July 12, 2013, by Brigitte Nerlich
Science, politics and certainty
I have recently begun to think about what it might mean to give scientific advice (or science-based advice or evidence-based advice or, even, expert advice), a topic that I have not thought about before. I know that there is a lot of literature on that topic and that many people have written about it, but I have to confess that, as I am on a hill/beach somewhere (see photo) in deepest Dorset, I have not read that literature. So, what I am going to say here is very provisional and tenuous.
The topic I am exploring is also linked to other complex issues which would need more thought, such as science, politics, experts, advisors, expertise, evidence, decision making and so on, some of which will be explored by experts, such as Harry Collins, at the ‘Science in Public’ conference that will take place on 22-23 July in Nottingham.
I will talk about ‘scientists’, but this is, of course, a very heterogeneous category (and the same goes for scientific advisors, policy makers, experts, etc). They work in different disciplines, across different disciplines; they work on topics which have policy relevance or not; they work in different political, cultural and institutional contexts (and these contexts are rapidly changing under pressure from the impact agenda). They have different personalities, some being attention seekers, some recluses, some being team players, some not. Some like dealing with politicians, some (and I guess most) don’t. Sometimes they agree on something, sometimes they don’t, and so on. I assume, however, that most scientists share a desire to find things out about the world which haven’t been found out before. Most would, I assume, never say that what they want to find or have found out is ‘the absolute truth’, as that would inevitably put an end to their very existence and also because it’s just not possible. This phrase may however crop up once science and politics come into contact, based on some misunderstandings about the nature of science and the nature of scientific advice, as well as about the nature of truth and certainty.
Constatives and performatives
In the early 1960s the philosopher of language John Law Austin made a distinction between constatives and performatives. (Second proviso: In the following I’ll use this distinction as a heuristic device to think about scientific advice, but as speech act theorists will tell you, the distinction is very fuzzy and so is, in the end, the distinction I make between science and advice).
Austin was one of a number of philosophers who realised that language is more than a device for representing thought; that engaging in language is engaging in action; and that every utterance is a ‘speech act’. In his book How to Do Things with Words (1963), Austin made (initially) “a reasonably clear-cut distinction between constative and performative utterances. According to him, an utterance is constative if it describes or reports some state of affairs such that one could say its correspondence with the facts is either true or false. Performatives, on the other hand, ‘do not ‘describe’ or ‘report’ or constate anything at all, are not ‘true’ or ‘false.’ . . . The uttering of the sentence is, or is part of the doing of an action, which again would not normally be described as saying something.’ Marrying, betting, bequeathing, umpiring, passing sentence, christening, knighting, blessing, firing, baptizing, bidding, and so forth involve performatives. …. Whereas the constative utterance is true or false, the performative utterance is felicitous or infelicitous, sincere or insincere, authentic or inauthentic, well invoked or misinvoked.” Austin later replaced this dichotomy with a distinction between what is said (locution), the act performed by saying something (illocution) (which now subsumes constatives and performatives) and perlocution, the effect or consequence of saying something. I will here talk about performatives in terms of illocutionary speech acts.
Stating scientific findings and giving scientific advice
I think it might be possible to apply the albeit fuzzy distinction between constatives and performatives to elucidate some aspects of the relation between science and scientific advice or between what scientists do when they do science and what they or politicians do when they issue scientific advice (and, again, things are more complicated of course than I make out here).
I believe there is a difference between scientists announcing, describing or stating scientific findings and scientific advisors, policy makers etc. giving scientific or science-based advice. Statements about scientific findings can be regarded as ‘constatives’ and can be true or false. In most cases they are something in between and they are couched in terms of a whole array of uncertainties, probabilities, error bars and so on (e.g. under certain circumstances, given a certain genetic make-up, with such and such provisos etc. etc, smoking can be fatal). (Sometimes statements of that sort can be or become performative and have consequences, of course, but I disregard that here for the moment.)
Evidence-based or scientific advice, by contrast, is more clearly ‘performative’. Based on certain scientific findings (or statements of evidence) policy makers who have digested them, give advice, alert people, give warnings, for example not to smoke (Don’t smoke! Or: I hereby warn you: Smoking will kill you!). These speech acts are normally ‘categorical’ or what Sheila Jasanoff calls ‘binary‘. They propose certain courses of action that should be carried out or not carried out (to avoid smoking, for example). If these speech acts are surrounded by too many hedges, provisos, disclaimers and so on, actions will not follow and the advice, warning etc will fail. Advice as a speech act is neither true nor false, but it can succeed or fail. For advice as a speech act to succeed demands certainty, whereas stating scientific findings is grounded in (various degrees of) uncertainty. (Another blog could be written about managing degrees of certainty within advice giving, and the political contexts in which what one may call absolute certainty works and those when it doesn’t).
Truth and certainty
So there is, it seems, a certain difference between announcing or stating scientific findings as a result of doing science and giving science-based advice (on the basis of these findings) as part of doing politics. At the same time there is also a difference between the meanings of certainty and uncertainty in these two domains. Whereas science is inherently uncertain (and scientists are very well aware of this and say it frequently when issuing statements about ‘scientific facts’), science-based advice is inherently certain. (Yet another blog could be written about how being open about scientific uncertainty can help or hinder advice giving and persuading people to adopt certain courses of action)
It is important to note that the words certain and uncertain when used in these two domains are not just words with opposite meanings (antonyms). They are used in very different ways and have quite different meanings (they are what Wittgenstein would call part of different ‘forms of life’). One can perhaps distinguish between ‘constative certainty’ (which can never be achieved and only approximated) and ‘performative certainty’ (which has to be achieved or at least be approximated). Antonyms would then be constative uncertainty (the norm in science) and performative uncertainty (to be avoided in politics).
These differences between speech acts and between the meanings of certainty/uncertainty related to them sometimes lead to confusion when scientists speak to policy makers/politicians etc. and when politicians/policy makers etc. speak to the people that may be affected by that advice. This might also lead to misunderstandings about truth and certainty in science. As I said at the beginning, there never ever can be absolute truth in science. However, what policy makers and decision makers most often seek is some sort of holy grail of science that provides them with absolute truth and certainty. In the real world, they have to deal with giving advice and making decisions under conditions of (constative) uncertainty. And yet, what they say in terms of advice has to be couched inevitably in terms of categorical (performative) certainty.
Science, advice and politics
Courses of action that people are advised to follow or life-styles they are advised to adopt are of course and have of course to be political by their very nature and are political in their consequences, whereas statements of scientific findings are not and should not be (although they sometimes become so). Statements of scientific findings are true or false (and mostly everything in between), whereas the advice based on those findings is right or wrong (or something in between). This is also linked to the distinction between facts and values. Given these distinctions, one can understand the reluctance of some scientists to step from one ‘form of life’ and one type of language into the other. One can also understand the confusions that might arise from conflating these two worlds, the speech acts used in them and the types of truths and (un)certainty on which they are based.