Monday, May 23, 2005

(Reading Roundup: Callon on Translation)

Originally posted: Mon, 23 May 2005 22:19:22

Although Bruno Latour is most often associated with actor-network theory, Michel Callon's work predates him. Callon's work with the sociology of translation tends to focus on economics more than Latour's does, but it's the same basic idea: examining how actants are changed so that they can forge alliances, work with one another, and circulate. Translation itself has undergone several conceptual changes since it was first derived from Michel Serres' 1974 work, so I'll focus on differences in accounts within these four articles and compared to other ANT writings.

Callon, M. (1986). Some elements of a sociology of translation: Domestication of the scallops and the fishermen of Saint Brieuc Bay. In Law, J., editor, Power, action and belief: A new sociology of knowledge? , pages 196?233. Routledge, Boston.

This is a famous and often-quoted study of how scientists studied scallops in St. Brieuc Bay, and more generally, "a new approach to power, that of the sociology of translation" (p.196). Translation has

three principles, those of agnosticism (impartiality between actors engaged in controversy), generalised symmetry (the commitment to explain conflicting viewpoints in the same terms) and free association (the abandonment of all a priori distinctions between the natural and the social) (p.196)

In this article, translation has four moments. Notice these, because although this article is frequently used to define translation -- it gets a cite in Latour's Pandora's Hope, for instance -- it's described differently from place to place. In Pandora's Hope, for instance, some of these moments get shifted away from translation and are made equal components in "mediation." (See my recent review.) But here, translation is described thus:

Four ?moments? of translation are discerned in the attempts by these researchers to impose themselves and their definition of the situation on others: (a) problematisation: the researchers sought to become indispensable to other actors in the drama by defining the nature and the problems of the latter and then suggesting that these would be resolved if the actors negotiated the ?obligatory passage point? of the researchers? programme of investigation; (b) interessement: a series of processes by which the researchers sought to lock the other actors into the roles that had been proposed for them in that programme; (c) enrolment: a set of strategies in which the researchers sought to define and interrelate the various roles they had allocated to others; (d) mobilisation: a set of methods used by the researchers to ensure that supposed spokesmen for various relevant collectivities were properly able to represent those collectivities and not betrayed by the latter. (p.196)

Using these four moments, Callon goes on to systematically study the empirical case, in which researchers try to determine ways to conserve and cultivate scallops. In Callon's symmetrical reading, the actants included the researchers, the fishermen (who, based on public demand, were tempted to overharvest the scallops), and of course the scallops themselves. The vocabulary of translation allowed Callon to study these actants in the same terms. Scallops "voted," for instance, in that counting the number of larvaethat anchored themselves to collectors was no different from counting the number of fishermen who voted to take part in the project themselves. In both cases, "A series of intermediaries and equivalences are put into place which lead to the designation of the spokesman" (p.216), and "To speak for others is to first silence those in whose name we speak" (p.216).

Callon describes a temporarily stabilized settlement here, in which all actants are interested and invested in the outcome. If successful, the researchers turn themselves into an obligatory passage point that all actants must pass on the way to their own objectives. "But this consensus and the alliances which it implies can be contested at any moment. Translation becomes treason" (pp.218-219). For instance, scallops initially attach themselves, but in subsequent years, they refuse to do so; fishermen initially restrain themselves from overfishing, but give in to temptation one Christmas eve (p.220).

This essay does a great job of laying out translation in four moments and describing how a sociology of translation works. But translation is itself translated in later iterations.

Callon, M. (1986). The sociology of an actor-network: The case of the electric vehicle. In Callon, M., Law, J., and Rip, A., editors, Mapping the dynamics of science and technology: Sociology of science in the real world, pages 19?34. Macmillan, London.

The same year that Callon contributed the scallops study to John Law's edited collection, he coedited another collection with Law. That collection included this essay on electric vehicles. Here, Callon comes up with some terms that apparently didn't survive in later ANT, such as the "actor-world," which Callon tries to conceptually separate from an actor-network without much success (p.22 et passim). In this essay, Callon says that translation has three components: the translator-spokesman; the obligatory passage point; and displacement (p.24).

At first glance these seem rather different from the four moments of translation described in the scallops article, but on closer examination Callon is talking about the same things. It's just that in the scallops article he focuses on moments (points or processes in the translation), while here he focuses on components (elements or entities that must be involved). The new focus is confusing in that components are used in multiple moments and the moments themselves are cyclic rather than linear, so reading through this explanation, it's easy to detect very different moments than the ones we saw in the scallops article! But careful tracing suggests that these two articles are indeed talking about the same thing. In this case, Callon discusses the development of an electric vehicle, but the principle is the same as with the scallops.

Callon elaborates again on translation: "Translation is a definition of roles, a distribution of roles and the delineation of a scenario. It speaks for others but in its own language. It is an initial definition. But ... no translation can be taken for granted for it does not occur without resistance" (p.26).

So, then, we're getting a coherent picture of translation. And this brings us to the next article.

Callon, M. (1991). Techno-economic networks and irreversibility. In Law, J., editor, A sociology of monsters? Essays on power, technology and domination, pages 132?161. Routledge, London.

Five years later, Callon introduces the concept of techno-economic networks: "a coordinated set of heterogeneous actors which interact more or less successfully to develop, produce, distribute, and diffuse methods for generating goods and services" (p.133). These seem to be essentially actor-networks, but specifically ones that relate to economics. I'll treat them as actor-networks here, recognizing that there may be some differences because of focus.

In any case, what Callon says about TENs applies to actor-networks more broadly. First, actors are drawn into relationships through intermediaries; "an intermediary is anything passing between actors which defines the relationship between them" (p.134). They include inscriptions ("relatively immutable media that resist transport"); technical artefacts; human beings (with "the skills, the knowledge, and the know-how that they incorporate"); and money (p.135). These intermediaries "describe their networks in the literary sense of the term. And they compose them by giving them form" (p.135).

A few notes on these. Callon addresses apprenticeship briefly here -- an unusual move in the ANT literature, which tends not to concern itself with learning, development, education, or competence in detailed terms. In apprenticeship,

the instructor describes the operation of an object: the network 'inscribed' on it is set out and inspected. What are the links between technical objects? And what are the roles that humans play? ... In this way the machine is interpreted, deconstructed, and inserted back into its context -- though possibly not the way intended by the designer. The written traces of such efforts to put objects into words are to be found everywhere, as are the controversies to which they lead. Codes, checklists, maintenance manuals and user handbooks, all of these escort objects on their travels. (p.137)

Furthermore, skills can be regarded as networks. A description of skills involves reconstituting the network in which they are "expressed and put to work" (p.138). Again, this is remarkable in that Callon is actually discussing human competence and learning. He goes on to discuss what activity theorists would call mediation: Consider Mr. Smith, who wants to go on vacation. He uses a travel package provided by Club Med, "a mixture of humans and non-humans, texts, and financial products that have been put together in a precisely co-ordinated sequence" (p.139). These actants include "computers, alloys, jet engines, research departments, market studies, advertisements, welcoming hostesses, natives who have suppressed their desire for independence and learned to smile as they carry luggage, bank loans and currency exchanges" (p.139). This TEN might be complex, "but in principle it works just like any other intermediary. If Mr. Martin uses a fork to mash potatoes this is just another (albeit simpler) intermediary" (p.139). Notice that unlike activity theory's version of mediation, Callon's intermediaries can be people as well as artifacts.

An actor is just another intermediary in this reckoning (p.140). Just as the hostess or smiling native serves as an intermediary for Mr. Smith, he serves as an intermediary for them. "Actors define one another by means of the intermediaries which they put into circulation," Callon says in the italics that he liberally uses to emphasize his key points, and "the social can be read in the inscriptions that mark the intermediaries" (p.140). Is a given thing an actor or intermediary? It depends on "where the buck stops" -- what you're trying to study -- because all actors are intermediaries (p.142).

Here, Callon begins to talk about networks in more detail, drawing on Deleuze to make some of his points (p.142). He describes a triangle of translation, again similar to the triangle used to describe mediation in activity theory. The triangle is of the translator, translated, and the medium in which the translation is inscribed (p.143).

Callon then elaborates by discussing convergence and irreversibility. "Convergence measures the extent to which the process of translation and its circulation of intermediaries leads to agreement" (p.144). It involves alignment, or the extent to which translation "generates a shared space, equivalence and commensurability" (p.145). "A translation that is generally accepted tends to shed its history. It becomes self-evident, a matter on which everyone can agree" (p.145). Of course, translation must "rarefy the universe of possible actors by organising imputation and limiting the number of translations that can be easily stabilised" -- that is, you want to limit the extent to which translations can occur so that you can strengthen and build on the network -- so actor-networks generate "conventions" or "codifying regulations" that Callon calls "co-ordination or translation regimes" (p.147).

This leads Callon to argue that one can make a relative judgment about the boundaries of a network based on its level of convergence (p.149). He has a brief methodological discussion here, without much substance, I'm afraid.

Then he gets to irreversibility, which I find to be quite interesting. Irreversibility is

(a) the extent to which it is subsequently impossible to go back to a point where the translation was only one amongst others; and

(b) the extent to which it shapes and determines subsequent translations. (p.150)

"All translations, however apparently secure, are in principle reversible," Callon declares (p.150). No matter how widely it is accepted that the Earth is flat, or that spontaneous generation occurs, or that mankind is only 6,000 years old -- and no matter how many other things are built on these beliefs -- they are reversible. And so are the things that replace them. Facts can be walked back. But some facts are harder to walk back than others: the longer a network is, the stronger it is, because the more depends on its translations being stable (p.151). Irreversible translation is normalization (p.152).

Callon briefly discusses apprenticeship in this context:

Apprenticeship is a case in point. In this the elements involved in a translation become dependent on one another in a process of mutual adaptation. A skilled machinist cannot work without his machine. The development of a technology depends on engineers with a specialist training. The practice of this trade puts specific objects into circulation. And so on. In this way decisions become more and more dependent on past translations. (p.151).

Callon, M., Larédo, P., and Mustar, P. (1997). Techno-economic networks and the analysis of structural effects. In The strategic management of research and technology, pages 385?429. Economica International, Paris.

This final chapter postdates the others by quite a bit, and comes before his chapter in Complexities. Callon and his colleagues go into technico-economic networks in much more detail than I find useful here. But let's get the 50,000 foot version. Again, Callon discusses TENs as actor-networks particularly within the context of economics (that's my take, not a specific quote of his). Here, Callon is particularly interested in innovations stabilizing within the network (pp.386-387). Again, he discusses actors as circulating intermediaries (p.389). He then goes into an elaborate taxonomy of TENs based on several axes: networks can be lacunary or chained; dispersed or convergent; short or long; polarized or non-polarized. These axes all cover different aspects of a network, but Callon's objective in detailing them, again, seems to be in terms of how stable and powerful they are. The axes provide a more detailed analysis -- certainly more detailed than I particularly wanted to get into. I'm not sure how helpful they are.

In sum, these four articles represent a line of thought parallel to Latour's, and they certainly cite Latour a lot. But the notion of translation is a little undercooked in places, and the term seems to disappear in the last article. On the other hand, the principles of translation seem to be intact in here as well as in Latour's work; the basic idea is the same, though it gets chopped up in different ways.