviews of Harbart on the back of the dust jacket, so clearly they don't see it as an alternate translation. I still see the differences as editorial. The ND version is edited with a heavy hand to render it, I assume, more palatable. I would not need to consult a Bengali expert to suggest that change in the example above (though I would expect the translator would sign off on it and here it may just come down to a "pick your battles" kind of thing"). I do know ND did defer to their preferred expert regarding the name, so perhaps they pushed further. Again, I don't know. In the end, it's not like a less heavily edited edition won't be available—and in a handsome hardcover too.

    The Moser article clearly indicates that not only was damage done to the text of the translation, it calls into question the authenticity of every Lispector title published by ND. He convinced the publisher that Lispector must have a consistent voice in English throughout her entire oeuvre, one only he can determine. A conceit off the top because her voice changes from work to work. Every translator hired to translate one of her books has experienced the brutality of his edits (and this was not entirely unknown) but in this case the situation was far worse than anyone knew and finally someone had the courage to call him out. It makes me quite sick thinking of the Clarice Lispector books I have wondering what if any can be trusted. The editor and his/her ego can run rough shod over a translator if they have enough power.

    On another note, with respect to Leon de Kock's comments about Triomf, I haven't read it yet but I have the South African edition. It amazes me that they wanted it retranslated—yes it's full of slang but it doesn't impact understanding. Perhaps it's the bastardization of the language in the white lower class community. I would be surprised if the same extreme response would be needed today.

  6. If it's unprofessional for an artist to talk about her art, that's sad.

    Translation is an artistic and scholarly endeavor. The scholarly side depends on argument and evidence. Claims can't be evaluated without them. I mean, why, why would you suggest "reeled" in place of "whirled"? What's the argument? What's the evidence?

    On the artistic side, the final decision is often intuitive, creative. But translators are good at describing the problems they're working on and laying out the different solutions.

    I can't figure out how we know which changes are edits and which are translations. What's the evidence? I don't see how the blurbs work as evidence. The reviewers only saw one version. They didn't know the other one existed.

    This publishing stuff, it's all shadow. Open the windows, let in some light! Scholars do that. Publishers should do it, too. Show your work.

    1. I thoroughly agree. The heck with the secret-guild stuff.

  7. The interview I am imagining, by the way, is not to allow complaining about New Directions - I assume the translator was treated well - but to discuss the translation to English of complex Bengali texts, with examples drawn from her work. The kind of thing that will get a post at Languagehat.

  8. As someone who cohabits with a publishing professional, I can tell you that translations get edited, and the edits are not always run by the translators. Some editing is done for "clarity" (a moving target depending on who the reader is), and some publishers have "house styles" that are imposed on all the prose that goes out their door.

    Not all translators, of course, are created equal, and not all translators are necessarily a good fit with the author's style or subject matter. Not all authors have the patience to answer endless clarifying questions from translators or publishers, and sometimes translators and/or editors have to guess what the author meant, just like any other reader.

    I actually helped edit a translation from German a few years ago, turning the clunky English into less clunky English. The original German was not clunky. I did look some things up in the original text, but my priority was to make the prose better as English prose. Real "death of the author" attitude in play.

    In general I'm in favor of the idea of bringing the work to the reader rather than "preserving the flavor of the original" or whatever (whatever that might even mean, and surely that's another moving target depending on who's reading), and surely translation is always some kind of a paraphrase. Though I wonder how Shakespeare would feel about English paraphrases of his work ("No-fear Shakespeare" or whatever it's called). I would likely be annoyed at English-language paraphrases of my own novels that were sold as some kind of equivalent to the novels themselves. None of this is an argument against any kind of translation, really. I'm pro-translation, even bad translation, if the translator believes he's doing it properly. Translational intent is a new concept to fight about, isn't it?

    Jhumpa Lahiri decided a few years ago that she would only write in Italian. A dual-language book of her essays just came out. Lahiri did not do the English translation of her Italian original. The essays aren't great, but it's an interesting experiment, I think. When she reads the English translation, does she nod and say, "Yes, that's what I meant when I wrote it in Italian"? I don't know. I don't know how much it matters. I have strong doubts about the accuracy of language to capture thoughts anyway. Writing is a mere paraphrase of thinking, of living.

  9. In this case, the author is deceased. The translator was on her own in that sense.

    I was just looking at William Weaver's great essay in The Craft of Translation. He says when he was translating Gadda and he would call the author about difficult bits, Gadda would always end by saying "You know, just cut that part." Then Weaver would call a friend of Gadda and they would work something out.

    Margaret Jull Costa says that every question to Saramago was answered by Saramago's wife.

    Anyway. In this case, deceased.

    I still have this idea that one aspect of a good translation is its accuracy. There are many ways to be accurate, yes. Banerjee's translation is a move towards accuracy. But there are also ways to botch things, and a publisher like New Directions with an important book on its hands should have accuracy among its goals. My impression is that they do. They did in the past.

    I am against bad translations. I am for better-than-nothing translations. But if clarity and less clunk means, for example, cutting the hard parts, that is bad news.

    Venuti's 2nd edition includes as part of a case study the English edits of Henning Mankell's Wallander books, American translations Anglicized and then some, and then some more. Mankell's British publisher thinks its readers are simpletons.

    I mean, if you were helping with something for an airline magazine, then who cares, but if it was a previously unpublished text if Goethe's, that's something else.

    Shakespeare? Shakespeare wouldn't care. That man did not care about his texts.

  10. I guess I don't really have strong feelings about translation, except that there should continue to be more of it. It raises interesting questions, but I have profound doubts regarding "accuracy," not only to original texts but also within original texts. One opinion I have is that the "hard parts" of a text are often difficult to understand because the author couldn't fully articulate something he himself did not fully understand, or in fact wanted to remain vague. Yes, at some point you can say that the text is a fixed object, that a target language can be some kind of a mirror, that the reflected image should show some kind of fidelity to the original, etc. I just don't know what kind of fidelity, or if it really matters much. The real Goethe specialists will read German, of course.

  11. Oh no, a mirror, I hate mirrors. I have strong feelings about both translations and mirrors (as literary metaphors). Translation is far more interesting than holding up a mirror. More creative, more challenging - "an art, though a very modest minor one" as Donald Frame writes in a superb essay that is also in that Craft of Translation book, which I recommend highly to everyone.

    Frame is maybe a little too modest, but he translated authors where every part was a hard part: Montaigne, Rabelais, Molière.

    Are you using "accuracy" in some kind of metaphysical way I do not understand? I am using it instrumentally. Translate a line, show the original and translation to ten people who know both languages well. How many say "Hey, not bad, pretty close" and how many say "Ooh, no, that's not it"? This kind of accuracy (many say "pretty close!") is generally achieved by professional literary translators.

    Real specialists in many subjects read texts in translation and depend on their accuracy.

  12. No, no, I apologize. I have been short on sleep all week and I'm simply exhausted, muddle-headed and pessimistic. I have no idea what, if any, argument I'm trying to make. I think my original thought was just to say that all kinds of mischief happens within publishing houses that has nothing to do with the intentions of authors or translators. Sales and marketing departments also influence the editorial hand on prose, sometimes.

    Beyond that, yes, "accuracy" is being used as a metaphor for something, part of a philosophical argument about epistemology that you weren't having, that I didn't realize I was having.

  13. Well, I appreciate the argument. So often, it is the meaning of the terms. Almost always.

    And I know you are right about some parts of publishing. The so-called "young adult" sector seems pretty much corrupt, full of ghost-writers and named authors who barely contributed to the final text. A nightmare.

    I wish I had the Venuti book in front of me, but someone took it to her office. Maclehose publicly states that it Anglicizes translations by Americans, but it is clear, with the Mankell novels, that they also - well, it is like with juvenile fiction - they lower the reading level. They cut out metaphorical language. It is pathetic, although I am not sure who exactly to pity.

    But this is New Directions! I know, Laughlin is long dead.