Creature of habitus: Latour on Bourdieu Pt. 1

Hi guys. Back from my road trip and returning to active duty. I definitely want to weigh in on the plasma/secularism question, but in the meantime here is the first of a few posts I started preparing before the holidays, on Latour and Pierre Bourdieu.

The subject of Latour's relation to Bourdieu is so complicated that I'm going to have to split it up over several posts. Before I get started, I'll quickly indicate what I think the complications are. There are basically four: (1) Bourdieu was arguably the central figure in French sociology from the early 1960s until his death in 2002 and Latour is, from what I understand, frequently seen to be his successor in this role; (2) Latour is heavily critical of Bourdieu in Reassembling the Social; (3) Bourdieu, for his part, attacked Latour and science studies in one of his last books, Science of Science and Reflexivity; and finally (4) Bourdieu has been a very strong influence on me personally, basically serving as my entry point into the world of sociology and social theory — so for me to accept Latour's pretty much wholesale dismissal of "critical sociology" requires a good deal of rethinking and unlearning on my part, a process which is still taking its way. (And this is interesting, actually, because one of the points of disagreement between Bourdieu and Latour is, essentially, about the way learning works, and how permanent and conscious it is — more about this in a moment.) So we have (1) what sociology thinks of Bourdieu and Latour, (2) what Latour thinks of Bourdieu, (3) what Bourdieu thinks of Latour, and (4) what I think of both of them, and whether they admit of being combined — that's a lot to cover. I'll do my best.

In this first post I'll have nothing to say about (3) — Bourdieu's Latour — and I'll focus on (2), with (1) and (4) perhaps bleeding in at the edges.

According to the index of Reassembling the Social, there are ten direct references to Bourdieu, though he's present by implication in many other passages — whenever Latour starts referring to "critical sociology" or "reflexivity," for example. In fact, despite the general spirit of opposition, a number of the citations of Bourdieu's work are rather admiring, though nearly all of these refer to Bourdieu's 1977 Outline of a Theory of Practice (cf. 101, 169, 175). Thus it seems fairly easy to see that the Bourdieu that Latour values is the close examiner of local interactions and the thinker of habitus, not the analyst of fields and capital that emerges in Bourdieu's work from Distinction on. This is a Bourdieu that Latour can plausibly associate with interactionism and ethnomethodology, a "micro" or "local" Bourdieu, as opposed to the "macro" or "global" Bourdieu we get with the subsequent development of field theory. This is consistent with Latour's general resistance, all throughout Reassembling the Social, to "contexts" or "frameworks" of any kind: it makes sense that he would like the "genetic" Bourdieu that is attentive to the tiniest behavioral tics and traces, but not the "structural" Bourdieu that uses all this data to build new versions of the class structure to correlate with this behavior. That is, Latour doesn't want to know, or claim to know, what the outside of the network consists of; he doesn't want to "fill in the blanks" as he puts it somewhere.

So one could imagine many of Bourdieu's analyses of social causation of seemingly natural human behavior adapted to actor-network theory — in a sense he's all about breaking up the nature-society divide, after all — but not the claim that such causation forms a rational structure or system. The point here is, Latour likes habitus, and he makes advances on the concept in the latter part of his chapter "Second Move: Redistributing the Local." For instance, here's Latour at his most Bourdieuvian (in thought, if not in style):

How many circulating clichés do we have to absorb before having the competence to utter an opinion about a film, a companion, a situation, a political stance? If you began to probe the origin of each of your idiosyncrasies, would you not be able to deploy, here again, the same star-like shape [i.e., a network] that would force you to visit many places, people, times, events that you had largely forgotten? That tone of voice, this unusual expression, this gesture of the hand, this gait, this posture, aren't these traceable as well? And then there is the question of your inner feelings. Have they not been given to you? Doesn't reading novels help you know how to love? How would you know which group you pertain to without ceaselessly downloading some of the cultural clichés that all the others are bombarding you with? (209)

Again, though it's a little more lyrical in tone, this could be taken from Distinction, Bourdieu's massive "social critique of the judgment of taste." So far it sounds as though Latour is merely recapitulating Bourdieu's own meditations on habitus: we incorporate into our bodies something from outside, which then gets reproduced as seemingly spontaneous and context-specific action (judging a film, taking a political stance, loving a person, et al.). But where Bourdieu always insists on the habitus as a "structuring structure" — that is, not so much habits or skills as a way of seeing the world ("principles of vision and division," is the way he frequently puts it) — Latour characteristically wants to keep each habit or competence discrete. That word "downloading" should tip us off that something different is going on here: and in fact Latour's (admittedly, somewhat sketchy) rethinking of habitus is built around the internet-era idea of "plug-ins," "borrowing this marvelous metaphor from our new life on the Web":

When you reach some site in cyberspace, it often happens that you see nothing on the screen. But then a friendly warning suggests that you "might not have the right plug-ins" and that you should "download" a bit of software which, once installed on your system, will allow you to activate what you were unable to see before. What is so telling about the metaphor of the plug-in is that competence doesn't come in bulk any longer but literally in bits and bytes … Being a fully competent actor now comes in discreet [sic] pellets or, to borrow from cyberspace, patches and applets, whose precise origin can be "Googled" before they are downloaded and saved one by one. (207)

Rather than just somehow absorbing a principle of vision and division and then unconsciously letting it dictate the way we respond to any given situation (the way Bourdieu argues that class habitus, preeminently, does), Latour proposes that we grab very specific competences that will only come in handy in a very few situations, and then either use them or not. Also, where Bourdieu would insist that a habitus, once acquired, can only be altered or replaced by stringent conditioning, Latour rather suggests a "use it or lose it" proposition: we have to "ceaselessly download" competences or we forget them. In sum, Latour has a more volitional and specific idea of how habitus works: it's not a mysterious creature that gets inside you and bursts out of you just when you think you're all alone (in An Invitation to Reflexive Sociology Bourdieu compares it to the creature in Alien), but a guest you willingly invite inside (more vampire than alien?), or, better, a party favor you take home, but are always at the risk of misplacing. The difference from Bourdieu gets underlined here: "If you take each of the rubrics as the mere 'expression' of some dark social force, then their efficacy disappears. But if you remember that there is nothing beyond and beneath, that there is no rear-world of the social, then is it not fair to say that they make up a part of your own cherished intimacy?" (209)

So, to finish this up in the spirit of Latour — and very much not in the spirit of Bourdieu — with a personal reminiscence. I spent basically a year reading through Bourdieu's work (and, as you've reported elsewhere, it really takes about a year of serious reading to get fully comfortable with a theorist). I gained a tremendous amount from it: it gave me an amazing sense of the field of theory, as well as providing numerous stylistic and conceptual tools I could put to work, and it even affected my political and moral opinions. In a real sense, then, it changed the way I think. On Bourdieu's view, I significantly altered my habitus by reading him: I made a series of epistemological breaks with my old, pre-Bourdieuvian self and ended up with new "principles of vision and division" that governed how I saw and thought about everything.

The trouble was, it also made me deeply unhappy in a lot of ways. (I can't lay all of this at Bourdieu's doorstep: last year was a tough one for a number of reasons, but certainly reading him was a major part of it, and the only one that's relevant here.) I felt, as I think many graduate students feel when immersing themselves in a single theorist's work, taken over or subjected t0 Bourdieu's thought more than I wanted to be. The influence was too powerful, too total. And beginning to read Latour seriously has been part of a larger project of attempting to extricate myself from Bourdieu a little, or eliminate some of what Bourdieu had managed to incorporate into me (along the lines of the logic of the pharmakon: the cure for too much social theory is… more social theory!). But I also haven't wanted to lose what I gained from reading Bourdieu and being, in some sense, "taken over" by him. On Latour's view — and this is the great thing about it, I think — I don't have to make yet another epistemological break and throw over Bourdieu for Latour, reflexive critical sociology for ANT (however much the rhetorical antagonism toward "the sociology of the social" throughout RTS might make that seem to be the case). In fact, on his view, I never acquired a Bourdieuvian habitus in the first place: I just subscribed to so many Bourdieuvian "plug-ins" that I was tricked into thinking that he really had taken me over and changed the whole way I see the world. So what I can do now, if I want, is download some new plug-ins from Latour, which may or may not be compatible with those previously acquired from Bourdieu — and if they're not and I find Latour's more useful, I can delete the old plug-ins. Not only this, but if I don't keep "ceaselessly downloading" Bourdieu's plug-ins — by rereading him, and talking about him, and putting his ideas to work — then they will stop acting on and in me whether I voluntarily delete them or not. And this seems true: what's surprising to me now is how much cognitive work I have to do to access Bourdieuvian concepts that at one time I could barely manage to think outside of. Depending on how you look at it, then, this whole question of "theoretical influence" has become either much less terrifying or much more daunting: because you don't have to worry so much about being "taken over," but you do need to do a lot more maintenance work.

OK, so that's habitus. I'll move from here into a post on structure and agency, and then finish by considering Bourdieu's critique of Latour.

Curved minds

Graham Harman says, in response to my last post: "I’m never convinced by the notion of Latour as a devious strategizer or rhetorician." Fair enough: it's a bit too Sokalian, and makes him sound too much like a sophist (in the colloquial, derogatory sense). But I happened to come across a passage just now in Latour's article "What If We Talked Politics A Little?" (2003) that obliquely addresses this very issue (I love it when that happens):

We have to be careful here so as not to draw the hasty conclusion that it is enough to be devious in order to utter political talk accurately. Unquestionably, politics is imposture; we are well aware that the virtue of autonomy can be secured only at the price of a fundamental vice, betrayal, both there and back; we acknowledged that lying — as opposed to the supposedly easy truth of faithful transfer of information — is an integral part of the work of composition; we know that expecting a spokesperson to "tell the truth," to be "authentic," amounts to killing the process of transubstantiation. However, this does not, for all that, mean that to be a good politician it is enough to lie, to be a phoney. That would be too easy. The Prince of Twisted Words would simply have replaced the White Knight of Transparency … One can walk skew, think curved, cut across, be sly, without necessarily drawing the political circle. It is not because they all differ equally from the straight line that all acts of envelopment are similar. "Curved minds" are clearly distinct from one another, even if they are all an object of ridicule for "straight minds." (153)

So perhaps a better way of putting it would be to say that Latour is devious, but always in the service of enveloping something, of drawing a political circle, of catching something in his net. And in this he's not far off from the devious politicians who, as he sees it, are always so despised.

Phenomenology of Reading Latour

As per your suggestion a little while ago, here are some scattered notes on what it's like (for me) to read Latour.

I'll start by jumping off from something you wrote to me (in an e-mail): "he actually is quite a great article-writer... usually I find that you can oppose article-writers to book-writers, but he's great at both!" I think this is true, although I've got to say that so far I give a slight edge to his books: Latour is such an ambitious theorist, always wanting to lay down the maximum bet and shake things up as much as possible (sometimes at the risk of caricaturing others' positions, as you point out in your most recent post), that I think he does better with a little room to stretch out and indulge his theatrics, rather than being forced, for reasons of space, to keep one line of argument going from beginning to end. Another way of putting it would be to say that, like Norbert in Aramis, Latour's method is to make as much of a mess as possible, and then see what's required in order to clean it up — an approach that seems crazy until you realize how closely it reflects the socio-logic of how actors actually make sense of the world. So I guess what I'm saying is that in the articles he sometimes doesn't have enough time to make a mess and clean it up.

You also mention that reading Latour gets you "talking more in terms of strategy rather than about ideas": "Something about how you get a little more frank and unashamed while reading this stuff has to be communicated to people..." Again, I totally agree, and this is something I get from Bourdieu as well: the sense of non-stop devious strategizing (which in Bourdieu, of course, is also thematized) is, once you get used to it, actually very useful as a model, a manner you can imitate, at least to a certain extent, to make your own arguments (which obviously don't need to be the same kinds as the ones Bourdieu or Latour make).

But Latour is such a rhetorical writer that it can even be slightly off-putting, not to mention confusing, because he's also got such a rhetorical view of life. (In this, he's like Nietzsche, who keeps coming up, which perhaps shouldn't be so surprising: as I recall reading somewhere — maybe Harman's book? — he's one of Latour's biggest unacknowledged influences.) In other words, Latour is always trying to convince you that everything in the world is trying to convince everything else: as if reality is in fact made up of innumerable little Bruno Latours. He pays nearly everything the compliment of assuming it's as clever as he is. (He does make some exceptions, usually for other sociologists.) I realize this sense on my part contradicts some of what you said towards the end of your last post about BL's positive valuation of a certain kind of "stupidity," or "forgetting," that goes along with being on the inside of a network. If we're on the inside, and we can just forget what we're doing and sort of go limp, then there's no real need to strategize and rhetoricize all the time. We don't really need "the logic of practice" to explain how anything happens. And there does seem to be a general shift in Latour's work away from a Bourdieusian assumption that everything is infinitely clever (which we find in Pasteurization of France and Science in Action) towards an assumption that everything is kind of stupidly enthusiastic.

But this is rather vague, so I'll finish with a specific passage that I think illustrates many of the hallmarks of Latour's style pretty well. It's from "Where are the Missing Masses?" (1992), which I believe we both read a couple of weeks ago:

The most interesting (and saddest) lesson of the note posted on the door at La Villette is that people are not circumspect, disciplined, and watchful, especially not French drivers doing 180 kilometers an hour on a freeway a rainy Sunday morning when the speed limit is 130 (I inscribe the legal limit in this article because this is about the only place where you could see it printed in black and white; no one else seems to bother, except the mourning families). Well, that is exactly the point of the note: "The groom is on strike, for God's sake, keep the door closed." In our societies there are two systems of appeal: nonhuman and superhuman — that is, machines and gods. This note indicates how desperate its anonymous frozen authors were (I have never been able to trace and honor them as they deserved). They first relied on the inner morality and common sense of humans; this failed, the door was always left open. Then they appealed to what we technologists consider the supreme court of appeal, that is, to a nonhuman who regularly and conveniently does the job in place of unfaithful humans; to our shame, we must confess that it also failed after a while, the door was again left open. How poignant their line of thought! They moved up and backward to the oldest and firmest court of appeal there is, there was, and ever will be. If human and nonhuman have failed, certainly God will not deceive them.
(166-167)

What's going on in this passage? Let's taxonomize:

1. Humor. This is probably one of the first things that strikes readers of Latour: he's very funny, and not afraid to make jokes about pretty much anything, at any point of his argument. (Graham Harman says somewhere that he initially got interested in Latour because he was the only funny person in Continental philosophy.) Here the jokes are at the expense of French drivers (they don't obey the speed limit), the authors of the note ("How poignant their line of thought!" — treating them as if they themselves weren't making a joke), and Latour himself as a scholar (he regrets being unable to properly cite the authors of the note). And the improbably specific details ("French drivers doing 180 kilometers an hour on a freeway a rainy Sunday morning when the speed limit is 130," and then, a little later in the sentence, "the mourning families") are also funny.

2. Symmetry. These kinds of parallel constructions are so characteristic of Latour as to be almost an obsession: "In our societies there are two systems of appeal: nonhuman and superhuman…"; "They first relied on the inner morality and common sense of humans… Then they appealed to what we technologists consider the supreme court of appeal…" One of the odd things about Latour is that, for a thinker so committed to asymmetries and hybrids and quasi-objects, he divvies up and juxtaposes and dichotomizes like Derrida never gave dualisms a bad name. Often his symmetrical formulations are ironic, i.e. he's caricaturing what some other person or group thinks (the moderns, the epistemologists, the technologists, et al.), but he's still clearly addicted to them. And I should say, I don't think this is a problem; that is, I don't think he's really a dualistic thinker, or at least not at all a rigid or limited one. In Latour these pairs ("epistemological couples," in Bachelard's parlance) are always getting pulled apart and recombined and mapped on to others (and not simply reversed, as in dialectic). But stop him at nearly any point and he's always working with a clear-cut distinction.

3. Sacred/theological language. This is another persistent verbal habit of Latour's: it pops up as early as Irreductions and plays a big role in the rhetorical vocabularies of both We Have Never Been Modern and Aramis. One way to deal with this would be simply to read it under the heading of (1), that is, as a joke — and sometimes Latour does use sacred language as a joke, or for merely hyperbolic or blasphemous effect. But knowing that he's a practicing Catholic puts an interesting spin on his continual recourse to a religious idiom. I don't quite know what to say about it here, but I'd just like to add a little more to the passage I quoted above:

I am ashamed to say that when I crossed the hallway this February day, the door was open. Do not accuse God, though, because the note did not make a direct appeal; God is not accessible without mediators — the anonymous authors knew their catechisms well — so instead of asking for a direct miracle (God holding the door firmly closed or doing so through the mediation of an angel, as has happened on several occasions, for instance when St. Peter was delivered from his prison) they appealed to the respect for God in human hearts. This was their mistake. In our secular times, this is no longer enough. (167)

Who is Latour reassuring here? On one level he seems to still be joking: who would possibly think to "accuse God" of leaving a door open? And again, he's hyperbolically specific: that "as has happened on several occasions, for instance when St. Peter was delivered from his prison" is obviously an unnecessary flourish, and a bit of a goof on Latour's own multidisciplinary scholarly authority (oh, you're a theologian now?). But even without knowing anything about Latour's own beliefs, the last three sentences are charged with a sort of melancholy resignation: "instead of asking for a direct miracle … they appealed to the respect for God in human hearts. This was their mistake. In our secular times, this is no longer enough." Of course, the whole point of Latour's article is that much of what we think of as autonomous human morality is in fact preordained by clever engineers, who make it practically impossible not to wear a safety belt, or exceed the speed limit on a residential street, or leave a door open. As you'd expect from him, he seems for the most part to see this as a good thing, as well as a technological marvel, and he would like sociologists, technologists, and moralists all to see that what they talk about is so densely intertwined. But that final caveat — "In our secular times, this is no longer enough" — suggests a somewhat darker (almost Heideggerian?) picture of the "missing masses," a sort of nostalgia for a (probably imaginary) past in which humans really did make moral decisions, in which autonomy was real and society did not make so many of the crucial choices for human beings ahead of time.

Of course, I could be way over-reading (an occupational hazard), and Latour's "secular times" is just total sarcasm: he does cast doubt on the very idea of secularization in We Have Never Been Modern and numerous other places. But it's the sort of troubling ambiguity that he opens up, again and again, through his use of sacred language, and also, for different reasons and to a different degree, through his use of the language of politics.

But I'll stop here, having made a mess and not cleaned it up. Suffice it to say that the more one thinks about Latour's style, the more fascinating a rhetorical case study it becomes. And I haven't gotten to his own writings on style (of matters of concern and so on). I'm sure this is a subject we'll both have a lot more to say about as we go on.

Everybody's Autonomy

I've now delayed my post on Aramis for so long that I'm beginning to feel like the RATP. (Rim shot!)

Anyway, I doubt that I can do justice to such a strange and provocative book in one entry, so I will just offer some scattered comments, and we can go on from there. First, I want to do a little ad hoc taxonomy of its structure, which Latour actually makes pretty easy: certainly the differentiation between sections and subsections is much clearer than any of the divisions within the Aramis project itself. To wit, the elements of Aramis (the text) are:

1. the tale of Norbert (a sociologist and obvious stand-in for Latour, possibly named for Norbert Wiener) and the narrator (a young engineer doing an internship);
2. interviews with engineers and administrators involved in various stages of the Aramis project;
3. documents pertaining to the case;
4. speculative interludes on the nature of technological projects, many of them explicating or applying concepts that crop up in other ANT literature;
5. prosopopoeiae written in the voice of Aramis, the unfinished train line, itself (and occasionally other entities as well)

To make my life a little simpler I'm not going to deal with (4) or (5) at all right now, even though those are some of the richer passages in the text, and may be some of the most interesting to return to once we get into Harman and speculative realism. Please feel free to adduce some of those passages in your response: it's not at all that I'm not interested in them, it's just that I feel they are marked off as properly separate from the "story" of Aramis, which is what I want to start with.

Towards the end of the book, Norbert says:

I'd actually like to do a book in which there's no metalanguage, no master language, where you wouldn't know which is the strongest, the sociological theory or the documents or the interviews or the literature or the fiction, where all these genres and regimes would be at the same level, each one interpreting the others without anybody being able to say which is judging what. (298)

Is Aramis that book? I'm not sure that it is, completely. Certainly the narrative passages aspire to that state, and Latour brilliantly realizes it by making extended use of the dialogue form in the conversations between Norbert and the unnamed engineer. (I know Latour doesn't like this word, but the look is truly, and literally, "dialectical.") And as Norbert never tires of pointing out, the explanations offered by the various interviewees are very nearly as sophisticated and convincing as any of those advanced by the sociologists, making the text as a whole a true contest between "genres and regimes … each one interpreting the others." (Of course this is what Actor Network Theory would claim is always going on; it's just that usually Latour doesn't sacrifice the "friendly tour guide" persona of his first person voice. This movement into dialogism makes for a much more disturbing, chaotic text.) But in the speculative passages (4), he pushes back against these tendencies, marshaling all of the conceptual forces of ANT against the actors — which is why these sections, though full of great insights and formulations in themselves, ultimately don't seem so convincing as a "refined sociology" of Aramis.

But perhaps that's the point. In one way, Aramis is yet another handbook, yet another methodological tract, like Science in Action or Reassembling the Social. Yet instead of showing how perfectly ANT (or science studies or refined sociology or whatever Latour wants to call his method) accounts for the phenomena it's supposed to account for, Aramis takes the somewhat perverse tack of showing how confused the sociologist can get by his objects. Ultimately this is the moral of the book: one must love research, which entails loving its cul de sacs, breakdowns, and unexpected branchings as much as its eureka moments and positive successes. This is the error the actors make, that Norbert chides them for in his big Hercule Poirot speech in the epilogue:
“Oh, you do love science! … But you still don’t love research. Its uncertainties, its whirlwinds, its mixed character, its setbacks, its negotiations, its compromises — you turn all that over to politicians, journalists, union leaders, sociologists, writers, and literary critics: to me and people like me. Research, for you, is the tub of the Danaides: it’s discussion leading nowhere, it’s a dancer in a tutu, it’s democracy. But technological research is the exact opposite of science, the exact opposite of technology.” (291)

A great moment, one that actually moved me a little (I have a thing about research). And one that connects in a pleasing way to the two other themes in Aramis that I wanted to mention: utopia and autonomy.

One interesting motif in the book is the frequent allusion to 60s utopianism, often by the interviewees: e.g., M. Liévin: “‘It was fashionable at the time — Personal Rapid Transit, PRT. Everyone was excited about it … around the sixties. The Kennedy era. Private cars were on the way out — that’s what everyone was saying’” (15); or, later, M. Henne: “‘Before 1975, there was a period of innovation — new cities, all sorts of wild gimmicks. After 1975, it was all over; security was the only thing that counted…’” (47); or M. Cohen: “‘It’s also a question of the times, you know. I have trouble imagining an industrialist today who’d say, ‘We don’t have a medium-distance ultrasound link-up? Okay, let’s go for it — we’ll invent one. There’s no motor on the market? Never mind, we’ll develop one.’ And it was all like that. Today, everybody sticks to his own job. People don’t take so many risks’” (54); or M. Girard: “‘I went back upstream, as it were, back to the somewhat utopian thinking of the 1960s … we need something like cars that join together, trains that split apart’” (136); or M. Coquelot: “‘It’s a project from the culture of the Sixties, transporting people in a private Escape instead of in public conveyances… Now, obviously, we’re culturally out of phase…’” (155). In Latour's multi-layered allegory, one thing Aramis seems to stand for is the intellectual confidence, even arrogance, of France in the 1960s — a theme that is echoed, in another key, by Norbert's frequent references to figures like Sartre, Foucault, and Lévi-Strauss. What's so interesting about this nostalgic theme, however, is that it shows how 60s utopianism — which one often associates with an anti-technological stance, like the one Norbert attributes to Habermas on page 280 — in fact pervaded the sphere of technology and engineering itself.

This qualifies in an interesting way what some people have seen as Latour's technocratism, his own "love of technology," which is suspect to many intellectuals (chiefly Heideggerians, but not only them). Because what Latour loves above all is the technology that could exist, and the social interactions that (can) help to bring it about. He's not an apologist for the neoliberal order or the effects that technological culture has had, but he does insist on the reality of the networks that science and technology have created, and the possibility of using them to different ends than the ones they are currently used for. From this standpoint, he looks like quite a utopian thinker himself. And while Norbert, like Latour, is reluctant to attribute the failure of Aramis to leviathan-sized macro-actors ("Are you going to accuse the social system? Capitalism? Napoleonic France? Sinful man, while you’re at it?," 197), there is more than a tinge of pathos in the fact that an innovation that would have helped solve ecological as well as transportational problems was scuttled by technocratic management. And while many of the explanations given by the actors are as good as any a sociologist could come up with, some are a lot less enlightened — e.g. M. Chalvan: “‘Cars belong to individuals; everyone looks out for them. But Aramis would have been collective property. The first time anything went wrong, people would have blown the whole thing up’” (71): a rare appearance of straight-up ideology in Latour’s work. [Cf. also "The Fear of Mob Rule" in "Do You Believe In Reality?"]
Which brings me to the second question, the question of autonomy. This is not a concept that I would normally associate with Latour, given that his network ontology seems to make the question of autonomy not even really worth asking: if everything — human and nonhuman — is constantly negotiating with everything else, as Irreductions would have us think, then in what sense could autonomy even matter? If human minds aren't privileged, if Kant was wrong, then autonomy is pretty much a non-issue, right? But Aramis proves to be (among many other things, of course) a sustained meditation on autonomy, although it gets in through the weird allegorical back door (or "hidden staircase," as Norbert likes to say) of the operation of the Aramis cars themselves.

From quite early on, we are encouraged to think of the Aramis system as akin to the networks that Latour, in his other work, sees everywhere: M. Soulas, president of the RATP, says: “‘It wasn’t a line like a subway, but more like a bloodstream: it was supposed to irrigate, like veins and arteries. Obviously the idea doesn’t make sense if the system becomes a linear circuit — that is, if it ceases to be a network’” (9). But one of the big questions becomes, how does one keep the elements circulating within this network from destroying one another other, or the passengers they carry? The enormous technological difficulties of programming the cars to be autonomous and ensuring that the system is fail-safe is one of the major challenges faced by the Aramis team: “‘The big challenge with Aramis is that the cars are autonomous; they don’t touch each other, yet they work together as if they were part of a train. They have nonmaterial couplings — nothing but calculations. So you can imagine how autonomous they are. Every car has to know who it is…’” (54) Everyone has to know who they are, and where others are; they don't touch each other, yet they work together as if they were part of a greater whole. This is the social, as Latour sees it, and autonomy turns out to be a more important concept for it than I had previously thought. In one of the book's wilder passages, he transposes this concern for autonomy into a theological register in another odd prosopopeia, a dialogue between a Catholic priest and someone named "Lamoureux" who compares the Aramis cars to Leibnizian monads: “…they’ll be connected by a vinculum substantiale. Nothing material will link them together to keep them on the right path. They’ll have to make independent decisions, check themselves, connect and disconnect, in conformity with the laws of the world system to be sure, but freely, without touching each other and without being the slaves of any automated mechanisms…’” (63-64)

But it isn't Latour alone imposing this anthropomorphic autonomy on Aramis: the actors themselves do it constantly as well. The social scientists who conduct a study of the project's public image conclude that “Aramis has to be perceived as the prelude to a new philosophy of transportation, addressed to responsible adults” (186). [One might argue that it's the autonomy of the riders that's in question here, but as in other areas of social life, one has to grant the autonomy of others — in this case, train cars — as well as maintaining and asserting one's own.] The engineers in charge of Phase 3B describe the fixed sectors of traditional guided-transportation systems as being “‘like a cop who takes away all the flexibility from trains and subways in exchange for a considerable margin of security'" and “‘MATRA comes up with a radical solution … It involves doing without the sacrosanct fixed sectors, in exchange for increased intelligence on the part of the cars’” (208, my emphasis). Maybe Latour owes more to la pensée '68 than we thought! The great social struggles of the (long) 60s, the quests for civil liberties and freedom from police surveillance and oppression, are recapitulated in the attempt to imagine a true transportation network, one that requires a radical — and perhaps impossible — degree of intelligence and autonomy on the part of its constituent actors.

I'll leave it here, having barely scratched the surface of this marvelous, hilarious, frustrating, unpredictable book. I welcome your thoughts, whether or not they're a direct response to what I've written here. Let's see if we can link up and act as a train.