David Wootton’s The Invention of Science (Allen Lane, 2015) is witty and learned and gloriously ambitious, and although I am not convinced that science as we know it came into existence in the seventeenth century, as Wootton argues, I do think that the seventeenth century is the one that lays the strongest claim to hosting that momentous event. I also agree with much that Wootton has to say, in this book and elsewhere, about how the history of science should be done, and especially with his view that there is a place for hindsight in the study of past science. However I do think that there is a blind spot in Wootton’s hindsight. He suffers from some of the same confusions as his adversaries, such as the historian and sociologist Steven Shapin, and the gap between Wootton and Shapin becomes much smaller when we clear up these confusions.
What Wootton wants to be allowed to do in his history books is this. When he finds someone from the past holding a belief that we now consider false, he wants to be able to say that it is false. For example, when he finds someone saying that ice is heavier than water, or that blood-letting cures syphilis, he wants to be able to say that these beliefs are false. Similarly, when he sees Galileo saying that ice is lighter than water, and when he reads a paper from 1857 claiming that blood-letting cures very little, Wootton wants to be able to inform his readers that these claims are true.
Why would Wootton want to do this? One reason is that he wants to make interesting observations about the past. For example, it is interesting to observe, as Wootton does in Bad Medicine: Doctors Doing Harm Since Hippocrates (OUP, 2006), that European doctors were much better at curing diseases in the twentieth century than they had been in the previous twenty-three centuries.
Wootton’s aim as a historian is not just to observe events but also to explain them, and the second reason he uses hindsight is to assist his explanations. He wants to explain why false beliefs flourish and why they fall into obscurity. He also wants to use his knowledge of nature and humanity to explain the experiences of people who, for whatever reason, did not have his knowledge. For example, he says in Bad Medicine that one of the reasons doctors were wrong about blood-letting for so long is that they made unconscious use of the placebo effect. If hindsight were banned in history, this explanation would be banned because the doctors in question did not know about the placebo effect.
Let me now turn to the historians that Wootton has in his sights in Bad Medicine, in The Invention of Science, and in the pages of the London Review of Books (14 Dec 2006) and the Times Literary Supplement (19 Oct 2011). I will call these historians ‘relativists’, since that is Wootton’s name for them. Stephen Shapin is prominent among them, though Shapin might not accept the label ‘relativist.’
What do these relativists want? Well, the main thing they want, when it comes to explaining past science, is to make room for sociological explanations of true beliefs. They want to be able to say that the 1857 refutation of blood-letting was written by someone with an axe to grind, or with a cause to promote, or with money on the line. The mere fact that the author reached the right conclusion does not mean that his route to that conclusion stands up to rational scrutiny. The fact that he was right by our standards doesn’t mean that his reasons were good by our standards. Conversely, the fact that most doctors of the day were wrong about blood-letting, does not mean that this conviction was the result of malice or subterfuge or self-delusion. Maybe they were rational and disinterested but unlucky. One can be wrong for the right reasons as well as right for the wrong ones.
Can Shapin and Wootton both get what they want, without conflict or compromise? I think they can. Shapin wants to suppress hindsight under certain other conditions, and Wootton wants to exploit hindsight under certain other conditions. Because the conditions do not overlap, their respective projects should not come into conflict.
Consider again the 1857 article, which was written by the Edinburgh professor John Hughes Bennett. Shapin wants to ban the following inference: Bennett’s conclusion was true, therefore he had good reasons for holding it. If we make this concession to Shapin, what exactly have we conceded? Not very much at all. We can still state the premise of the banned inference (ie. Bennett’s conclusion was true), as long as we don’t draw the banned inference. We can also use that premise in ways that are not covered by the ban. For example, we might say that the inefficacy of blood-letting is part of the explanation for Bennett’s belief in its inefficacy. This is not the same thing as saying that the inefficacy of blood-letting was one of Bennett’s reasons for rejecting blood-letting. (Compare: there are synapses in my brain that help to explain my belief that today is Thursday, but if you asked me to justify this belief, I would refer you to my calendar, not to my brain chemistry). Finally, there is nothing in Shapin’s ban that prevents us from saying such things as ‘one of the reasons so-and-so died of syphilis is that blood-letting does not cure syphilis, and blood-letting is what her doctor gave her.’ It looks like Wootton can do everything he wants without treading on Shapin’s toes.
So why do Shapin and Wootton spit daggers at eachother in the pages of literary magazines? My view is that neither of them have been able to express their position precisely enough to see how neatly their two positions fit together. Let me start with Shapin. The position that I have been calling ‘Shapin’s ban’ is not easy to find in Shapin’s writings. Or rather, when he states that position he usually follows it up with much stronger claims, such as the claim that historians should pretend that the natural world has nothing to do with what scientists believe about it, or the claim that history is an entirely descriptive enterprise that has no room for evaluations of the kind ‘Bennett was right about blood-letting.’
It is not surprising that Wootton takes issue with these strong claims and that he argues for the permissibility, and indeed the necessity, of some kinds of historical hindsight. However Wootton could be clearer and more charitable when he discusses his opponents’ position. Consider the following passages from The Invention of Science, where he discusses the Strong Programme in the sociology of knowledge:
“Its essence lies in the principle of symmetry: the same sort of explanation must, according to this principle, be given for all types of knowledge claims, whether they are successful or not” (p. 43).
Shapin would probably agree with this characterisation. Shapin might also agree with Wootton’s illustration:
“Thus if I meet someone who claims the earth is flat, I will seek a psychological and/or a sociological explanation for their peculiar belief; when I meet someone who claims that the Earth is a sphere floating through space and orbiting the sun, I must look for exactly the same sorts of explanation for this belief too” (ibid).
So far, so good. But Wootton’s next sentence is bound to raise hackles among sociologists:
“The strong programme insists that it is illegitimate to say that the explanation for the second belief is that it is right, or even that people believe it because they have good evidence for it” (ibid., my emphasis).
The view that Wootton seems to be ascribing to the strong programme is that the ‘evidence’ for a belief can never be used to explain a belief, and that only ‘psychological and/or sociological’ factors can explain beliefs. Now, it is true that some versions of the symmetry principle can be read in this way. But this is not the case for the ‘principle of symmetry’ that Wootton stated in the first sentence I quoted above. That sentence implies that we can explain true beliefs in terms of the evidence, as long as we also explain false beliefs in terms of the evidence.
The symmetry principle that Wootton quotes later in his book is even milder. It says that
“all beliefs are on a par with one another with respect to the causes of their credibility….regardless of truth or falsity the fact of their credibility is to be seen as equally problematic” (p. 581, quoting Barry Barnes and David Bloor).
All that Barnes and Bloor seem to be saying here is that all beliefs should be explained. They do not say that all beliefs should be explained in the same way.
Alert readers will notice that I misquoted Wootton two paragraphs ago. He wrote ‘good evidence’, whereas I wrote ‘evidence.’ This is a crucial distinction. There is nothing ahistorical in the claim that (for example) Galileo’s belief in the motion of the earth was partly due to the fact that he believed that there are bodies orbiting Jupiter. This is a descriptive claim that we can verify by reading Galileo’s Dialogue on the Two Chief World Systems (1632). It is quite different from the normative claim that the existence of bodies orbiting Jupiter is a good reason for believing that the earth moves.
A related distinction is that between explaining a belief and explaining the truth of the belief. Galileo’s belief in the moons of Jupiter explains the fact that he believed in the earth’s motion. But it does not explain the fact that the latter belief was a true belief. To explain that fact we would need to show that Galileo had done something that is conducive to holding true beliefs. In other words, we would need to show that his reasons were good ones.
So we have a series of distinctions—between excluding evidence and including it impartially, between explaining all beliefs and explaining all beliefs in the same way, between evidence and good evidence, between a belief and the truth of the belief—and all of these distinctions are crucial if we want to give a clear account of the symmetry principle. Shapin, Barnes and Bloor tend not to make these distinctions, and as a result they tend to rule out hindsight altogether. Wootton reaches the opposite conclusion—hindsight is almost always legitimate—but he is not much better than the sociologists when it comes to distinguishing the legitimate uses of hindsight from the illegitimate. The result is an unnecessary polarisation of opinion.
There is an easy way to end the deadlock. Wootton can get what he wants if he admits that Shapin has a point—scientists can be wrong for the right reasons and right for the wrong reasons. And Shapin can get what he wants if he admits that this point only goes so far—the symmetry principle, properly understood, does not prevent Wootton from making observations about the efficacy or otherwise of past medicine, and nor does it prevent him from explaining these observations or from using them to explain other observations.
This road map for methodological peace is not a perfect one. My plan means that both Shapin and Wootton would need to give up some of their more extreme claims. But they would both retain the claims that matter to them as historians, i.e. the ones that make room for the kind of books they want to write.
o-o-o
Postscript on progress. In the above post I have said nothing about progress. This may seem remiss, since Shapin and Wootton frame one of their exchanges (see here and here) in terms of progress. If Shapin and Wotton can agree about hindsight, will they automatically agree about progress? No, but their remaining disagreements would not be susceptible to the kind of road map I have sketched above.
They would at least agree on one thing concerning progress, namely that the fact that statements about progress are judgements does not make them ahistorical. If Shapin relaxed his ban on judgements, as I think he should, then he would also relax his ban on discussions of progress.
But Shapin has other reasons for shunning discussions of progress, besides the fact that such discussions involve judgements. One of these reasons is that progress is always linear and triumphal, and that modern historians have no truck with narratives that are linear and triumphal. Shapin does not make this argument explicit, but it is implicit in the first three paragraphs of his review of Bad Medicine.
The reason this argument is not susceptible to a road map is that it doesn’t deserve one. It’s just such a bad argument. Why not say, as George Sarton and many others have said, that science makes progress but that it does so in a piecemeal, complicated and zig-zagging fashion, with a great many dead ends and false starts? Shapin appears not to consider this possibility. He gives us only two options – either science makes linear, triumphal progress, or it makes none at all.
Shapin has two other reasons for rejecting discussions of progress. These are beyond the reach of road-maps because they are too large. They rear their heads in the latter part of Shapin’s review of Bad Medicine, where he assumes for the sake of argument that historians can discuss progress, and where he shows that Wootton has missed out important considerations in his own discussion of progress—considerations such as the fact that modern medicine has created illness as well as curing it, and that there have been changes in the criteria we use to decide whether a given consultation was successful.
The large issues here are empirical (does modern medicine have a better track record than past medicine, all things considered?) and philosophical (is there an objective set of criteria for judging medicine, criteria that we can meaningfully apply to medicine of all epochs?). No road map is going to bring Shapin and Wootton into agreement on these issues. They are issues that require debate and deliberation and empirical research. But I suspect that they will be easier to resolve if we can reach agreement on hindsight, and the debate about hindsight is on just the right scale for road maps.
o-o-o
Featured image: https://upload.wikimedia.org/wikipedia/commons/9/93/Galileo_Donato.jpg