A couple of weeks ago there was an interesting exchange in The Guardian between George Monbiot and Nicholas Maxwell, a philosopher of science from University College London. In his piece, Monbiot presents an excellent, if overly pessimistic, analysis of the psychology behind climate change denial. In his response, Maxwell draws on some interesting results from the philosophy of science to make some outlandish and possibly dangerous conclusions about climate science. He argues that climate science, and science more generally, doesn’t seek truth and nobody will believe what scientists say until they own up to this fact.
(Update: In the comments on this post, Maxwell claims that his views are misrepresented here. Scroll down to see the debate.)
Monbiot’s analysis of the psychology of climate change denial
Monbiot points to a problem with the climate change debate that I have been interested in for some time. He explains that the actual science is too complex for the public, in general, to make an informed judgement about so we must take what the experts say on trust. But taking things on trust seems to be counter to the lessons we have learnt from science.
The detail of modern science is incomprehensible to almost everyone, which means that we have to take what scientists say on trust. Yet science tells us to trust nothing, to believe only what can be demonstrated. This contradiction is fatal to public confidence.
Monbiot’s column finishes on a very pessimistic note. After discussing some very interesting research about the psychology of evidence and belief, he suggests that maybe there is nothing we can do to convince people of the real dangers of climate change. “There goes my life’s work,” he concludes.
I think that Monbiot is unnecessarily pessimistic and i think the deep tension he points to regarding trust in science can be resolved — but I’ll return to that issue in another post. In this post, I want to have a look at the bizarre claims made by Maxwell in his response.
Maxwell’s misuse of philosophy of science
Maxwell’s response is that scientists are giving people less reason to believe what they say since they are lying about their whole enterprise anyway! If they want people to believe what they say, according to Maxwell, they should stop lying.
The odd thing is, that according to Maxwell, the lie that scientists are going around telling us is that their research aims at truth. They should own-up, he says, and admit that they don’t seek truth at all.
The obvious problem with this advice is that if it turns out that science doesn’t aim at truth, then why should anyone believe it? Most people want their beliefs to be true and if they think that science doesn’t aim at truth, they won’t want to believe science.
So it’s clearly not good advice for scientists who want the public to believe what they say. But is Maxwell right about science? What reason does Maxwell give for thinking that science doesn’t aim at truth?
My first thought was that Maxwell would appeal to the kind of antirealism defended by philosopher Nancy Cartwright. In the introduction to her book How the Laws of Physics Lie, Cartwright says,
In modern physics, and I think in other exact sciences as well, phenomenological laws are meant to describe, and they often succeed reasonably well. But fundamental equations are meant to explain, and paradoxically enough the cost of explanatory power is descriptive adequacy. Really powerful explanatory laws of the sort found in theoretical physics do not state the truth.
I don’t agree with Cartwright but her work is important and very interesting. Maxwell’s claims do sound a bit like Cartwright’s but, at least in the Guardian piece, seem far less challenging. He says,
Physics only ever accepts theories that are unified – that attribute the same laws to all the phenomena to which the theory in question applies – even though many empirically more successful disunified rivals can always be concocted… The aim of science is not truth per se, but rather truth presupposed to be unified, or explanatory.
Proper definitions of measures of unification are quite complicated. For present purposes, it is useful to think that for one theory to be more unified than another, it must either explain more phenomena or make fewer fundamental assumptions, or both.
Maxwell’s article concludes,
In short, in holding that the intellectual aim of science is truth alone, scientists seriously misrepresent its real, problematic aims, and thus prevent urgently needed critical assessment by scientists and non-scientists alike. More honesty about the nature of science might improve science, and public attitudes towards it – and might even encourage scientists to produce less gobbledegook.
Maxwell’s point seems quite different to Cartwright’s. Cartwright’s (questionable) point is that no real-world situations actually obey the laws of physics: the laws of physics apply only to imaginary situations like frictionless planes. The generality or unification of physics comes at the price of truthfulness, she argues. The generality gained by the unification of the laws of physics comes at the price that they only accurately apply to imaginary circumstances.
Maxwell, on the other hand, is not arguing that our current laws of physics are wrong because they do not accurately describe the real world, but rather that they only describe the world as accurately as infinitely other, less unified theories. He says that scientists thus don’t seek truth, but rather “truth presupposed to be unified, or explanatory”.
As a matter of fact, he is right. There are infinitely many equally empirically adequate theories. And one of the main reasons for accepting one theory over another is unification — we choose theories that carry the least number of assumtions while explaining the most phenomena.
Why unification is a good guide to truth when choosing theories is an interesting question and its not clear what the answer is. One is tempted to simply say that unification has been a good guide in the past, so we should expect it to be in the future. But its not clear that this defense really works. (For one thing, the defense seems to assume some kind of unification between the past and the future which might introduce some circularity.)
Another line of defense is that when a theory is better unified – when it explains more phenomena – it is also better confirmed. A theory is thought to be confirmed by the number of phenomena that it explains and so the more unified a theory, the better confirmed it is.
Or perhaps the right answer is that given by some pragmatists: what we mean by truth, is that which is, in some way or other, useful. Theories completely lacking in unification do not explain much at all and as a result are of very little use. If that’s right, then it’s almost a matter of definition that the more unified theories are more likely to be true.
Let’s forget about truth. What matters is usefulness.
But the important thing to point out is that the way Maxwell presents the issue of unification is somewhat misleading. It is not that scientists aim at unification instead of truth but rather that they think that unified theories are more likely to be true. Given the choice of two theories, the more unified one is thought a better contender at truth. As I’ve pointed out, why that has been the case in the past and why we think it will be the case in the future is an interesting question and one worth investigating.
At the heart of Maxwell’s view is the claim that usefulness and truthfulness come apart — that true theories might not be the most useful (for things like predicting the future). We know that theories that are more unified are more useful, but Maxwell claims that they are not more likely to be true. If that’s the case, then when it comes to matters of practical significance such as climate change, it’s not at all clear why we should care about what Maxwell is calling truth. What matters is that the theories are useful: that they provide the right predictions. Thus, even if Maxwell is right about scientists not aiming at the real truth, he doesn’t give us any reason to think that they should.
It seems to me unlikely that truth and usefulness come apart as Maxwell thinks they do. But if they do, he needs to tell us why a “truth” that doesn’t explain much — and therefore doesn’t predict much — is worth pursuing. What we need is a theory that makes the correct predictions, not one that conforms to Maxwell’s peculiar definition of truth.
Kitcher, P. (1981). Explanatory Unification Philosophy of Science, 48 (4) DOI: 10.1086/289019