Wednesday, 14 September 2011

Some Thoughts on Translating Wittgenstein

The sharp-eyed amongst you might’ve noticed that I’ve been quoting from the 2009 translation of Philosophical Investigations, rather than Elizabeth Anscombe’s original 1953 version (the new edition is based on Anscombe’s translation, but amended by Peter Hacker and Joachim Schulte, with the help of the Wittgenstein editorial advisory committee). Actually, the 2009 translation was the catalyst that sparked this blog; I borrowed it from a friend a few weeks back just to see what changes had been made. Then I decided it was about time I read it again; then I decided to read it carefully; then I realised (to my shame) that there were certain fairly key passages I only partly understood; then I realised that correcting this would mean taking notes and working through examples; then I decided that keeping a blog would help spur me on and force me to focus my thoughts; and then… well, here we are knee-deep in language-games and the definition of “definition”.

Anyway, I approached the Hacker/Schulte translation with a certain amount of trepidation. The Investigations is one of those rare philosophical works that is admired not just for its insight, but also for its style. It is gnomic, elliptical, aphoristic, playful, discursive, conversational and yet always (somehow) serious and to the point. Delving into it after reading, say, Descartes’ Meditations or Hume’s Treatise on Human Nature, you cannot help but realise that you are meeting a mind that is simply not like that of other philosophers (perhaps only Nietzsche’s works come close in this respect). That it is such a pleasure to read is, for English speakers, due at least in part to Anscombe’s excellent stylistic interpretation. So what changes had been made to this classic? To put it bluntly, had they mucked about with it?

The answer, mostly, is: no. Anscombe’s original words remain largely intact and, so far as I can tell, substantive changes to the text are based on sound methodological reasoning (I should point out that I speak no German, so I’m not qualified to comment on the accuracy of the translation itself). Having said that, I have still found myself balking over small changes to cherished passages. Today, for example, I looked up §371 and was disconcerted to see “Essence is expressed in grammar”. In? In?! How could they change “by” to “in”? It just doesn’t read right!

Of course, a good deal of my grumpiness about this was brought on by nostalgia. The Investigations has been part of my life for over 25 years and so even a change for the better would be a bit like seeing a new wrinkle on the face of an old friend. All the same, it’s sad to see that some of the “flavour” of Anscombe’s prose has been jettisoned in order to bring the text up to date. As the editors remark, “Anscombe’s translation is now more than 50 years old, and English has moved on apace”. Accordingly, they’ve stripped out some of her more archaic usages, such as spelling “show” as “shew” (which Word considers a mistake), her fastidiousness over the use of “shall” and “will”, and the use of words like “queer” (meaning “odd”) and “fishy”.

Personally, I’ve always found such quirks rather charming, but there’s a more important point here that’s not simply a matter of taste. Ascombe’s translation was itself a product of the time and place in which the Investigations was written. She was a pupil and close friend of Wittgenstein’s, and such was his faith in her ability that he asked her to translate his work even before she had learnt German. Her linguistic quirks might not have been the same as his but they were from (roughly) the same milieu, and I can’t help thinking that’s a mark in their favour.

To put things in perspective, imagine how it would sound for a translator of Faust to say, “I’ve updated the verse-style because German has moved on apace since Goethe’s day”. In such a case I think it would be fair to wonder if making things easy for the modern reader was the only valid consideration. Sure, give us the sense of what he wrote, but wouldn’t a suggestion of Goethe’s archaic syntax, spelling and vocabulary also provide us with an insight into his creation?

Now, obviously the Investigations is not a poem, and different priorities are bound to operate when translating philosophy. But how far – even here – should bare ease of reading predominate? This seems a particularly pertinent question in the context of a philosopher who once remarked that a great work of philosophy should also be a work of literature, and whose masterpiece bristles with such a keen sense of the different ways in which language can convey meaning – indeed, the different forms that “meaning” can take.

Furthermore, it’s strange that updating archaic idioms seems more relevant to translations than to original-language texts. Here’s a quote from Locke’s Essay Concerning Human Understanding, taken more or less at random: “If we will disbelieve everything, because we cannot certainly know all things; we shall do muchwhat as wisely as he, who would not use his legs, but sit still and perish, because he has no wings to fly” (Book 1, §5, 2004 Penguin edition). I think it’s fair to say that English has certainly “moved on apace” since 1690. Why are we still forcing students to read this stuff? Where are the modern-vernacular editions of Locke, or Berkeley, or Hume?

Well, let’s not get carried away. Over all, the new edition of the Investigations is a fine piece of work. And regarding its style we are, at present, only talking about a few discrete tweaks here and there. All the same, this sets a precedent and I can’t help wondering what the 150th anniversary edition will look like. Wittgenstein considered his writing to be against the spirit of his age, and – precisely for that reason – the Philosophical Investigations is very much a product of that age. To read it outside of its historical context is to miss something of its meaning. Anscombe’s prose helped anchor the translated version firmly in that context. I think it’s a shame the connection has been diluted, and worry about how far the trend may continue. Or maybe I’m making a mountain out of a molehill? I don’t know.


  1. Really interesting post - I find the 'in' instead of 'by' change most curious, I'm not sure how that can be justified as simplifying, and it slightly changes the feel for me (first reaction is that 'in' = contained within, whereas 'by' = via?).

    I think there is definitely a place for new translations - and indeed 'translations' from older English texts into modern language - to aid understanding, but I agree that you miss something by not reading the original (or at least a translation so close to the original as Ascombe's).

  2. Hello and thanks for dropping in!

    Regarding "by" vrs "in", it could be either (a) that "in" is closer to the German word "in" or (b) a way of saying that grammar REVEALS essence as opposed to BEING essence. No, I'm not convinced either.

    As for updating, yes it can be helpful and we shouldn't deny ourselves this help. But I think it's important that the original linguistic quirks are still available. Locke's prose style (for example) is part and parcel of the 17th century mind. So long as grasping the sense of his writing is not extremely difficult I think that aspect of his writing should be retained. At least in part the medium is the message.

  3. I just discovered this blog through a mention in Duncan Richter's blog, where I have recently been a fairly regular commenter. Looks quite good to me!

    A[n]scombe's translation was itself a product of the time and place in which the Investigations was written. [...] Her linguistic quirks might not have been the same as his but they were from (roughly) the same milieu, and I can’t help thinking that's a mark in their favour.

    In 1999, John O. Nelson published a very interesting little piece titled Is the Pears-McGuinness Translation of the Tractatus Really Superior to Ogden's and Ramsey's? In it, he makes exactly the same argument about the Ogden-Ramsey translation of the Tractatus. And I think he has a point. I certainly have always preferred the Ogden-Ramsey translation, for much the same reasons Nelson advances, and I have been much troubled by its cavalier erasure from contemporary scholarship. Somehow I'd never thought of the parallel with Anscombe's translation, but it seems instantly obvious now that you wrote what you wrote.

    It's always been fascinating for me how Wittgenstein is largely read through his translators, and not just the English ones - and how taken for granted the translations are in spite of their status as "mere" translations. (I myself left a looming career in academic philosophy and Wittgensteinian scholarship to work in document translation, for which I have no qualifications at all other than sheer fluency; whatever talent I have has always been packed exclusively at the linguistic end of the spectrum and not the mathematical/scientific end.)

    Here in Finland, all of Wittgenstein's canonical published works, with the exception of the not so canonical Philosophical Grammar, were translated into Finnish in the '70s and '80s by von Wright's erstwhile research assistant Heikki Nyman. His translations are as revered and taken-for-granted as those of Anscombe and others are in the English-speaking world, and for equally good reason. This month saw the publication of my own translation of Wittgenstein's Koder diaries, which were discovered only after Nyman had already retired from academia, and I must say that his boots were really intimidating to step in at times. The disjointed and private character of the text, compared to the more heavily edited and exclusively philosophical ones Nyman translated, was fortunately somewhat helpful in freeing me from his shadow.

    Part of the fascination for me is the fact that translators into languages other than English have often used solutions that are very different, but which nevertheless appear as the obviously correct ones to readers psychologically familiar with them, in much the way that (say) Anscombe's solutions appear to readers familiar with them. Thus, Nyman translates §371 as "Olemus saa ilmaisunsa kieliopissa" - which would translate into English as "Essence receives its expression in grammar". This is the very different translation I grew up with, so I can't get too excited about whether "is expressed by grammar" or "is expressed in grammar" is better. And such inability is itself something interesting and noteworthy.

    Why are we still forcing students to read this stuff? Where are the modern-vernacular editions of Locke, or Berkeley, or Hume?

    Well, as a matter of fact, Jonathan Bennett has prepared many such and made them available through his web site, Some Texts from Early Modern Philosophy. He also supplies a rationale which is certainly very debatable, but I don't think it can be dismissed out of hand. If only for the reason that it's as much a question of aesthetic taste than historical accuracy.

  4. Interesting. I hadn't made the connection with translations of the Tractatus but I can see how the same argument would apply - and I'm willing to bet more English-speaking people know the quote "Whereof one cannot speak..." than the Pears-McGuinness equivalent.

    And I agree that the psychology of translation is fascinating. The Bible is an obvious case in point; those who love the King James version tend to assume non-English speakers have to "make do" with their versions because (obviously) the KJB is THE translation of the text. They can't quite believe it could be as stirring in French or Russian - in fact they'd probably only grudgingly admit it might be as good in the original languages!

    RE: Locke et al, in a way I think this supports my point. Updating a 300 year-old text is controversial when it's a case of English-to-English, yet updating a 58 year-old text is regarded as normal when it's German-to-English. By this logic we're going to need a modern English version of Naming and Necessity in 18 years' time!

    Hope your translation of the Koder diaries is well received!

  5. Hi, Philip -

    Arriving via mention in Philosopher's Carnival.

    I'm currently reading PI for the first time, so I'll be following your rereading with great interest. I started a similar series of postings (in lieu of margin notes) when I started but found it took too more time than anticipated. But I did a few, some including translation issues. I don't speak or read German (despite taking German in college decades ago), but am reading the Hacker et al German-English edition - by now (just got into the 200s), mostly in German. I immediately became suspicious of the translation and look up many words in a dictionary rather than accepting the accompanying translation.

    Of special interest to me - owing to a speculation that we process language in units larger than individual words - concerns the interpretation of "ganz" in Remark 31, addressed in the last paragraph here:

    Since I have no previous direct exposure to W's writing, I'd appreciate your opinion re my alternative interpretation.

    Other comments/questions on the remarks you've already covered will follow.

    Thanks for your efforts here.

  6. Hi Charles

    I sympathise with your finding a blog on the PI too time-consuming!

    I'll certainly check out your site and give what comments I can. Off the top of my head, I suspect the rate at which we process words is an empirical question rather than a conceptual one and so outside the scope of the Investigations. But I'd have to give it some thought.

    Thanks for reading my stuff and hope it sheds some light.


  7. Charles - comment on this posted directly at your original post on your blog.

    Philip - I burst out laughing at your mention of Naming and Necessity. That was a very good point about its being not that much younger than PI.

  8. Philip -

    You are, of course, welcome to read the cited comment anyway, but Tommi has set me straight on the translation issue I raised. My error.

  9. Charles - OK, I shall certainly have a look at your post.

    Tommi - Ha! Actually, I should've said "15 years" as the Hacker/Schulte translation came out 3 years ago. Mind you, although it's only 18 years, a pretty large cultural gulf separates 1953 from 1971.

  10. The worst change in the Hacker version is replacing Anscombe's 'in a flash' with 'at a stroke.' 'At a stroke' is not any contemporary idiom that I am aware of. It rather strikes me simply as confused and confusing.

  11. Modaloperator, it's not true. "At a stroke" is a very common phrase. I hear it a lot and use it all the time.