Ether Wave Propaganda is pleased to present as a special serialized feature this Q&A session with Cardiff sociologists Harry Collins and Rob Evans regarding their Sociology of Expertise and Experience (SEE) program. As a special feature, it will not adhere to the usual length restrictions we try to keep on posts, and, therefore, will run unusually long. Also note that this post does not represent a spontaneous exchange. Collins and Evans have asked for clarifications on the original questions, and have carefully crafted joint responses. They have also asked me to ask them to modify their responses if I deemed them inadequate or wrong, so as to make the Q&A as useful as possible. I saw no need to do so for the first question.
Will Thomas: SEE is an attempt to move beyond the sociology of knowledge into a sociology of expertise, and is explicitly formulated as a “third wave” in the sociology of science. What is it that most distinguishes SEE from other attempts to move beyond the initial insights of SSK, such as the Actor-Network Theory, or Pickering’s “mangle”?
Asked to clarify what I had in mind as the proposed novelty in ANT and the Mangle, I replied:
Will Thomas
On Pickering’s mangle and ANT, I see both as ways to respond to the initial problem with SSK, which is that it didn’t really take into account how important the encounter with the non-human world is in science. Of course, I agree that the meaning of those encounters is determined by what can be sociologically agreed upon, but, at the same time, the appeal to experience and experiment is a persuasive force that needs to be acknowledged. So, how to incorporate that persuasive ability into a sociological account of science? ANT, as I understand it, allows agency to be granted to non-human actors, so long as they are given interpreters. The mangle, again as I understand it, refers more to “resistances” that actors experience, and which may (depending on other factors in the mangle) cause them to reconsider their claims.
So, what makes SEE into a whole new “wave” where these programs are not? Basically, if I were to anticipate your answer, I think it allows that experts can be distinguished from contributing experts or non-experts on more than, for lack of a better word, contractual grounds. By framing the problem from a policy viewpoint, you explicitly claim that communities of people with differing but consistent forms of expertise can formulate common goals and can come to a non-compelled agreement about what action is most appropriate. This tactic deflects attention from whether this or that piece of knowledge is “correct” to whether or not you, an expert, can tell me something about my problems that makes me rethink what actions I will take in the future. Say I’m negotiating an oil contract; if you’re a petro-geologist, you can tell me things, and I will have no independent means of ascertaining whether or not they are true—but, if I later conclude that my contract was better because of your advice, my expertise in evaluating contracts will allow me to make an appraisal of your ability to contribute without relying on your credentials.
So, I suppose if the main difficulty with Wave Two is the problem of determining who to listen to, and if the way out of that difficulty is agreed to be rooted in the persuasiveness of experience and experiment, their approach is to confine the persuasive experience to the individual encounter with nature and to build upward from the credentials attributed to the small, privileged community of experiencers (which allow us to treat their claims as established fact); whereas SEE allows other kinds of experts to “feel” that encounter in their own deliberations, actions and policies, even if they do not benefit directly from the initial experience, as in the oil contract example.
I learned that Collins and Evans saw their own approach as more orthogonal to the “encounter with nature” problem, and that their classification of ANT and the Mangle and these programs’ concern with this problem as part of Wave Two is closely related to the “epistemological chicken” debate (alluded to and cited in my intro to this Q&A series). I also realized the importance of “tacit knowledge” to their commitment to drawing a distinction between sociological and epistemological or historical inquiry.
Harry Collins and Rob Evans: As far as we can tell ANT (Pickering’s mangle can be included within it as far as this discussion is concerned), was not an attempt to move beyond the sociology of knowledge (SSK) but to do pretty much the same thing (that is, to describe in great detail, and in a naturalistic way, how science gets done). To the extent that ANT appeared to move beyond the major initial insights of SSK, it was a matter of philosophical adventurousness rather than any change of aim. Collins’ views on the consequences of this move are documented in the ‘Epistemological Chicken’ article he co-authored with Steve Yearley in 1992 and Evans shares a broadly similar view.
Harry Collins
In summary, the ‘Epistemological Chicken’ argument is that the beauty of SSK – its ‘special thing’ – is its determination to ignore the force of experiment and the influence of the natural world on what people believe about nature: SSK was sociology of scientific knowledge and sociology is about the social not the material. The point of SSK, certainly under the Bath School model, was that, as a principle of methodology, it refused to give the last word to anything but the social. The substance and meaning of data or other kinds of evidence was to be explained by reference to the social. The claim is that this is the only way to introduce enough rigour and determination into the study of the impact of the social. Allocating some influence to the social and some to the natural does not produce enough rigour on the social side. Under these circumstances the choice of how to divide up the explanation can be ad hoc or self-serving – whenever social explanation gets hard it is all too easy to invoke the material.
As a result of its methodological intransigence, SSK developed an empirical grounding for established philosophical debates about induction, falsification, and the like, and gave rise to a whole string of new findings that challenged the common sense view of science. For example, SSK showed that in scientific controversies measurements were rarely trusted and that refereed journal papers were often completely ignored. These findings caused real difficulty for the `scientificness’ of science and force us to re-think if and why it is special. In comparison, ANT’s ‘hyper-symmetry’ brought in both the social and the material which resulted in a rather prosaic model of the world in which `people do this’, `nature does that,’ and the outcome is some mixture of the influence of the two rather as astute scientists have always believed. That the resolution of scientific controversy was really a mixture of the social and the natural is exactly the argument that the `science warriors’ threw at SSK in opposition to its determination to examine only social causes. The script of ANT and the like can be found in, for example, in Alan Sokals’ contributions to The One Culture (edited by Labinger and Collins, Chicago, 2001). If they had only understood it properly, Alan Sokal and the rest of `science warriors’ could have refrained from their attacks on ANT – in essence, it already fulfilled their requirements.
Rob Evans
But suppose ANT and Sokal are right and what causes one thing to be believed rather than another is a mixture of the social and the material/natural. In that case, if you want to know about the social aspects of science you should ask sociologists and if you want to know about the natural aspects – door-closers or scallops, to use Latour and Callon’s most well known examples – you should ask engineers and fisherman/biologists. What has hyper-symmetry to do with it? It is just old fashioned science and technology with a bit of the social thrown in. If, on the other hand, ANT’s hyper-symmetry is meant to indicate that one is talking only about humans’ interpretations of door-closers and scallops as resistances or compliances then we are back with SSK.
It is as if ANT offers a fancy way of speaking – for example, `inscription devices’ and `immutable mobiles’ take the place of meters and papers – but it doesn’t really challenge anything. Rather, it takes something away. With an inscription device and an immutable mobile you can sometimes say “the meaning has to be constructed” and sometimes say “the meaning is carried across space because it has been `inscribed’ and is `immutable’.” You can choose to invoke fluidity and interpretation sometimes and stability and portability other times – whatever suits the argument at that moment. With SSK there is none of this choice or confusion – everything is fluid unless it is made non-fluid and it can be made non-fluid only by humans.
Or consider the social analysis of the relationship between humans and machines. Here hyper-symmetry takes attention away from the important questions by dissolving the difference at the outset. And, because under hyper-symmetry there is no special role for humans, it submerges the idea of tacit knowledge – but tacit knowledge is central to the proper analysis of science. It is strange that a theory of science that cannot encompass the idea of tacit knowledge can be so pleasing to social scientists.
For these reasons, the popularity of ANT seems rather puzzling to us. Neither of us can claim much success in grasping how hyper-symmetry has added to the understanding of the world even though we have tried very hard. It is odd and uncomfortable to be in such a small minority so if anyone can provide us with some examples of real differences made to the world by ANT’s invocation of `hyper-symmetry’ — differences which are something more than new ways of speaking, or new political manifestos, and which could not have emerged from SSK or sociology of technology — we would be grateful. In sum: what counter-commonsensical findings emerge from treating non-humans and humans as equivalent – each being simply `actants’? (This is not to say that Latour, Callon, and their colleagues have not made significant contributions to the `second wave’ analysis of science that began with SSK, it is just that these contributions seem unrelated to the hyper-symmetry which is, so far as we can see, the only thing that licences any claim that ANT has `moved beyond the initial insights of SSK.’)
In the case of the Third Wave, on the other hand, there is a clear attempt to go beyond SSK and do something different. The advance claimed by the Third Wave is to take the insights of the Second Wave – that, to a large extent, scientific truth is made by humans – and show how we can still feed science and technology into policy-making. Actually, you do not need ANT, SSK, or anything similar to realise that this is a difficult question because, even for the most adamant realist, the speed of scientific consensus formation is often slower and than the speed of policy-making. In our catch phrase, `the speed of politics is faster than the speed of science.’ What we have done is to say that you can use science and technology even if it is ill-formed (either in the short- or the long-term) by turning the gaze from truth to expertise. Since expertise is a possession of humans, turning, in our analysis, on the possession of tacit knowledge, the development is incompatible with ANT.
The aim is to use the expertise about expertise developed by SSK to classify and assess expertises so that, long before anyone knows what truth will emerge (or come to be seen as having emerged), some judgements can be made about who ought to contribute to technical decision-making. The aim is to challenge the claim that science is politics by other means and to provide non-political criteria, or criteria that are as non-political as you can get, for policy-makers to evaluate expert claims.
Turning the gaze to expertise has already produced a Periodic Table of Expertises plus a number of counter-commonsensical discoveries and insights both philosophical and empirical. For example, the idea of interactional expertise flies in the face of existing theories about the relationship between the body and conceptual life locating the significance of the body at the collective level but not at the individual level. It helps to explain how managers manage and, perhaps, it explains the division of labour as a whole. The idea has also given rise to a new experimental methodology – the imitation game [ed. also see this article on Slate]– which has shown that, say, blind people are indistinguishable from the sighted in their discourse whereas the sighted cannot pretend to blind. The new experimental technique has all manner of applications. And interactional expertise is just one cell in a table with around ten new expertise categories all held together by the notion of tacit knowledge. The programme of research is growing out from this table in many different directions.