Cole Swensen

In Praise of Error


Much thinking on translation is based on the assumption that translation has something to do with two languages, but this is wrong; it’s a matter not of two languages, but of the gap between them, and far from reducing that gap, it for the most part increases it by making us suddenly aware of what cannot be said, not only in the other language, but also in our own. In short, it makes us aware of a poverty we never before realized we suffered, and we quite naturally resent this—and (on the kill-the-messenger principle) this distances us just a little more from the other language. It’s not only what we can’t say that we notice, but also nuances of feeling and insight that we simply can’t access. These all seem to fall head-first into the inter-lingual chasm, reawakening our innate suspicions about the futility inherent in speech in the first place. So don’t look to translation to bridge cultural gaps; it’s only going to aggravate them.

However and on the other hand, as one of the greatest contemporary threats to vibrant, productive culture is the homogenization that results from increasing globalization, translation offers great benefits. It maintains particularity, refuses acquiescence, and increases the misunderstandings that linguistic difference often occasions.

Misunderstanding is greatly underrated. At its best, it can offer a whole new view of things, a fresh interpretation, which doesn’t replace the first, but adds a parallel ghost, a sosie that doesn’t quite fit. It augments what was intended, contributing its own details and flair. And the nice thing about misunderstanding is that you don’t need a lot of it—a little, as we’ve all noticed, goes a long way. In fact, slight misunderstandings often create the sort of offset that ignites a reverberation. We’re more familiar with this in the visual field—for instance, the slight offsets that animated much pop art in the 1960s—but the same thing occurs in the mental realm: two thoughts that are very close but not quite congruent set up a ringing between them, creating a slight additional dimension that brings them both out more strongly.

Is understanding ever perfect? Can we ever know, sense, feel precisely what another intended? I doubt it. What we call understanding are simply those instances when the additional meanings are too slight to become conscious. They slip by unnoticed, and are therefore often wasted. Understanding is, in short, pretty useless; it merely gives us more of what we’ve already got—it’s in misunderstanding that the promise really lies. Which leads us to ask how much difference we can tolerate between utterance and interpretation before we say that understanding has missed.

We can put all this on sound theoretical footing by appealing to information science and its concept of self-organization from noise. The concept is a simple one: whenever a message gets transferred from one site to another, whether it’s from speaker to listener in a room or over the telephone, or from writer to reader in a book, or from parent to child through biological reproduction, there’s always the chance that the message will become changed in the transmission. We tend to think of such change as damage, but, as in the case of the alterations to the genetic message that led to the opposable thumb, for instance, change is not necessarily bad. Like most things, it depends on what you do with it.

Alterations that occur to a message within the transmission and reception system are called “noise” (think of the classic crackling “bad connection”), and noise always makes the system more complex, which means that its potential for meaning has been increased. To apply this to translation, we have to think of the translation and its original as a single system, one which also incorporates the writer, translator, and reader; texts don’t work on their own, but must be involved with these other elements to attain any meaning at all, and in fact, it is the texts themselves that generate the system. Any text’s system is always growing; it includes all translations of it and all other writings on it, as well as all its readers and their interpretations of it and, eventually, all the effects—thoughts, deeds, etc.—of which it is the cause.

Even within the skeletal version of the system constituted by text, translator, translation, and reader, no message is ever received precisely as sent, so a translator is always working in an environment in which meaning is a priori threatened by everything from changes in connotations and references to reader distractability. To try to be “faithful” in such an environment would require an approach like Pierre Menard’s to his Don Quixote, in which, through tremendous effort and extreme concentration, he was able to translate Cervantes’ classic into an exact copy of itself. The only way to avoid that pitfall is to accept that a certain amount of noise will inevitably be incorporated, but that the attentive reader will be able to organize that noise into something meaningful.

Children recognize the value of this principle and use it as the basis for the game Chinese Whispers. It delights them because they put a different interpretation on loss. So much thinking in the field of translation is conditioned by the conviction that something is necessarily lost, but I’d like to counter that. In the first place, such an approach starts the discussion off on a morbid note; we enter already dressed in mourning. It also introduces a tinge of nostalgia for ahh “that which is no longer.” Both attitudes have us looking backward, not forward, and accepting defeat before we’ve even begun because that way we have less to live up to in the long run. In the second and more important place, nothing is lost, every bit is right there where it has always been: in the original; it is perfectly safe, and we will know right where to find it when we need it. Therefore, the translation can afford to lose something, and in fact must lose something, for a translation that loses nothing will not gain anything either. If, however, we regard difference not as loss (which is simply an emotional judgment placed upon change), but as augmentation, we can argue that translation always adds something—and that is the truth of the matter.

This is not to say that we shouldn’t cultivate the illusion that something is lost, for, like misunderstanding, loss is abundant resource not to be discounted. The notion that something is lost allows us to enter into the rich and rewarding process of searching for it, which, while not likely to yield the sought-after element, does often cause us to run across other useful and delightful things we never would have thought to look for. Missing things are useful, too, in that they stimulate the imagination. Readers who assume that much has been lost will imagine much more powerful, moving, and lovely things than are likely to have been there in the first place. It makes readers stretch their minds an extra measure, as they prefer to risk outdoing the inaccessible original rather than feel like they’ve missed out on any of what’s been advertised.

Thus the lost becomes a ghost, a spirit that hovers over the text, an undifferentiated energy that, like stem cells, can be tapped and adapted to meet the needs of the moment.

Misunderstanding and loss are the two principal types of error open to the enterprising translator, for to translate is to err, both in these ways and, perhaps most of all, in the French sense, as in errer: to wander, to roam. Translation, like all writing, is intimately linked to walking. Writing is the walking of the mind. Like the walking of the body, it combs through the cells, gets them aligned, and like the body’s walking, its essential medium and product is rhythm. It establishes the rhythm that puts the heartbeat and the breath in sync; it keeps time with the outside world.

If all writing is walking, translation is walking of a certain sort, the sort that wanders, meanders; it must take an uncharted path. While original writing charts a path, translating takes that path, but without following it, takes it as though it were raw territory, thus tracing the known, but maintaining the tension of unknowing. Translating is an open-sided walking that demands (and cultivates the faculty for) attention without intention. It demands that one be receptive but not directive, with a sharp ear out for the overtones and undertones, for the peripheries. It is to err into adjacent regions, which is to make these regions suddenly perceptible.

We tend to think of a text as an isolated instance, as an island of language in mid-air, when in fact, it is more analogous to an archipelago of certain words lit up (emerged) within a sea of words that surround it. Every text exists within a context of reasonable (or unreasonable) assumptions and associations that accompany any word. For instance, you can’t have “a wake” without either a boat or a body, though they need not be mentioned, and if you’re deft, you can get both. Boats and bodies in turn bring with them all sorts of other images and territories, and a good translation can wander into some of this territory, making it perceptible and broadening the path of the original.

In Stephane Bouquet’s book Un peuple, he cites the case of a French translation of Virginia Woolf’s Mrs. Dalloway in which, toward the end of the party scene, when Mrs. Dalloway is by herself, thinking of Septimus, the translator has added Mrs. Dalloway’s thought “la morte est une étreinte.” It’s not in the original, nor in any other translation. It is an error, but above all, it is an errancy. It’s a perfect example of the translator’s wandering into regions immediately adjacent to the original, regions implied by the original, even made necessary and inevitable by the original. The line was an overtone that this translator heard so loudly that it became audible to us all.

To err from the text is not the same as to err in the text—to err in it simply has us running in circles, creating noise that may or may not be useful, whereas to err from the text extends it; it creates space. It is to start with a thread and draw it outward into a world, even if the world is only one phrase long. In the Mrs. Dalloway case, it was five words, but beyond that, it was a generative action; it set the text in motion again by establishing within it a new vector that will be further extended in the reader’s mind, as it was in Bouquet’s when he leapt from there to his own next thought: “elle dit que la mort aussi est un instant de la matière.” It is translation and its errant errors that has brought that beautiful thought into being. This cannot be wrong.

The word errant also evokes the image of the “knight errant”: The errant always has something of the Don Quixote about it, and in turn, the sign of Quixote is always hovering over translation, which is why Borges chose this figure for his seminal speculative work on the subject. Quixote is the patron saint of translation, revealing it as both gallant and foolish, always on the side of right, and always just a little bit wrong. Like Quixote, translation creates a world based on, but not limited to, an independently pre-existing one, and like Quixote, it doesn’t recognize the difference between the two.

And like Quixote, it is always about to stumble in when uninvited, which is how many translators feel—they feel intrusive, and because of that they refuse to err. They follow the path of the original text very closely, but often this amounts to translating the book out of the original language, but not then writing it into the target language. Translation doesn’t end when the words have been exchanged; it must then start the wandering that will let the new text find its own path. This path is greatly limited, however, by our fetish for semantic meaning. That fetish, or tremendous bias, means that we trust the translator as proxy to choose different sounds but not different senses. It means we consider a text as meaning couched in sound, and not the other way around and not an equal collaboration of the two. And, of course, this makes translating poetry impossible.

To explore what might be possible if the translator were given as much liberty with sense as we normally accord him or her with sound, I began translating some of my own poems, noting how increasing the incidence of errancy opened new avenues into and out from the poem, changing it dramatically, but ultimately creating a text that was more “faithful” overall. Intrigued by some of the new directions introduced in the translation, I then re-translated them into English with equal errancy, creating chains that, like children’s Chinese Whispers, continue wandering outward. Here are three of those chains: