Polly Barton is a Japanese to English literary translator currently based in Bristol, UK. Having published prolifically in multiple journals, she is also the recipient of the 2019 Fitzcarraldo Editions Essay Prize, the 2017 PEN/Heim Translation Fund Grant, and the 2016 Kyoko Selden Memorial Translation Prize. Her works include the translation of ‘Spring Garden’ by Tomoka Shibasaki (Pushkin Press, 2017) as well as ‘Fifty Sounds’ (Fitzcarraldo Editions, forthcoming), a personal dictionary of the Japanese language, and a record of a process of linguistic and cultural assimilation. Konstantinos Doxiadis spoke to her about her views on translation, language and meaning.

Interviewer: Having first studied Philosophy at Cambridge University, you only began learning Japanese at the age of twenty-one. What motivated this switch? Was it a sudden decision, or had you planned it long in advance?

I’d not planned it at all, in fact! As I was coming to the end of my time at Cambridge, I had a strong sense that I wanted to get somewhere far away for a while, so I applied for the JET Programme, which is where the Japanese government employ native and near-native English speakers from countries around the world to work as language assistants–not quite fully fledged teachers–in public schools around the country. At the time I applied, I was reading quite a bit of Japanese fiction and I’d always been drawn by Japanese aesthetic culture, so it wasn’t totally an eenie-meenie-minie-mo decision, but it wasn’t too far off that either. Then I found out that I’d been given a place on the programme, and the place was to be on a remote island, two and a half hours by ferry from the mainland port… I think it was only then that I started thinking, gosh, I probably really should try and get a bit of Japanese under my belt. It feels quite mad to look back on now.

Japanese is notorious for having a very distinct grammatical form and sentence structure to English. For example, sentences tend to follow a Subject-Object-Verb (as opposed to the English Subject-Verb-Object) structure, and verbs as well as adjectives are agglutinating (meaning morphemes are conjoined to form more complex words). Do these structural differences give you more liberty as a translator, allowing you to simply choose a consistent form you can write well in, or does it make it harder to convey what the author meant in Japanese?

I think there definitely is a liberty aspect, yes, which comes from the radical difference between the two languages. Which is to say that I can imagine when one is translating between, say, two romance languages or similar, one might feel an obligation not only to include cognates for individual words, but also to replicate as closely as possible the overall sentence structure – or at the very least, an underconfident translator would. Whereas when you’re working with Japanese, that option is off the table from the beginning. Of course you still worry about all kinds of fidelity issues, or at least I did, but I think that from my very first translations on I did feel a rush of creativity – there was never any illusion that translating was simply switching words around.

That said, the gulf that yawns between the two languages can be the cause of many headaches. In terms of how one puts together a sentence, Japanese offers more flexibility than English does, and there are times when I mourn how little wiggle-room there is with this language, how strangely rule-bound it is. Or, for example, owing to the loose S-O-V construction you mentioned, the most semantically loaded position in the Japanese sentence is what comes last, whereas in English it’s pretty much the start of the sentence. And so that requires a juggling act a lot of the time. Japanese does this thing often where the sentences are front-loaded with all this information, leading up to a punch-in-the-guts kind of conclusion (or a hilariously anti-climactic one), and that’s really tough to replicate in English. At the beginning of my career I was loath to cut up any sentences at all, but I’ve since realised that imposing a rule like that is really shooting myself (and the texts I translate) in the foot, so these days I wield my sentence-scissors where necessary.

Would you say that translating more liberally makes it easier for the translator’s voice to ‘seep into the text’? If so, is it something you would actively try to avoid as a translator, or would you instead just gravitate towards authors who have a similar voice to yours?

I think that when we begin down this line of discussion, we need to be really aware of the kind of language we are using, and the hidden assumptions and belief that lie behind the vocabulary we employ. For so long, everything demanded of the translator has been codified with this concept of ‘fidelity’, which seems simple on the surface, but is in fact a highly slippery and complex notion, which means various things at various times. But for that reason, when we use a word like ‘liberal’, I think it already contains a grain of disapproval, or the possibility of snapping into such. And similarly, and more obviously, a phrase like ‘seep into the text’ sets up this fallacy that what is ‘naturally’ there is the author’s voice, and the translator’s stuff is an intervention–when in fact, what is ‘naturally’ there is absolutely nothing. It’s a blank page. Any voice realised in the target language is, by definition, a conjuring trick, an act of creation on the part of a translator. One can only hope, and strive, to approximate as closely as possible to the voice, the spirit of the original. And I think sometimes, because of various stylistic and cultural features attaching to the source text, that calls for a translational approach that is quite inventive and envelope-pushing; at other times, less so. I think of it as a kind of dance, where you need both knowledge and intuition, rule-following and creativity–although I imagine once you get really good you can forget you’re using any of those things at all, and just strut your stuff. 

I’m not blaming you, incidentally, for using this kind of terminology–it’s ubiquitous, and I still find myself dipping into it too. But I’d like to encourage people, in a kind of Wittgensteinian way, to start being more aware of what we’re actually doing with our language when we speak of translation. I feel like a good starting point is to ask yourself in what ways the language you use to talk about the work of a translator differs from that in which you talk about the work of a writer, because that can be quite revealing.

Hmm, that makes things clearer. I’ll try to rephrase. When authors are writing, many of the lexical associations they make are subconscious, with active deliberation only entering in later stages of editing. Given this, should translators choose to translate authors who make similar subconscious lexical associations with them, or can they successfully approximate any voice via actively deliberating on word choice?

Aah, that makes things tougher! I wouldn’t really feel comfortable making normative pronouncements in this area. I mean, on the one hand, the further away one’s conceptual schema is from that of the author one is translating, the more that one is going to have to work to break down one’s automatic choices, to actively assimilate with the worldview of the author. But then I also think that there’s a danger on the other side too, of finding an author whose view chimes so well with one’s own that one can start to believe that the two of you share an identical system of lexical associations, and that can lead one to not sufficiently interrogate one’s choices. In other words, I think over-identification can potentially lead to complacency and ultimately sub-optimal choices.

Speaking personally, there are some authors I know I could never translate, because elements of their value system are so different from my own that the act of running their words through my brain and out of my mouth would feel viscerally wrong–and it’s hard to imagine I could do them justice within such conditions, either. That said, I like that there’s a decent amount of variation in the authors I translate. There are some with whom I really identify, and others whose work I admire, but with whom I’m definitely more conscious of getting into a role when translating. To make the slightly tedious, slightly pretentious acting analogy, I personally feel like playing a range of characters is important, but that the range should still be within the bounds of one’s competency and empathy. How one defines those bounds, I think, is down to the individual, but requires a certain degree of self-awareness.

While reading your translation of Misumi Kubo’s ‘From the Left Bank of the Flu’ (GRANTA, online) I was taken aback by how familiar the dialogue and rhythm of the text was. Colloquialisms such as ‘Damned if I know,’ and ‘I really had got chucked’ not only allowed me to empathise with an otherwise foreign character and context, but also brought out the slightly humorous undertones of the text. Cicero once said of his translation of Demosthenes, ‘I have not felt myself obliged to pay out each and every word to the reader. Instead, I have paid out an equivalent in value.’ Is this how you approach translating something as general as a humorous voice?

This is a discussion I’ve had often, and I struggle with it. The short answer is yes, but I’m not sure that I’m clear on what ‘paying out each and every word’ would even be, when translating between Japanese and English. Certainly it wouldn’t be a good translation–my sense is that it would scarcely be comprehensible! And then there are times when it’s actually perfectly possible–the word ‘chucked’, for example, is an almost literal translation of what’s in the Japanese.

But yes, I think as I’ve grown in confidence as a translator, I’ve felt more and more comfortable to think on the level of overall value rather than units. I’ve been at a talk by Juliet Winters Carpenter, one of the most talented Japanese to English translators that we’ve seen, where she said something to the effect that, often in order to translate the spirit behind the words you need to deviate from the literal meaning, sometimes quite far. She gave an example, in fact, of how she’d very significantly altered a sentence in order to convey the incredible resonance and cultural significance of a particular piece of onomatopoeic language as it occurred in the Japanese. It was utterly inventive, and it utterly worked. I was very much bolstered by that. And that’s never more the case than with humour, for sure. I tend to be attracted to authors who write humorous prose, often dryly and quirkily so, and when you’re dealing with that kind of material it’s so crucial to get the voice, the tone exactly right. And when you’re doing that you just can’t be thinking in each and every word terms. It just kills it.

You say you were ‘bolstered’ by this creativity, but it also sounds quite intimidating. Not only are you tasked with conveying the spirit behind the words, but you are forced to do so in a way which is much closer to how one would expect an author, rather than a translator, to work. Do you think that a good translator ought to be a proficient writer, in order to take advantage of such liberties? If so, might this proficiency make it easier for the translated text to depart significantly from the source?

I definitely think a good translator ought to be a proficient writer–not just ought to be, in fact, but has to be. Translation is, after all, nothing other than writing. Granted the translator doesn’t have to dream up plots, but s/he won’t get far without a good deal of invention and creativity. I think the notion that translation is somehow the province of people who aren’t quite up to writing their own stuff is getting very tired, and not only because there are so many examples of people these days who choose to do both very well. Translators are writers–that’s really all there is to it. But if I read you correctly, your second question seems to imply that this acknowledgment is likely to inspire more needless deviations from the source? The problem is this judgment around what is necessary or not, right… And it’s so hard to be prescriptive, because so much of the ethics around this is about these relationships of trust, of intimacy that exist between the translator and the author, which play out over a million momentary decisions over the course of the translation. It’s a ruleless land. All I can say is that a good translator is not only a wordsmith–s/he also has a strong desire to remain true to the spirit of the source. Which is to say, I believe that in the ideal translation relationship, the translator’s creativity is harnessed in service of fidelity to the text, even if that fidelity differs from an orthodox understanding of what that should or could be. I don’t know if that answers your question though!

Yes, it certainly does! I’d like to follow up on this idea of ‘trust’ between the translator and author, and of remaining true to the ‘spirit’ of the text. Do you think a translator is ever justified in (significantly) altering the content of the source in pursuit of this spirit? E.g. If the source delves into a culturally restricted event which will inherently seem peculiar to any foreigner reading the translation, would one be justified in tempering or even altering these cultural norms to make the translated text more relatable?

The norms for these kinds of changes vary so much across different media forms and genres. Advertising copy, for example, and games software – media where there is more focus on the effect on the consumer and less on the primacy of the original – are often localised to a great degree, whereas traditionally the wisdom has been that such incursions aren’t justified when dealing with ‘literature proper’. But you look at English translations of traditional Japanese texts and almost all of them feature not alterations but glosses, of cultural objects and concepts which aren’t familiar to the typical Western reader. Especially when you’re talking about a period of time when the interior landscape, culinary tradition and way of dressing were all entirely different, the translator needs to be supplying some kinds of information there to help the unaffiliated reader on his or her way. But that’s very different to altering, even if it could be called a form of localising. I’d say I’m generally wary of supplanting cultural norms, because I think that one of the joys and benefits of translated literature is that it prevents us from falling into a cultural-norm-echo chamber. That said, there are times when I will make small alterations in things like gestures, where in the original those aren’t of great significance in and of themselves, but rather serve to make clear a character’s emotional reaction. Or another example is the large number of ritualised elements of Japanese dialogue that exist. Children going off to school will have a particular exchange with their parents that can be translated literally means “I’m going and I’ll be back,” and the response is “Go and come back!” And there’s a similar exchange when they return: “I’m home!” “Welcome home!” Now you could of course translate each and every one of those literally–that’s a valid choice. But mostly the function of those in the text is to create this sense of familiarity, regularity, to evoke a stable household situation, so I think that a translator would often be justified for substituting that for something that feels friendly and warm in English – “Bye!” “Have a good day!” and so on. I think when you’re making those decisions, you have to be asking yourself what function the text in the original is serving – although of course it’s always possible to doubt your interpretation…

That makes sense. I would like to conclude our discussion by returning to an earlier point regarding the linguistic peculiarities of Japanese. You spoke of how Japanese often seems more flexible (syntactically) than English, but I was wondering what implications this has on the way in which ideas manifest themselves, and are interpreted, in Japanese. For example, in Greek many nouns are treated as gendered (instead of neutral), such as ‘the wind’ being male. This supports greater personification in literary works, and gives nouns a clearly defined identity and voice. As a result, natural elements such as the wind and sea are much more prominent as themes in Greek poetry than they are in English. Is there any similar phenomenon that comes to mind in Japanese?

That’s interesting, I hadn’t thought about gendered nouns in terms of supporting greater personification. I’m not sure if they could really be classed as similar to that, but there are certainly many features of Japanese that are notably different from English. The wealth of mimetic language, for example, which can be integrated so flexibly and diversely within sentences, and which amplifies the sensory, embodied, immersive qualities of a sentence – but which, if translated directly into English, often feel quite infantile, because we have a very different relationship to onomatopoeia and mimesis. Actually, this is in part what I’m dealing with in Fifty Sounds, the book I’m writing for Fitzcarraldo. There’s also the visual aspect to mention – when you have three alphabets like Japanese has, one of them which is made up of two thousand or so ideograms, you have just so much more potential to play around visually – to riff around by translating the same sounds between different visual representations in a way that is so thrilling as a reader, but sometimes when I’m wearing my translator’s hat makes me want to weep for the paucity of the English alphabet… A great example of that, actually, is The Last Children of Tokyo by Yoko Tawada; Margaret Mitsutani did an absolutely awe-inspiring job with that. There are many more things I could mention too, but I’ll stop there! 

You can find samples of Polly Barton’s work at: https://www.pollybarton.net/work