Expert opinions

We no longer see scientists as unquestionable founts of wisdom. So where should we seek expertise when we need to make decisions involving science and technology? Matthew Reisz hears the views of two contentious academics. There is a big fashion among social scientists never to claim expertise in anything," declares Harry Collins, "but actually we think we're quite expert when it comes to understanding the nature of knowledge, scientific controversy and expertise - and we're not afraid to say so. What's the point of being an academic if you don't become more expert in something or other?"

When it comes to a specific issue such as the measles, mumps and rubella vaccine, Collins, a professor at the Cardiff School of Social Sciences, is happy to be equally forthright. "The social science community has not come out of the episode very well," he says. "It's been very good at representing the views of the parents but not at thinking about the MMR controversy as a piece of scientific knowledge. There never was a controversy - we know what scientific controversies look like."

"Expertise" is a hot topic within universities. In the preface to Rethinking Expertise , the stimulating new book Collins has written with Robert Evans, the authors make a point of thanking "those colleagues whose attempts to stop our work being published, or even referenced, make novels of academic life seem dull".

It is also a hot topic out in the real world. When matters of science and technology are in dispute, Collins argues, "the standard (academic) view is that who counts as an expert cannot be ascertained until the controversy is over. But a policymaker hasn't got the luxury of waiting. You've got to make a decision, even if it's a mistaken decision, and we have to find a rationale for making those decisions using science and technology - because if we stop using science and technology we've no longer got a civilisation".

When I visited Collins and Evans in Cardiff, they eloquently laid out for me the views on expertise that have proved so contentious. But they also wanted me to try a little test. And that is how I found myself sitting at a computer in Collins's office typing in the words: "What's your favourite chat-up line?"

Almost instantly I received two responses.

"I don't have one and can't remember one being used on me."

"Weren't we at school together? I think I remember playing football with your brother."

I knew that the answers came from volunteers in the next room, one of them a man, the other a woman pretending to be a man - and it was my job to tell them apart. After six questions, virtually certain that Person 2 was the man, I confidently guessed wrong.

This may, of course, say something about me. But it also reflects the fact that most women constantly interact with men, learn to speak their "language" and so can often give a convincing imitation. If Person 2 had been a nun, might she have been less successful at fooling me?

Such "imitation game" experiments are based on an old parlour game famously adopted by Alan Turing for what came to be known as the Turing Test, which challenged engineers to create a computer that could "pass" as human to a person firing questions at it. Collins suspects that the technique "will become a standard part of undergraduate courses in sociology and psychology". It can certainly be used to explore many issues of prejudice and socialisation, but for the moment its main purpose is to illuminate and test for something he and Evans call "interactional expertise" - roughly the kind of expertise one acquires by hanging out with a particular group of people.

A Danish colleague, for example, is exploring whether childless midwives can answer questions about childbirth in a way that convinces women who have actually pushed a baby out into the world. The colour-blind seem better at "passing" for the normally sighted, among whom they spend their lives, than the other way round. Might the same thing apply to black people who have grown up "in a white world" and been forced to learn its codes?

All this forms only one element of Rethinking Expertise . The book offers a rich and detailed "periodic table" of expertise, ranging from the kind of beer-mat knowledge useful only in pub quizzes to the levels of skill that enable people to make a contribution to cutting-edge science. It considers wine buffs and art connoisseurs, hoaxers, journalists and pseudoscientists. It looks at deep philosophical issues of "embodiment" - whether you need to move around in the world to acquire a language or the jargon of a specialist field - that have major implications for the study of artificial intelligence and computer learning. It is full of case studies, anecdotes and intriguing experiments. But at its heart are questions arising directly out of the authors' work in the sociology of science and the challenges of scientifically literate public decision- making.

Collins has spent years among gravitational-wave physicists on a study funded by the Human Sciences Research Council exploring the politics and dynamics of such research, the controversies that have greeted claims that gravitational waves have been "spotted", and "how the process of social negotiation eventually emerges into a consensus". The results of his ongoing work have been published in the huge first volume of Gravity's Shadow (2004).

It also led him to reflect on his own form of expertise. "I'm sitting around at lunchtime and over coffee with the physicists who are now my friends," he tells me, "chatting about physics, putting a devil's advocate argument and occasionally making a suggestion - which, on one extraordinary occasion, was actually taken up. Yet I can't do gravitational-wave physics. I can talk it, but I can't do it." In the terminology of the book, he is an interactional but not a contributory expert.

His ability to "talk the talk" means that Collins does pretty well in the imitation game when a real gravitational-wave physicist asks questions about the subject. His responses, in other words, often allow him to "pass" as the real thing. What is interesting is that other scientists - not just microbiologists but even the astrophysicists in the lab next door - tend to be pretty hopeless at this task. "Real, deep expertise," Collins concludes, "takes places in narrow crevasses ... One of the policy implications is that you can no longer trust the generic scientists in their white coats when they are talking outside their specialisms."

Along with this dethroning of "the men in the white coats" has come a recognition of those whom Evans (Collins's colleague and former PhD student who has done sociological fieldwork among economic forecasters and on a sustainable energy project) called the "specialist experts who just happen not to have certificates".

It was people in this category who emerged, for example, when doctors wanted to test a new Aids drug, AZT, in 1985. Initial plans were for traditional double-blind trials. But an activist group called Act Up objected: those put on placebos would be sick or dying in the meantime, so they were bound to try to get hold of alternative cures, thereby messing up the data.

At first the scientists ignored the members of Act Up, who had no medical training and "whose dress codes and presentation-of-self were as far from the world of medical orthodoxy as it was possible to be". But the activists soon acquired adequate levels of knowledge of microbiology and statistics, along with "experience of how Aids sufferers would actually respond to the demands placed on them", so they were co-opted by researchers, "not least because it enabled them to do better science". Expanding the range of expertise involved in the drugs-testing programme could only help everybody.

Yet it is here that we hit a problem. We have punctured the image of the scientist as an unquestionable fount of wisdom. And we have come to accept that those with relevant but non-certified expertise have a role to play even in specialist areas. This, agree Collins and Evans, is all to be welcomed. But they also want to retain the principles that "there are people with no expertise" and that "all else being equal, we ought to prefer the judgments of 'those who know what they are talking about'". These statements sound like the most banal common sense, but they are worried by the widespread currency of the ideas that everybody's view is worth hearing - and even that "ordinary people are wiser than experts in some technical areas".

They cite, for example, an Economic and Social Research Council publication called The Politics of GM Food: Risk, Science and Public Trust (1999). This document went out of its way to praise "ordinary people", who "demonstrate a thorough grasp of issues such as uncertainty: if anything, the public are ahead of many scientists and policy advisers in their instinctive feeling for a need to act in a precautionary way". It sounds striking, but turned out merely to mean that "there was a feeling that scientific and technical interventions moved food even further from its (desirable) natural state".

There could be few more simplistic notions than that "natural" simply equals "good", argue Collins and Evans, yet it is "quoted in this report as supporting the view that the public exhibited a 'reasonable' assessment of the evidence". Romantic and anti-intellectual notions of folk wisdom are surprisingly common in debate about science and technology, they claim, even among academics.

Collins himself long had a reputation as a young Turk and a radical relativist, and he still believes that "research in the sociology of scientific knowledge for the past 30 years has done some absolutely terrific work in showing that science doesn't produce the clear exact pieces of knowledge that everybody thought it did, and how tangled with politics it is". Today, however, he is keen to stress that this is only half the story. "It is incorrect to conclude that science is just a branch of politics. Having shown how much like ordinary life science is, we now have to embark on the somewhat more difficult task of showing the ways in which it is not quite the same. There is something special about it."

Much of the flak their work has received, Evans believes, comes from their "moving on from relativism to making evaluations about the truth of scientific knowledge, which we'd all trained ourselves not to do. For the policymaker, what we've been saying for the past 30 years isn't much good, because we've got to make evaluations".

To give a concrete example, Rethinking Expertise cites the issue of redundant oil rigs. One option is to sink them in the sea, which some oppose as "showing a lack of respect for the environment". Although this should be decided by the normal democratic process, it is not merely a political or lifestyle choice - there is also a factual question about whether it would cause pollution or rather, perhaps, "provide a safe haven for endangered fish".

If we don't ask this question, we are liable to confuse making a green gesture with doing something effective for the environment. And who but a scientific expert can provide a useful answer? Collins and Evans have produced a bold attempt to untangle some of the complex challenges. Now all I have to do is try and sort out the differences between men and women.

Harry Collins and Robert Evans's Rethinking Expertise is published by Chicago University Press, £16.00.