Richard Marshall interviews Tim Lewens in 3:AM Magazine:
3:AM: You’re interested in philosophy of science, bioethics and the science of human nature among other things. In trying to work out what science is you look at some borderline cases: you vividly describe economics as being less science more ‘Lord of the Rings with equations’, Intelligent design as hopeless and homeopathy’s effects as being no more than placebo effects. So how do you draw a line between science and non-science? Do Popper, Kuhn, Lakatos and Feyerabend et al still offer helpful insights or have they been supplanted by better approaches?
TL: As you’ve indicated, I think that philosophers are well placed to expose significant flaws in diverse pseudo-intellectual endeavours. Intelligent design theory, for example, really is laughable, and it’s not too hard to show why. Even so, these evaluative tasks don’t require that we have a single criterion that allows us to sort the scientific wheat from the chaff, and there are good reasons to think these is no single criterion. First, the sciences are exceptionally varied in their methods: even some forms of economics are valuable! The sciences need to be varied because the universe itself contains many different types of phenomena, which need to be probed with different tools. Second, debates over the propriety of fields of learning have many dimensions. The case of homeopathy illustrates this. Even if it turns out that homeopathic remedies draw solely on placebo effects, we need to remember that placebo itself is a fascinating and little understood phenomenon. Placebos differ in their intensity: placebo capsules are more efficacious than placebo pills, and four placebo capsules are more efficacious than two. The process of medical consultation with a professional also has a strong positive placebo effect. And placebo has a maleficent twin: the nocebo effect means that if you expect a drug to do you harm, it can end up damaging your health even if it’s just a sugar pill in disguise. We also need to remember that mainstream drugs aren’t always as beneficial as is thought: the best research suggests that standard drugs used for moderate depression do no better than placebo. The upshot of all this is that we shouldn’t write off homeopathy too quickly: if you have moderate depression, and if you are suspicious of mainstream medicine, you might be best off visiting a homeopathic practitioner. You will avoid the nocebo effect you would get from standard treatment, and instead you will get a big placebo boost from the elaborate, bespoke consultation the homeopathic practitioner is likely to offer.
So does this mean that we still have something to learn from Popper, Kuhn and the other big beasts of mid-century philosophy of science? Yes! Feyerabend is right, I think, to cast doubt on the existence of any recipe that will tell scientists how to go about investigating the world: the interesting question for us is whether this really means that in science anything goes. Evidently these are issues with hefty ramifications for decisions over funding, and the politics of what gets taught in schools. And we need to understand Popper’s appeal better. It’s still the case that scientists point to Popper when asked how they do their business. I suspect that they endorse an eviscerated, but very sensible, version of Popper’s falsificationism. They think that Popper tells us that scientists test their theories against the world, and that scientists never invest their views with certainty. But Popper’s views are far more radical than that: Popper gives us no grounds for thinking that any scientific tests can ever have significance, because he ultimately denies any epistemic authority to scientific data. And he doesn’t merely say that scientists aren’t certain, he says they have no reason whatsoever to think their theories are close to the truth.