Scylla and Charybdis: steering between Tolstoy translations

When, three years ago now, I was invited to review two new translations of Tolstoy's great 1877 novel Anna Karenina, I had no idea I was about to encounter a tempest in Tolstoy studies. Certainly, the appearance of two Tolstoy translations in the same year is a rare event - looking back at this blog post's list of previous Anna versions, you can see that such simultaneity hasn't occurred since 1960, and even when the novel was new to Anglophone audiences, it was only rendered in English four times between 1901 and 1918. I won't comment on the market forces that sent two translations careening on a collision course in 2014 - and indeed, my metaphor much exaggerates the reality. Both translations were by eminent translators at the peak of their careers; both were commissioned by university presses; both were well-received by critics. Rosamund Bartlett's version was published by Oxford University Press, while Yale brought out the Marian Schwartz translation. Nonetheless, as I discovered, precisely because of their expertise, skill, and confidence, each translator took a distinctive approach to her material; and any comparison of their translations opens up fascinating questions about how Tolstoy should be translated and, indeed, how any translation of a canonical text should be approached.

I recently had lunch with a former professor from my own department, long retired, who edited and partially translated one of the first major Penguin anthologies of Russian short fiction in 1981. He is a veteran of translating Turgenev and Bunin, both masters of nature description and therefore of demandingly specific lexis, the latter additionally infamous for twisty clauses and nuanced tenses. "Translating Tolstoy must be easy," he said. He meant that Tolstoy writes in relatively plain language, employing relatively short sentences compared to some of his peers. This notion could hardly be more wrong. Translating Tolstoy is very difficult indeed, for two main reasons: firstly because Tolstoy is a rule-breaker, deliberately writing incorrect Russian (especially in regard to gerunds and participles), and secondly because Tolstoyan simplicity, when translated into English, creates a false impression of undertranslation.

This is not to suggest that Tolstoy's style is illiterate or unpleasing: quite the contrary. However, his prose does confront the translator with unhelpful syntax: adjectival traffic jams; awkward, unmanageable, and not always even conventionally grammatical gerunds (which, as Eugene Lampert wrote in 1973, ‘fastidious translators do their best to obliterate’); and enigmatic, often incompletely cited, peasant idioms. Schwartz firmly believes that the 'unconventional and unsettling' effect of Tolstoy's style, the occasional 'roughness', the use of apparent "mistakes" and of course the repetitions, are all intended to "convey meaning, to express his spiritual and moral concerns' (Translator's Note, xxiii). An obvious example of repetition that both translators cite is the adjective veselyi (jolly) and its cognates such as veselost' (jolliness, good cheer), which Bartlett claims occurs 318 times in Anna (and she should know). Schwartz chooses to translate this word wherever it occurs by a single English equivalent – cheerful – and its cognates (e.g. cheer, cheery). She suggests that  by constantly referring to 'cheer', Tolstoy meant to provoke 'ominous associations' (xxv) in his readers' minds - a suspicion that the characters were in fact very far from cheerful. Because Russian is an inflected language with multiple derivations and affixations possible from a single stem, in the original, this repetitive technique creates a rich web of inferences and implications. In English, it causes most readers to wonder at the apparent poverty of the translator’s vocabulary. Surely Tolstoy couldn’t have been such a limited writer, constantly re-using the same word?

Bartlett resorts to a richer vocabulary, including ‘merry’, ‘livelier’, and ‘light-hearted’, in order – as her introductory essay explains – to convey the ‘richness of meaning implied in the original’. She asserts that Russian is simply more concise than English, and that therefore multiple meanings may be implicit in a single word; thus to fix on a single English equivalent for that word, as Schwartz does with veselyi, would be unduly confining for the translator (and repetitive for the reader). I quote her introduction to her own translation: 'This translation seeks to preserve all the idiosyncrasies of Tolstoy's inimitable style, as far as that is possible, including the majority of his signature repetitions, so often smoothed over by previous translators [...]. At the same time, it is a mistake to render Tolstoy too literally. He was often a clumsy and occasionally ungrammatical writer, but there is a majesty and elegance to his prose which needs to be emulated in translation wherever possible. Tolstoy loved the particular properties of the Russian language, but he would not have expected them to be reproduced exactly in translation, and would have surely expected his translators to draw on the particular strengths of their own languages. the aim here, therefore, is to produce a translation which is idiomatic as well as faithful to the original, and one which ideally reads as if it was written in one's own language' (Notes on the Text and Translation, p. xxxviii). There is a lot of good sense in this approach, and it certainly makes for a richer text for the Anglophone reader. And yet we must remember Tolstoy uses repetition for several reasons, including for emphasis; for the psychologically jarring sensation which Shklovsky would christen ‘defamiliarization’; and for the ‘Hansel’s breadcrumb’ effect, that is, using a chain of similar words to clarify the narrative’s symbolic underpinnings. The style is meant to convey meaning; to provoke discomfort; and to convey meaning by provoking discomfort, rather like a parallel process in cinema, Eisenstein's notion of intellectual montage, where contrasting or shocking images initiate an emotional or cognitive process in the viewer's mind. Unwise translators, by gobbling up the repeated words and substituting unrecognizable synonyms, may erase Tolstoy’s subtly laid ‘pathway’ through the plot - and forestall the thought processes that the author had intended to unlock.

Here is an example of a disrupted pathway: in Part Five of the novel, Vronsky is placed in an impossible situation by Anna, who is now living with him openly. Vronsky would like Anna to behave discreetly and accept that she must avoid society until their relationship is regularized by her divorce and remarriage. Anna, however, suffers bitterly from former friends’ contempt for her new status; additionally, she fears that their attitude will undermine Vronsky’s love for her. In St Petersburg, she defies unwritten social rules by attending the opera in full décolleté as if nothing were amiss. Vronsky has to witness Anna’s public snubbing, while fielding his mother’s mockery (despite her own chequered past, Vronsky’s mother hates Anna for spoiling her son’s career). In a short descriptive passage which follows Vronsky’s progress from his mother’s box at the opera to Anna’s, Tolstoy uses the same adverb to describe both women’s actions: nasmeshlivo (jeeringly). Spaced just a few lines apart, the repetition of this word forces the reader to compare these two apparently incompatible women. One is brave, passionate and despairing, the other is an immoral and cynical hypocrite: yet both treat Vronsky similarly during this crucial scene. Schwartz, predictably, rises to Tolstoy’s challenge by repeating the word ‘derisively’ for each use of nasmeshlivo. The effect of the original Russian is thus reproduced: there is, of course, a risk that readers will fail to recognize Tolstoy’s deliberate jar and blame the translator for bad writing instead. Bartlett, on the other hand, dodges out of this quandary by ingeniously making Vronsky’s mother’s tone ‘scornful’ and Anna’s expression ‘arch’. This creates a much smoother reading experience yet disarms that important Tolstoyan tripwire.

Thus the basic translation question that emerges here is: should translators risk rendering a provocative text in a version that reproduces the lexical and stylistic effects of the original? (The technical term is foreignization, as opposed to domestication). Or should they produce a 'smoothed-over' version that doesn't - shock horror - read like a translation? In both scenarios, the translator gets blamed. In the second scenario, however, only the academics will notice, as readers will be palliated by the more 'natural' style. Marian Schwartz's translation of Anna, precisely because it meets Tolstoy's challenge head-on and, consequently, reads in places more 'like a translation', has received an unfair degree of criticism.

Peaceful and rather scholarly debate spilled over into the global media when the journalist Janet Malcolm wrote an extended review of Tolstoy translations in The New York Review of Books, Socks, which attacked Schwarz's Anna for its 'awkwardness' and 'obtrusive literalism' and singled out one particular translation decision, her use of the invented word 'shapify' to translate the Russian verb 'образоваться', as 'an elaborately badsome English neologism' (pot, kettle, one might murmur). The viciousness of this review galvanized readers, translators, and academics to take sides or at least to discuss the issue: my colleague XIX век has collated many useful links to articles and blog posts on the topic here. A few months later, the NYRB published selected responses to Malcolm's article: of these, one of the most interesting is by the translator Judson Rosengrant. He is no kinder to Schwartz, calling her choice of shapify 'bizarre' and 'meretricious'. Rather than taking issue with Schwartz's style, like Malcolm, Rosengrant delves into the morphology of the word translated to explain why this translation might distort Tolstoy's style and mislead the reader. Essentially, he asserts that Schwartz has over-translated Tolstoy: by interpreting his use of the word 'obrazuetsia' as a neologism rather than an archaism, she created an unnecessary stylistic disruption and therefore (according to Rosengrant) effectively out-Tolstoyed Tolstoy. My own consultation of Dahl's dictionary suggests that there is a strong case for arguing that there is a fairly strong case for using a colloquial phrase rather than a neologism for 'образоваться.'  The verb is used by a servant, speaking casually but respectfully, to mean 'things will take their course'; Dahl suggests possible meanings such as 'to be shaped [by another]', to take form', but these tend to be in quite specific religious or ideological contexts, such as the growth of an unborn child or the blessing of a marriage. Hence, as Rosengrant suggests, the word is an archaic one used in an inappropriate context. It is a defamiliarization - but not quite a neologism. Why, you might ask, is 'shapify' so important? First of all, as a neat neologism, it makes a succinct title for book reviews (such as Carol Apollonio's TLS piece on the Schwartz and Bartlett translations). Secondly, as the word is repeated at key points throughout the novel, it becomes an important breadcrumb in Tolstoy's Hansel-and-Gretel trail of meaning through the narrative. And thirdly, as a radical translation decision, it summarizes the ongoing debate for translators of Tolstoy: should they take the rough with the smooth? Or just the rough? Or just the smooth?

Here comes my pennyworth. Another media flurry, so far positive, is in progress around Emily Wilson's new translation of Homer's Odyssey which, startlingly, appears to be the first ever rendering of this canonical text by a woman - unless you subscribe to the minority opinion that Homer was female. Interestingly, Anna Karenina  has been much more frequently translated by women or men working with women than by men alone: which makes me wonder whether one's choice of language for literary translation is strongly gendered and if so, what cultural and educational factors influence this gendering. Wilson's own views on translation and gender are uncompromising: 'the gendered metaphor of the "faithful" translation, whose worth is always secondary to that of a male-authored original, acquires a particular edge in the context of a translation by a woman' (Translator's Note, 86). Rather like Bartlett, Wilson brings impressive academic credentials (she read classics at Balliol and is now Professor of Classics at the University of Pennsylvania) to her role, which makes her translation decisions rather harder for would-be detractors to sneer at. When Wilson writes about the Odyssey, she could well have Anna Karenina in mind: both are texts that are 'deeply invested in female fidelity and male dominance' . Even though Anna's infidelity creates Tolstoy's narrative, the iron fidelity of her sisters-in-law Dolly and Kitty holds the plot together. In Homer, the marriage bed, carved by hand from a tree, forms the linch-pin of the relationship between faithful Penelope and her straying husband). Both texts are about women who stand by their men.

Tolstoy has often been compared to Homer, and not just because their most famous works have the most contested first lines in literature (compare "All happy families are alike, all unhappy families are unhappy in their own way", the opening line of Anna Karenina, with the less widely-cited but equally contentious first line of the Odyssey, discussed here by Wilson). Tolstoy read, adored, and emulated Homer. Stylistically, he has been usefully compared to Homer, and as Wilson reminds us, 'Homeric style is actually quite often redundant and very often repetitious' (82). Doesn't it follow, therefore, that best practice for translating Homer might also apply to Tolstoy? If we turn to Matthew Arnold's 1860 lectures called 'On Translating Homer', which some regard as the gold standard for this task, we are encouraged to aim to translate intention, manner, and inspiration, rather than word-for-word lexis. Arnold warns, 'To suppose that it is fidelity to an original to give its matter, unless you at the same time give its manner; or, rather, to suppose that you can really give its matter at all, unless you can give its manner, is just the mistake of our pre-Raphaelite school of painters, who do not understand that the peculiar effect of nature resides in the whole and not in the parts. So the peculiar effect of a poet resides in his manner and movement, not in his words taken separately'. This approach might remind us of Schwartz's with Anna Karenina, and could even justify her use of shapify where, arguably, no such neologism was required: Schwartz, having assimilated Tolstoy's 'manner', which tests and occasionally defies conventional lexis, was simply anticipating his 'matter' with this invention.

Emily Wilson follows Arnold's injunctions to preserve four aspects of Homeric style ('plainness, simplicity, directness of thought, and nobility'); the same might be said of both Schwartz and Bartlett's approaches to translating Tolstoy. At the same time, Wilson unflinchingly admits that no translation can ever be successful in the sense that it can reproduce the aesthetic effect of the original; as she puts it, all modern translations are inevitably and 'entirely alien from the original' (87). Our modern language is irreparably distinct from Homer's Greek; so is the world as we understand it; so is our culture. We might have more in common culturally and cognitively with Tolstoy's Russia, but not so much that the question of alienation can be avoided. In her Odyssey, Wilson aims for simplicity, making Homer as accessible as he would have been to his original hearers and readers, but she doesn't hesitate to make her text deliberately strange by including poetic effects and discordant language, in order to remind modern readers of the chasm between the Homeric universe and our own, and not least, to remind them that this is a translation. Her hope, as she expresses it, is to sustain 'a register that recognizably speakable and readable, while skirting between the Charybdis of artifice and the Scylla of slang' (87). What translator can hope for more?

Parts of this post have previously appeared in my review articles on translating Tolstoy in the East-West Review (Autumn 2015) and Translation and Literature (26: 2017), pp. 214-222.