In the blogosphere and on twitter I see lots of people wrestling with the question of whether we should “believe the science”. It’s clear that they are searching for some sort of Reliable Epistemological Principle, and struggle to articulate exactly what that principle is. Clearly the “science” is not always correct. So how do we know when to believe the science? And which scientists do we believe?

Richard Rorty made a career arguing that there is no Reliable Epistemological Principle. We need to look at each claim being made, and try to the best of our ability to figure out if it is true. The fact that lots of experts believe something is true is certainly one piece of evidence that deserves serious consideration. But that’s all it is.

I’m rather bemused by the anguish that some people clearly feel when they find out that science is wrong on some point, especially when the anti-science crazies took the opposite side of that particular issue. They seem to think this is some sort of threat to the Reliable Epistemological Principle, the final arbiter of Objective Truth.

People need to lighten up. We’ll be debating scientific questions from now until the end of time. Take comfort in the fact that science often has practical value. If you believe in the efficacy of vaccines, then you are likely to live longer than if you don’t.

Here’s Rorty:

We need a way of explaining why scientists are, and deserve to be, moral exemplars which does not depend on a distinction between objective fact and something softer, squishier, and more dubious.

To get such a way of thinking, we can start by distinguishing two senses of the term ‘rationality’. In one sense, the one I have already discussed, to be rational is to be methodical: that is, to have criteria for success laid down in advance. . . .

Another meaning for ‘rational’ is, in fact, available. In this sense, the word means something like ‘sane’ or ‘reasonable’ rather than ‘methodical’. It names a set of moral virtues: tolerance, respect for the opinions of those around one, willingness to listen, reliance on persuasion rather than force. These are the virtues which members of a civilized society must possess if the society is to endure. In this sense of ‘rational’, the word means something more like ‘civilized’ than like ‘methodical’. When so construed, the distinction between the rational and the irrational has nothing in particular to do with the difference between the arts and the sciences. On this construction, to be rational is simply to discuss any topic – religious, literary, or scientific – in a way which eschews dogmatism, defensiveness, and
righteous indignation.

There is no problem about whether, in this latter, weaker, sense, the humanities are ‘rational disciplines’. Usually humanists display the moral virtues in question. Sometimes they don’t, but then sometimes scientists don’t either. Yet these moral virtues are felt to be not enough. Both humanists and the public hanker after rationality in the first, stronger sense of the term: a sense which is associated with objective truth, correspondence to reality, and method, and criteria.

We should not try to satisfy this hankering, but rather try to eradicate it.

It seems to me that you can think of scientific institutions such as the CDC and FDA as having spent the last 18 months obsessed with rationality of the first type, obsessed with method. Trust the science became trust in rigid and arbitrary methods that were followed because no one in the bureaucracy had the courage to suggest another approach.

Alex Tabarrok spent the past 18 months engaged in rational analysis of the second kind, looking at all of the Covid-19 evidence in a reasonable way and trying to figure out the most useful way of interpreting that evidence. There is no cookbook for doing what he did; it’s either persuasive or it isn’t. Science is inevitably a messy human activity, reflecting all of our biases and assumptions. The output is not objective reality; rather it is useful models and tools that we can use to make our lives better.

And that’s OK.