Translating Infrastructure

I’m going to begin with something I’ve discussed many times before: David Olson’s contention that the scene of writing (it might be more precise to say the scene of prose) is predicated upon the reporting of speech outside of the presence of the listeners. So, to review: according to Olson, if, in an oral culture, one person, on one scene, is reporting the words of another person, spoken on another scene, the reporter will fill in the context of the original scene by imitating the speaker. Instead of saying something like “he angrily responded,” the reporter would speak in an angry sounding voice, affect an angry looking posture, and so on. The problem (alphabetic) writing, as a representation of spoken words, had to solve, then, was how to supply the context that would have once been part of the enactment of the speech. The problem is solved by the invention of new words, first of all verbs, and then nominalizations of those verbs, culminating in what Olson calls the “metalanguage of literacy,” which allows us to represent a whole panoply of imputed stances to the original speaker. And, of course, today, as literate men and women, we do this when speaking as well, rarely (but not never) enacting the attitudes we wish attributed to those whose speech we report.

I’ve done my best to radicalize this “hypothesis” of Olson’s, by drawing out a couple of implications, with the help of other elements of Olson’s lifelong study of literacy. It’s telling that Olson completely takes on board the notion of “classic prose” developed by Francis-Noel Thomas and Mark Taylor—on the one hand, Olson shares with Thomas and Taylor a basically scenic notion of writing; on the other hand, he also agrees with them that it is the nature of prose to efface the scene of writing, and create the illusion that writer and reader are present on the scene of reported speech. This is the understanding of good writing familiar to everyone: with good writing, the presence of the writer, and even of the language, is invisible—writing should be like a window onto whatever is being “shown” us. So, there’s a kind of massive “cover-up” built into the constitution of writing and literacy. It seems to me that any writing that plays along with this without drawing any attention to it is to that extent a falsification. But the falsification lies less in the style or straightforward reportage (but somewhat there as well) than in the entire vocabulary, generating an entire theory of “mind” and sociality, that conceals the media, or mechanisms, or infrastructure, that in fact makes the reported scene accessible. This is the—it seems to me quite relevant—truth of Derrida’s concept of “logocentrism.”

One next step to take here is to say that, if this is the way writing operates, and writing is a fundamental mode of technics, couldn’t it also be the case that a similar kind of concealment is constitutive of all technics? I think we’ll be getting to that question by and by. For now, we have not exhausted the implications of the infrastructural dimension of writing. Take one of Olson’s basic examples of an element in the infrastructure of the metalanguage of literacy: the verb “imply,” along with its nominalization, “implications.” If I write (or, as a literate person, say) of someone that “he implied the authorities were lying,” the word “imply” is doing some important work here. If we imagine an original scene of speech being reported here, we’d have to imagine the reporter indicating this “implication” through his reporting of the speech. The original speaker in question said something, and it wasn’t “the authorities are lying,” and from whatever he said, we are able to draw the implication that that is what he thinks, or “really means.” I’m not sure it’s very easy to imagine such a speech in a purely oral setting: we’d have to imagine, for example, someone saying something the authorities have said but in such a way, through vocal and bodily means, to enable his listeners to conclude that he considers the words he is actually repeating to be a deliberate falsification on the original speaker’s part. That’s quite a performance—to really make it work, we might have to further imagine an audience who knows the speaker and is familiar with his attitude toward the authorities.

If this kind of paraphrasing already strains our ability to imagine an original scene of speech upon which our understanding of the statement is nevertheless predicated, then we must accept that the infrastructure providing us with access to this understanding of the authorities has thoroughly supplanted any original scene of speech which could be provided transparently. And, to make this deconstruction complete, we must accept that the infrastructure was “always already” infiltrating the oral scene itself, in its scenic conditions or articulation. The importance of drawing attention to the metalanguage of literacy, then, rather than assuming its transparency to “mind,” is to read off of it precisely the infrastructural conditions of the utterance. A great deal of infrastructure must be in place for someone to imply the authorities are lying, and for others to grasp that implication. The authorities have various means of communicating its declarations and commands; its words are transported and repeated by people occupying various offices at various ranks; the possibility of introducing a note of skepticism into those declarations and commands itself depends upon differences among elites and dysfunctions in the chains of command, and what is taken to be an implication, and how that implication is received and recirculated, solicits the entirety of this system. So, when we talk about “implications,” “assumptions,” “premises,” “suggestions,” “considerations,” “understandings,” and so on, we are not referring to properties of mind to be sorted out philosophically but “properties” of infrastructure of knowledge and communication—but that, today, also means the infrastructures of sensing, measuring, data collection and curation, and algorithmic analysis that undergird knowledge and communication. This is all “in” our language, all the time, and can be elicited and extracted from our discourses regularly. To return to Olson: the creation of alphabetic writing involved turning language into an object of inquiry, one that can be broken down into such entities as “phonemes,” “words” and “sentences.” The next step is to make the inquiry into linguistic properties that all of us participate in as literate people itself an object of inquiry into the scenic conditions of possibility of the “objects” in question.

I’m reviewing this material in order address another question. The question is: what should discourse do? I’ve been moving away from metalinguistic cliches like “explanation” and “description,” which never really stand up to much scrutiny and, more importantly, presuppose the classic prose scene of transparency: when you purport to “explain” something, you are effacing your own participation on the series of overlapping scenes that includes both the one you are providing reasons for and those from which you are deriving the reasons. When you say something like “the revolution happened because of rising expectations frustrated by a stalling economy following a long period of growth” all of these phrases refer to formulaic summaries of areas of disciplinary activity, boiled down into abstracts, journalistic “think pieces,” etc.—that is, you’re referring to a whole infrastructure while pretending to be on a scene where some question is genuinely settled. You are, in fact, translating that infrastructure, which I don’t object to because that’s all you can do; it’s the pretense, usually involuntary, that is objectionable.

So, rather than speaking in terms of “explaining,” “describing,” “accounting for,” etc., I can speak in terms of translating infrastructure. You’re translating infrastructure, they’re translating infrastructure, I’m translating infrastructure, we’re all translating infrastructure. Now that that’s settled, toward what end are we translating infrastructure? Perhaps we should first ask, what are we translating infrastructure into? Something other than infrastructure?—what would that be? I would use an analogy here: translating infrastructure is like converting a declarative, along with the entire ostensive-imperative world it entrains, into a new ostensive. So, translating infrastructure is like an infra-lingual translation. You want what you say to have maximum impact, don’t you? And you want that impact to be one you could continue to recognize as one you initiated and can continue to contribute to, as that impact plays out, don’t you? That’s what I’m talking about here—making your utterances, your samples, performative, and doing so with increasing explicitness. There’s a kind of transparency desired here as well, but one that takes in rather than rendering invisible the surrounding enabling infrastructures.

Now, infrastructures are massive and distributed, and can’t be translated all at once, at least not in any way that can be brought within the scope of an utterance one might respond to. It’s always another’s translation of infrastructures that one is in turn translating. So, we’re talking about something like dialogue or conversation here, but one that accepts the staged, mediated, “disseminational” character of any dialogue or conversation—that is, that accepts the constitutive sense in which none of the participants are present for it. To translate is to carve out within one body of language an approximation to a selected portion of another body of language. The best English translation of, say, Cervantes, is one that would have the same relation to English as Don Quixote has to Spanish—a very challenging and probably unattainable way to think about translation. But to approximate is not to attain, and approximation is all we need. So, your sample has a relation to the whole that I try to approximate in my sample, remembering that the whole is now different because your sample has been added to it and my own sample is about to be. What my approximation adds to yours is the rendering of some relation to the infrastructure more explicit, and the request or invitation to have you do the same to my sample. What we would all be approximating in that case is a discourse which is the cooperative adjustment in infrastructural arrangements.

You can see that this is a discourse that is becoming more like technology insofar as it involves the exchange of imperatives that are continually measured through declaratives and generating new ostensives. To the extent that we are all respectively referring to the same sample, imperatives to introduce infrastructural adjustments that generate more ostensives that we can scramble to authenticate are being issued. We participate in society as inquiry as samples and inquirers—this model of discourse is the approximation of the originary scene once the replacement of the ritual order by the technological one is complete. We are now prepared to undertake the inquiry into all infrastructure, all scenic design practices, on the model of the installation of the literate order. Any device replaces, not so much an appendage of the body, as a scene or, rather, the results of a sequence of scenes that would have had to have played out to arrive at the results the device replaces while making unnecessary our own enactment in that chain of scenes. The implication for our relation to our devices is not that we should try and find our way back to the next scene that would have been in that series—we will never succeed because it doesn’t exist. What seems like the natural human relationship replaced by the device was another infrastructure. What you can do is work on translating the infrastructural inflection introduced by the device into our language as we now find it, transformed, however slightly, by the infrastructural modification. And you don’t have to be an engineer—you just need to hew very closely to emergent idioms, linguistic shifts induced by the infrastructural adjustments, and have faith that there must be some relation, which we can performatively translate, between each and every idiomatic difference and some rumbling in the infrastructure—even if we don’t know what it is yet. That’s the relation that we set about reciprocally provoking one another to approximate. And I refer to “faith” here because if there is something genuinely, irremediably wrong with a particular technological innovation that will show up in the language—there will be words that don’t work the same way, and can’t be made to give way to words that work better; there will be sentences with elisions that no one can “fill in.” And noticing such things will also be your guide toward repairing what has been broken, which will always mean creating something new on the model of infrastructural possibilities brought into view by your reparative inquiry.