Reading Latour 0: Ontology, Methodology, Compositionism

(Revised 12 Oct 2011)

Bruno Latour. Reassembling the Social: An Introduction to Actor-Network Theory. Oxford UP, 2005.

An Attempt at a “Compositionist Manifesto”, New Literary History, 2010, 41: 471-490.

Though it’s not so long ago that I decided to read Latour, I don’t recall the exact route that led me to his work. It might go like this: I was intrigued by the notion of Latour litanies, read a post by Graham Harman where he referred to one he’d quoted in his study of Latour, Prince of Networks, and so downloaded that book, as it was available as a free download. Once I got the book I looked up the specific passage Harman had mentioned, a quotation from Richard Rhodes on p. 103, and then looked around in Harman’s book, finally deciding I needed to read Latour himself.

I suppose I choose Reassembling the Social both because it is relatively recent (2005) and for the systematic exposition promised by its subtitle, An Introduction to Actor-Network Theory. What I read struck me as a fascinating and brilliant, if slightly odd, treatise on method:

To describe the social world, this is what you do, how you do it, and why you do it.

The many examples Latour gave were interesting but, given what he was saying, individually a bit thin (I assume he has thicker descriptive work in other books), as were the specific how-to-do-it suggestions. There was some talk of notebooks in one chapter, but that was as much a thought experiment as a practical suggestion. It’s as though the methodological flavor served as an elaborate metaphor for something else, indicated in the whys. And that something else is philosophy, more specifically, ontology.

It’s as an ontologist that Harman presents Latour, but as a peculiar one. Thus (Prince of Networks, 14):

Latour always insists that we cannot philosophize from raw first principles but must follow objects in action and describe what we see. Empirical studies are more important for him than for almost any other philosopher; later in his career he will even speak of ‘experimental metaphysics.’

A bit further on (p. 16):

Whereas Latour places all human, nonhuman, natural, and artificial objects on the same footing, the analytics and continentals both still dither over how to bridge, deny, or explain way a single gap between humans and the world. While graduate students are usually drilled in a stale dispute between correspondence and coherence theories of truth, Latour locates truth in neither of these models, but in a series of translations between actors. And whereas mainstream philosophy worries about whether things exist independently of us or are constructed by the mind, Latour says they are ‘socially’ constructed not just by human minds, but also by bodies, atoms, cosmic rays, business lunches, rumors, physical force, propaganda, or God. There is no privileged force to which the others can be reduced, and certainly no ceaseless interplay between pure natural forces and pure social forces, each untainted by the other. Nothing exists but actants, and all of them are utterly concrete.

Nothing but actants—most provocative, and most satisfying. That’s what I didn’t (quite) know when I started this project, and, correlatively, that’s something I’ve learned. And a very important lesson it is.

It is, however, one thing to give quick assent to the notion of a flat ontology in which no object is endowed with more Being than any other object. For reasons grounded in my own work—in particular, my work on music—I found that notion congenial when I first found it in posts by Tim Morton. But it is something else to see the idea systemically developed by a different thinker, from a different background. THAT is what I found in Reassembling, that is what gave off an aura of the new.

And I can see it there in the Jane Bennett’s very different work, Vibrant Matter: A Political Ecology of Things. While the primary scientific literature only shows up sporadically in the book—at least as I can see from a quick skim—it’s no more than a citation or two away from what DOES appear in the book. And that literature throws doubt on the nice neat division of the world into human, animal, vegetable, mineral, and all the rest, that division that’s embedded in our grammar. What happens if one is so bold as to insist on taking that neat division out of the specialized literature, out of the specialized disciplines, and bring it into the common sphere?

But I digress.

What I’m now wondering is whether or not Latour is neither a philosopher, nor a sociologist, but something new, for which he has provided a new term, a compositionist who thus practices compositionism. In saying this I am not suggesting that compositionism is a new discipline, taking its place along side the various old disciplines, such as physics, geology, history, sociology, and so on. For where would it situate itself in such a scheme, what School? Humanities? Social Sciences? Natural Sciences? Engineering? None of the above, all of the above?

No, I am suggesting that compositionism is something else, a new way of being disciplined, rather than yet another discipline trying to force its way into the scheme laid down in the 19th century German university.

As an oddly parallel case I offer cognitive science. The phrase was coined in 1973 by H. Christopher Longuet-Higgins, who was trained in and had appointments in the physical sciences but somehow ended up doing computational work on the human mind. He begins the preface of Mental Process, a 1987 collection of his papers, like this:

This is the log book of an expedition into the mostly uncharted territory of the mind. When in 1967 Richard Gregory and I packed our bags for Edinburgh, it was in the shared conviction that the workings of the mind could not possibly be as tedious as the psychologists made them out to be, or as peripheral as the physical scientists tended to assume.

I have no idea whether or not Longuet-Higgins was the first physical scientist to gravitate to a primary interest in the human mind—most likely not—but he certainly wasn’t the last.

As for the polyglot field he named, except for a program here and there, it never really took root as a university discipline. It has existed mostly as an interdepartmental mongrel, albeit one with its own journals and professional societies. A cognitive scientist specializing in cognitive psychology is, ipso facto, a psychologist; but a cognitive psychologist is not, ipso facto, a cognitive scientist. And so it goes with various linguists, philosophers, computer scientists, and others. To my mind, for my taste, the steam went out of cognitive science perhaps two decades ago. But that’s neither here nor there.

My point is that cognitive science was something else, a something else that never fit into the 19th century scheme of intellectual being. That in itself is not so strange. Biophysics, for example, is newer than that, and it is a well recognized specialization. The are many such hybrids in the contemporary university. The department where I was trained, for example, the Department of English at the State University of Buffalo, had programs in Literature and Psychology, Philosophy and Literature and Literature and Society. But these examples all span a relatively ‘compact’ interdisciplinary arena. Cognitive science sprawls across linguistics, computer science, logic, philosophy, psychology, anthropology, and neuroscience, a much wider and somewhat more mysterious and tenuous remit.

Compositionism feels more like that. Here’s Latour’s own gloss on the word (PDF):

Even though the word “composition” is a bit too long and windy, what is nice is that it underlines that things have to be put together (Latin componere) while retaining their heterogeneity. Also, it is connected with composure; it has a clear root in art, painting, music, theater, dance, and thus is associated with choreography and scenography; it is not too far from “compromise” and “compromising” retaining with it a certain diplomatic and prudential flavor. . . . Above all, a composition can fail and thus retain what is most important in the notion of constructivism (a label which I could have used as well, had it not been already taken by art history). It thus draws attention away from the irrelevant difference between what is constructed and what is not constructed, toward the crucial difference between what is well or badly constructed, well or badly composed.

Compositionism is about construction, about how things are assembled into other things.

There is, of course, nothing inherently new in that, as Latour acknowledges in listing kindred precedents: “. . . art, painting, music, theater, dance, and thus is associated with choreography and scenography . . .” For that matter, mathematics is almost pure construction; one starts with axioms and postulates and builds a mathematical system. Engineering is all about composition, design, and construction, about creating something from nothing, as I’ve argued in this post. And biology, while classified as a science, has, to my mind, an engineering feel about it. The cell is a chemical factory, multi-celled organisms are constructed of single cells organized into tissues of various types, organisms interact with one another in complex ecological fields, and so forth. And much of cognitive science has this engineering feel about, this compositionist aura. I’m thinking about Chomsky’s notion of a grammar as a set of rules from which sentence structures are composed in the manner of a mathematical proof and of all those simulations of mental models. All this feels much closer to engineering than to, say, physics, where one derives a universe from a handful of laws.

If, then, compositionism has been around for some time in various forms, where is the novelty in Latour’s proposal? That, it seems to me is at one and the same time, obvious and not at all obvious. It’s there in the flattened ontology and in the emphasis on description, which is no small thing, no, not at all. In Reassembling the Social Latour mostly urges description upon us rather than producing extensive descriptions himself. That is why I’ve introduced graffiti into so many of my posts. While the descriptive work I’ve done is relatively informal, there is enough of it, and of enough kinds, to give some sense of what it would be like to describe graffiti-world in some detail.

Beyond that, compositionism’s novelty can be seen in Latour's call to redraw “the boundaries between sociology and psychology” (213). But, as I have indicated in the appendix, Three Objects for OOOIII, he hasn’t really provided any conceptual tools for doing that beyond his metaphor of the plug-in. The cognitive and neurosciences, on the other hand, had plenty of tools for doing that and, in my description of the music-making group (see RL11: Plug-ins and Couplings, RL11: The Cartesian Individual, and in the appendix) I have composed a neural-level object that crosses, and in a sense even dissolves, that boundary.

A thorough description of graffiti-world would have to ‘reach’ into the graffiti itself and deal with those designs, which I have not done in these posts. THAT descriptive work, inscribed on a bridge between the social and the psychological, that would be deeply NEW. Given that description, what THEN, and ONLY then, would we be able to see about graffiti? Of course we do not know. We have to do the work first, but not only for graffiti world, but also, for example, semi-conductor fabrication world, quilting world, YouTube video world, urban hydroponics world, Arab Spring world, and so forth. In all those cases, and many more, we need to weave our compositionist webs back and forth over the abyss currently separating the psychological and the social. There, on the far side of descriptions as yet unwritten, there, I suspect, is where the compositionist frontier lies.