Translation : <

To say that “difficulties” of this sort occur “structurally” in natural languages and that they crop up across historical moments and locations would not seem terribly different from the rigid fantasies of designation and of value-indexing attached to the “possible worlds” of the global market and of the commodity-practice of translation. But it is. Let us say that in place of widgets, in place of the commodity I hope to produce with my Nicaraguan partner, my translator and I are discussing the related commodity that my translator is hoping to produce for my consumption, the second-order commodity or commodity-practice that is his or her translation. We are not talking about an object either of us can hold or produce as one would a widget (presuming for the moment that a “widget” is a really subsisting, material object): what we are discussing, roughly speaking, is the object we call “a natural language,” or two such languages, and the relation between them. We are discussing the uses to which that relation is to be put; the costs associated with moving between them; the value of the product my translator delivers; and so on. The Mandarin or Spanish translator and I will have to agree that we are engaging in some sort of intentional, communicative expression; and we will have to have a minimally similar understanding of what it means to “translate” when I set about hiring him or her.

Here, where the object under discussion is not a widget but the means of designating a widget in two different languages, that is, where the object designated is what we call “translation,” the threat of a poisonous regress is unavoidable. This poisonous regress distinguishes this level of discussion concerning this sort of object-which-is-not-an-object, called a “natural language,” from the discussion my translator and I might have regarding a widget, however the reality of such an object is to be imagined. And note: poisonous though it may be, this threat of regress is also, crucially, the source of unassimilable values and designation-effects—values and effects immeasurable according to the ecologies of either language and of either market, original or target.

When we say that some word or expression in Mandarin or in Spanish is a translation of the English “widget,” and that this holds true in all possible worlds because of the logical requirement imposed by the notion that we are indeed “translating” between languages, then we seem to be required to stipulate that designation conventions are, if not identical, then sufficiently similar among natural languages that “widget” may be said to designate in much the same way in the original language as in the target—independently of what, in fact, is designated by “widget” and its translation in Spanish or Mandarin.

What obtains for the concept of “designation convention” is also true for the concept of “translation,” and in general for metalinguistic terms that natural languages use to designate themselves. Such metalinguistic terms may indeed be said not just to “designate,” but to “translate” the natural language into the condition of being a discursive object (composed of statements, a lexicon, conventions of usage, syntax and so on) named or designated by a metadiscursive statement, or a name, in that same natural language. As you might suppose, what’s happening here—whatever process I am designating when I describe this objectification of a language from within that language, this metaphoricization of the language by means of itself—is that this process is neither instrumental nor aesthetic simply.

As to its strictness or rigidity across possible world, you will get a sense of how tricky the situation can become when you note that metalinguistic terms are not always and exclusively metalinguistic in a given context (metalinguistic statements, like performative utterances, are always, an early Derrida might put it, parasitized by context); when you note that certain “object” terms can be (and often are) used tropically with a metalinguistic function; and that the rules for deciding when a term is being used in one way or another are themselves conventional, partial, changing, and contentious, when not contradictory, and when they are not also the consequence of such metalinguistic uses.

Let us call the widget that produces intra-linguistic translation- and designation-effects of this sort Machine Translation, the spider in the linguistic cup before Turing and before the Cold War. This widget produces the flickering, undecidable movement between statements’ (or names’ or terms’ or metaphors’) linguistic and metalinguistic status. It works between statements and what they designate; it spiders away to tip a particular thing designated into the ensemble of designation conventions, lexicon, and syntax that make up the natural language within which that designations take a place. It is internal to the metalinguistic statement or term, its designation, and the object it designates. What does this sort of machine translation look like?

I will give you an example taken from the canon of phrases regarding translation, repeated mechanically since Quince (one of the “rude mechanicals” of A Midsummer Night’s Dream) first uttered it some four centuries ago in an infamously haunted forest, treading heavily among fairies, monarchs, and translated lovers. We are the audience at Shakespeare’s most subjunctive play, his most extravagant exploration of the fantasy of possible worlds, the most radical staging of philosophical modality in the theatrical canon. (Prose is another matter.) Shakespeare peppers his text with the most severe marker of unconditionality, the word “never”—used here over thirty times, more often than in any other comedy (with the exception, oddly, of Much Ado About Nothing). “Never”—yes, but set against the most basic and pervasive form of counterfactual wondering. What if the woodland held a Fairy court? Is a world possible in which a weaver can become an ass and stay a Bottom; where domestic dilemmas of Athenian politics are solved extramurally by sprites, mechanicals, actors; where senses become disjoined and garbled, voices can be seen, and faces heard (“I see a voice: now will I to the chink, / To spy an I can hear my Thisby’s face,” says Bottom’s Pyramus?)¹⁶ The theatre, of course, is always structurally counterfactual, a staging of modality—but this is Shakespeare’s most alarming, seductive and rigorous enacting or translating of that staging.

^{16. William Shakespeare, A Midsummer Night’s Dream, Act 5, Scene 1, Lines 192-193 in The Riverside Shakespeare: The Complete Works, ed. G B. Evans and J J. M. Tobin (Boston: Houghton Mifflin, 1997).↩}