Reading Latour 6: Recouping Constructivism

With a Note on the Sciences of the Artificial

Bruno Latour. Assembling the Social: An Introduction to Actor-Network Theory. Oxford UP, 2005. From the chapter “Fourth Source of Uncertainty: Matters of Fact vs. Matters of Concern,” pp. 87-120.

For what it’s worth, this is the second longest chapter of the book, the longest being the penultimate one, “Third Move: Connecting Sites”, pp. 219-246.

Now Latour confronts the sociology of science, though in its newer guise as “science studies.” This is where he first made his mark, where I first heard of him, though did not read him. And this, as we all know, has been the site of some of the nastiest intellectual fighting of the past three decades, a war that bled out of the academy and into the general sphere. And it remains there, and in strange forms and strange company indeed.

But that’s an aside.

The set-up (p. 88): “We now understand why the word ‘social’ could entail so much misunderstanding; it confused two entirely different meanings: a kind of stuff and a movement for assembling non-social entities.” And it’s on that distinction that Latour will recoup ‘construct’ as in ‘science is socially constructed.’

We all know, of course, that science, like pretty much everything else in human culture, is socially constructed. That, as they say, is banal and trivial. And yet, that’s the fulcrum point on which Latour will re-construct social construction. Indeed, that’s what he’s been doing in the whole book, showing us why ‘the social’, but the social considered (constructed?) as assemblages of heterogonous networks of entities, is not at all trivial. That is, when you look, carefully, at how science or anything else is constructed, it’s not trivial at all. When seen carefully it all looks like those parodic contraptions we know as Rube Goldberg machines, named after, you guessed it, Rube Goldberg, cartoonist, engineer, and inventor.

That seems innocent enough. But, alas, that’s not how ‘the social’ was construed in much of this newer work on science in its social setting, and that’s certainly not how the opponents of this work construed the social. In these discussions it was the other social that was in force, the social as a “kind of stuff”. In that reading, the social construction of science meant that scientific knowledge was constituted of/by this social stuff and, as such, was pretty much arbitrary with respect to how the world works. That is, science wasn’t objective it all. It was just a bunch of ideas that the boys in the club agreed upon.

Them’s fightin’ words, and fight they did. But we know all that.

Construction, for Real

p. 89:

. . . in all domains, to say that something is constructed has always been associated with an appreciation of its robustness, quality, style, durability, worth, etc. So much so that no one would bother to say that a skyscraper, a nuclear plant, a sculpture, or an automobile is ‘constructed’. This is too obvious to be pointed out. The great questions are rather: How well designed is it? How solidly constructed is it? How durable or reliable is it? How costly is the material? Everywhere, in technology, engineering, architecture, and art, construction is so much a synonym for the real that the question shifts immediately to the next and really interesting one: Is it well or badly constructed?

And then we go: “Even more so than in art, architecture, and engineering, science offered the most extreme cases of complete artificiality and complete objectivity moving in parallel.”

And that’s that. Science is artificial and objective. More, science is objective because it is artificial.

Alas, others didn’t see it that way. For many (p. 90):

To say that something was ‘constructed’ in their minds meant that something was not true. They seemed to operate with the strange idea that you had to submit to this rather unlikely choice: either something was real and not constructed or it was constructed and artificial, contrived and invented, make up and false. . . . but it flew in the face of everything we were witnessing in laboratories: to be contrived and to be objective went together.

And thus a simple, but not so simple, confusion about ‘the social’ led to a mis-construal of construction and from that, the science wars.

I agree with Latour, of course, on the constructed nature of science, and of the deep compatibility between artifice and objectivity. Artifice can be used for many ends, objectivity is one of them.

But, rather than continue along that line, or continue further into this chapter—we’re only four pages in—I want to look at a different part of the intellectual landscape, the cognitive sciences.

Constructing Minds (Construct Science)

It is no accident that one of the great informal expositions of cognitivism should have ‘artificial’ in the title. I refer, of course, to Herb Simon’s The Sciences of the Artificial (1981). In the first chapter we find these words (p. 6):

Natural science is knowledge about natural objects and phenomena. We ask whether there cannot also be “artificial” science—knowledge about artificial objects and phenomena. Unfortunately the term “artificial” has a pejorative air about it that we must dispel before we can proceed.

My dictionary defines “artificial” as, “Produced by art rather than by nature; not genuine or natural; affected; not pertaining to the essence of the matter.” . . . Our language seems to reflect man’s deep distrust of his own products.

It’s not the same as Latour’s problem with ‘the social’, but it’s cousin to it.

That little book has chapters on a number of things, including economics and social planning, but mostly it’s about human thinking: How do we think? What are the mechanisms and processes?

Much of cognitive science, as you know, proceeded by way of computer modeling. Some modeling was done in service to psychology, constructing models intended as accounts of how people thing. And some was done in service to achieving some practical result, like medical diagnosis or translation from one language to another. The latter work tended to be done under the banner of artificial intelligence while the former generally flew under cover of cognitive psychology or linguistics. But often it was, and remains, difficult to tell the difference between the two.

And so we have a rather large intellectual community devoted to constructing the mechanisms of (human) thought.

And that’s what it really was and is, construction, engineering, artificial. So, in one division of the academy, Latour and his colleagues are figuring out how the social and, in particular, scientific practice, is constructed. In another division, Simon and his colleagues are constructing the mental. Complementary activities, no?

Yes.

Culture as the Mental and the Social

But where and how do they meet? The constructors of the mental were not, for the most part, concerned about communication, how minds interact. And I say this knowing that machine translation has been one of the core problems in this domain, as well as designing the man-machine interface: How do humans interact with machines? But these are not conceived as fundamentally problems of communication; rather, they are engineering problems with a communication component.

The knowledge models that have been created, have pretty much been construed as knowledge in a ‘transcendental mind’ that knows everything in the relevant domain and so has no problems about the compatibility of one chunk of knowledge with another. And yet such problems abound in the real world. No one physicist knows all of physics. Nor even does one expert in, say, classical mechanics know that whole domain. So the knowledge of these various experts is partial, even within their respective specialties. Which means that no existing mental representation of scientific knowledge, that is, existing in a real human brain, covers the whole domain. Each representation is partial.

So how doe these experts, each with partial knowledge, partial representation of the domain, keep current with one another? How does a particular scientific community act AS THOUGH it had access to some ‘transcendental mind’? It seems to me that THAT question is firmly in Latour’s domain, the social construction of science.

About that, I offer one thought. It’s about the role of replication in scientific practice. Laboratory A conducts an experiment and publishes the result. The result is interesting, and potentially important. So scientists in Laboratory B and Laboratory C attempt to replicate it. That is, if they use the same procedures, will they get the same result? If so, the result will have been replicated. If not, if no replication, then the result will likely disappear.

Replication is a way to check the science. But just what is being checked? The particular proposition that is regarded as true in consequence of the experiment? It seems to me that that’s where the emphasis lies, and perhaps properly so. But it also seems to me that what is in fact being checked is the entire assemblage of things and processes that enter into making and communicating the observation. A replication can fail because the hypothesis is wrong. But it can also fail because the original experimenters inadvertently left out a crucial bit of information about apparatus or procedure.

So, when knowledge is distributed among a collection of investigators, just how do we draw the line between the mental and the social? Culture, it seems to me, is like that, both mental and social, residing in individual brains, but allowing them to operate as if they were one. Culture is magic, true magic.