Translation as Class Struggle

Capitalism, Marx taught, is all about getting something for nothing. Labour “produces” because workers give capitalists the free gift not only of part of their time, but also of part of their inheritance in the commons.

If that inheritance disappeared, so would profit. Hence the paradox and challenge of capitalism. Insofar as they accumulate at all, capitalists have to deny, through their daily practice, the existence of what makes accumulation possible. Commons must be enclosed. This denial is elaborated and made intellectually plausible by the gigantic, imposing body of orthodox economics. Yet that same denial, pursued far enough, leads straight toward failure to make a profit. Capitalism's millions of ingenious techniques for transforming human activity – or the commons – into an income stream, it turns out, are ultimately both necessary and impossible.

At the core of all capitalist labour lies interpretation. That's no surprise, since interpretation lies at the core of all social activity. What may be less obvious is that the advent of automatic machinery only expands the need for labourers' interpretive work. Computers carry the process even further. Today nobody needs scriveners like Herman Melville's Bartleby to read and copy legal documents by hand. But they do need a thousand times more workers to decide when a complaint about a photocopy machine merits a repair visit, to interpret what a fast-food customer wants, or to proofread word-processed documents and make sure the results arrive at their proper recipients.

Even more so than in the 19th century, capital's profits in an age of throwaway service labour rely on exploiting and monetizing the basic linguistic socialization given free of charge to each of its future employees between the ages of zero and five. True to form, capitalists are compelled to deny this dependence. Nobody talks about everyday interpretation as “work” at all, much less as skilled labour. Some may even dream that the obligatory drive to mechanize interpretation will someday not only put human interpreters in their place but make them superfluous. Rule-following would be so much more streamlined and efficient if it didn't require interpretation at all. Yet every move toward mechanizing or eliminating interpretation runs hard up against the reality that without the interpretive labour of ordinary workers, no capitalist could accumulate anything. As Wittgenstein pointed out, no rule can ever contain its own interpretation.

Some of what capital finds recalcitrant about interpretation as a labour process lies in the judgement calls interpreters have to make hundreds of times a day between (on the one hand) holding beliefs fixed and solving for meaning and (on the other) holding meanings fixed and solving for belief. Was that guy on the news really referring to Obama's “clowning achievement” or did I mishear him? And if I heard him right, is that actually what he meant to say? No innovation of capitalist civilization can short-circuit this process or render it less autonomous without paying a price.

Dictionaries, which correlate meaning with meaning and let belief fall out where it may, can never be more than an occasional, bookish appendage to interpretive labour. Google Translate may well speed up the global circulation of capital, but at the same time it multiplies potential misunderstandings that legions of human workers then have to correct. There's always more to what Marx called “contradictions” than is spelled out in the pages of Capital.

Of course, what's being introduced here as if it were a “capitalist” dynamic is more than that. Innovations like Renaissance perspective, or the movement toward a more “universal” time spurred by the development of polyphony, helped to make capitalism; capitalism did not make them. So too for the invention of reified and official languages and exchangeable units of meaning. These phenomena already mark eras usually designated as precapitalist, even if they develop much further afterwards. But to make this historical point is only to re-emphasize that interpretation is a site of political contestation, even class struggle.

It makes no difference to what is at issue here to resort to an example from (inter-linguistic) translation rather than (intra-linguistic) interpretation. The normal professional English translation of the Thai word ngaan is work. An English speaker will probably promptly reinterpret this to mean wage labour. As it happens, that translation may not turn out too badly, since over the last century of Thai history wage labour has come to dominate the meaning of ngaan as it did with work in Europe in earlier eras. The complication is that older meanings have continued to haunt the term more than they do work. Ngaan also signifies festival, ceremony and party as well as denoting various religio-agricultural practices, commons activities and other pursuits that have nothing to do with capital accumulation.

Respecting this diversity but reorganizing it in a capitalist framework, a professional or Google Translate approach necessarily re-enacts the split between salaried labour and cultural activity by leaving the word work out of translations of, say, ngaan wat (temple festival), ngaan taengngan (wedding) and ngaan chalong pii mai (New Year's celebration). A professional process like this is well-suited to minimizing friction in circumstances of particular concern to certain classes: conferences on economic development, time-sensitive commercial negotiations, tourism.

But what if these are not our circumstances? Suppose our livelihoods are tied up instead with contesting proletarianization and its disciplines. Then why shouldn't we encourage a translation of ngaan jaang as hired ritual instead of salaried work and ngaan sop as corpse work instead of funeral? That might help reverse the polarity according to which wage labour is dominant over the care and defence of commons and undermine any smooth vision of capitalist universality, while at the same time facilitating a useful Thai cultural critique of the West. By making Thai strange to English speakers, it would stimulate further inquiry rather than closing off its possibility. By making English itself stranger, it would honour Rudolf Pannwitz's rhetorical point that – to paraphrase – translations of Thai into English should not render Thai in English but make English more like Thai.

For professional translators, of course, the weirdness of a phrase like corpse work signals failure. Something must have been lost in translation. For vernacular translators animated by a different political awareness, however, nothing is ever lost in translation because there is nothing to lose. For them, it is sheer fetishism to posit pre-existing “meanings” or “beliefs” that are fixed ahead of the translation process itself. In accord with the rest of democratic politics, that process will always be negotiable. In this connection, it may be fitting that the same Amazonian intellectual tradition that helps sustain some of today's most fundamental, cutting-edge resistance to neo-extractive capitalism apparently also sees translation in this way, as what Eduardo Viveires de Castro describes as a process of “controlled equivocation – 'controlled' in the sense that walking may be said to be a controlled way of falling”.

Capital is likely to continue to have certain structural preferences for “instant” – or even mechanized – translation: for a political process that cuts off intercultural inquiry at a point where pre-existent shared meanings or beliefs (or “local knowledges”, or “individual preferences”) are said to have been reached. Correspondingly, effective anticapitalist activism – not to mention the commons that capital ultimately rely on – will continue to have to defend strategic spaces for translation processes that are not only slow (in the way that “slow food” is slow) but also tirelessly open-ended, indeterminate in outcome, and resistant to such mystifications.

Powerpoint presentations for the University of East Anglia and Binghamton University (available above) illustrate some of these points.

For a related, more recent exploration of the political issues involved in efforts to mechanize translation, see "Interpretation Machines: Contradictions of Artifical Intelligence in 21st-Century Capitalism," http://www.thecornerhouse.org.uk/resource/interpretation-machines. See also http://www.thecornerhouse.org.uk/resource/what-kind-future-world-rainfor....