I began posting on the symmetry principle in March 2013, in response to a post by Vanessa Heggie on the H Word. After eight posts and nearly eighteen months, it is time to bring this desultory marathon to an end. In the interests of brevity and coherence, here is a six-step guide to saving the symmetry principle. Each step corresponds to one or two posts in the series.
1. Get the fallacy right. What kind of inference does the symmetry principle forbid? Here is one kind of false inference: the earth really does move, so the motion of the earth is a sufficient explanation of Copernicus' belief that it moves. Very few historians have ever made this inference, so there is little point in fulminating against it. The real worry is inferences like this: the earth really does move, so the evidence for the earth's motion is a sufficient explanation of why Copernicus believed it to move. The difference is between explaining a belief by its truth and explaining a belief by the believer's arguments and evidence.
2. Take distinctions seriously. To get the fallacy right we need to distinguish between the truth of a belief and the evidence in favour of a belief. This may seem like an esoteric distinction, but we have no choice but to make the distinction if we want to understand the symmetry principle. The distinction is especially important if we want to communicate the symmetry principle to non-historians. If we do not make the distinction ourselves, we cannot expect non-historians to do so, and if they do not do so then we should not be surprised if they reject the symmetry principle. Another reason to make the distinction is that evidence gets a bad name if it is constantly confused with truth, since evidence usually explains more than truth.
3. Make room for truth. Truth nevertheless does explain something. The motion of the earth may not be a sufficient explanation of Copernicus' belief, but it is surely part of the explanation. There are other ways in which truth can legitimately enter into writing about past science. For example, it is philosophically respectable to say that theories are true insofar as they correspond to reality. And it is undeniable that factual evidence has played a role in the beliefs of past scientists. A blanket ban on truth does more harm than good in the history of science.
4. Be a minimalist about symmetry. Every historian of science knows that she should treat true beliefs in the same way as false ones. But what does this mean in practice? Does it mean that both sides of all past debates in science were equally reasonable? Or that historians should aim for completeness, always reporting the good reasons on the losing side and the bad reasons on the winning side? Or does it mean that we should ignore good reasons altogether and focus on social factors rather than cognitive ones? Or that, as a matter of fact, the social factors dominate in most cases? Properly understood, the symmetry principle does not imply any of these things. It simply forbids us from inferring the goodness of a believer's reasons from the truth-value of their belief.
5. Face up to the skeptical challenge. The reason this inference is forbidden is that people can be right for the wrong reasons and wrong for the right reasons. But what if we apply that principle to today's science? Maybe our best theories are wrong, despite the good reasons in their favour. Historians tend to welcome this thought as an antidote to dogmatism. In fact it is a serious threat to the symmetry principle, because few people will accept a principle that implies that today's theories are no better than Aristotle's. We need to work harder to reconcile symmetry and scientific realism. For example, we can observe that the argument from realism to asymmetry only works for consensual beliefs and not for those under dispute.
6. Aim for understanding not self-delusion. Suppose that symmetry is indeed consistent with realism. Shouldn't we nevertheless suspend our realism when we do history, just to make sure that it does not distort our results? Shouldn't we forget that the earth moves when we study Ptolemaic astronomy? No, we should not. If our knowledge of our former skews our view of the latter, then we have failed to grasp some fundamental points about past science. We should try to grasp these points rather than playing evasive mind games.
If we take these steps then we have a good chance of rescuing the symmetry principle from the false or irrelevant principles that are sometimes associated with it. These steps focus on what the symmetry principle is not, rather than what it is. But we need both to grasp the symmetry principle. So here is one more rendition of the minimalist, truth-friendly, and realist-friendly, version of the principle that I have defended in this series: the truth of a belief does not imply that the believer had good reasons for holding that belief, and the falsity of a belief does not imply that the believer had bad reasons for holding that belief. In short, historians should not reason from truth-value to goodness-of-reason.
Expand post.
Malheureusement je ne traduis plus tous les posts que j'écrit sur ce blog. Si vous voudriez lire les posts qui ont été traduits, veuillez cliquer ici.
Agrandir ce message.
I think this Simon Schaffer quote started the whole ball of wax rolling (it stimulated Vanessa Heggie's post):
ReplyDelete[I]t was extremely unpromising, to put it mildly, to suppose that social principles are only acting when folks get things wrong. So for example, it didn’t look remotely plausible to say that Isaac Newton thought that there was an inverse square law of gravity acting instantly at a distance through empty space between the centers of distant bodies because there is an inverse square law acting instantly from the center of one body to another through empty space, and Leibniz disagreed because he was German.
A demand for even-handedness --- thus "symmetry". E.J. Aiton's definitive treatment of the vortex theories of gravity (Leibniz's and others) is wonderfully symmetrical in this sense, focussing almost entirely on the participants' own arguments, and spending very little time on social factors. (Nowhere does he bring up Leibniz's nationality.)
Maybe we shouldn't read too much in Schaeffer's joke (made in an interview, not a published paper). But the rest of the interview certainly strives hard to deprecate internalist approaches. I haven't run across any plausible target for Schaeffer's quip.
I like to summarize your point (6) as, "Historians can walk and chew gum at the same time."
At one time I planned to offer this as an "Asymmetry Principle":
ReplyDeleteTrying to discern the unstated motivations of historical figures is always a dicey business.
though now I prefer to call it the Diciness Principle. Whence the asymmetry? Well, it's a much dicier business than technical, internalist analysis.
We have an asymmetry between:
(1) The texts where the participants made their best attempt to explain their views, as clearly as they could,
vs.
(2) Guesswork about the participants unstated motivations.
The imbalance between (1) and (2) is palpable, but that doesn't mean we should exclude (2) from historiography. But I believe historians should display a little more diffidence when invoking so-called bad reasons as explanations.
A good example comes from two celebrated biographies: Michael Faraday, by L. Pierce Williams, and Michael Faraday: Sandemanian and Scientist, by Geoffrey Cantor. Williams' work does delve into Faraday's personal life, including his religion, and possible effects of his religious views on his science. But mostly it sticks to technical arguments and records of experiments, when trying to elucidate the growth of Faraday's scientific theories.
Cantor's book was reviewed by Williams (in Isis 85:1), who had this to say about it:
It is not legitimate historical method to state that someone has read a text, pull quotations out of a work so complex as the Bible whose central concern is not with the laws of nature, and then insist that the subject uses these texts to establish concepts of natural philosophy quite alien to the context.
There is, in these cases, at least some evidence offered. In other cases there is pure speculation. [Williams quotes a passage on the "Forgan thesis", about public and private spaces.] There is not a scintilla of evidence to support this passage, and I don't believe that much of historical worth can come from this kind of fictional history.
... [Cantor's] general approach remains the same. He is determined to show that the major elements of Faraday's science and scientific method were derived from his religion. These are not just attitudes, which I think everyone would concede, but quite specific aspects of his science.
I won't quote the whole review, but Williams makes clear he is not allergic to this sort of speculation, so long as it is clearly labelled as such, and doesn't infect the whole historical enterprise. Indeed, Cantor does include some throat-clearing about the need to demarcate speculation, but as Williams comments, "As Cantor gathers momentum, he gradually moves from the speculative to the declarative; the result will not please most historians."
To bring this back to the Symmetry principle: anytime we have a historical dispute between two figures (call them X and Y), we have a matrix of reasons:
"good reasons" by X | "bad reasons" by X
"good reasons" by Y | "bad reasons" by Y
Suppose the current verdict of science vindicates X. The symmetry principle means we should not confine ourselves to the main diagonal. The diciness principle means we should treat the right column (typically psychological guesswork) more gingerly than the left.
And finally, the historian's mandate means that all four entries are fair game.
I'm afraid I'm still unconvinced by your point (5), "Face up to the skeptical challenge". It's a big non sequitur to go from "Today's theories might still be wrong", to "Today's theories carry no more weight than old discarded ones".
ReplyDeleteHi Michael,
ReplyDeleteThanks for your comments, which are sprightly as always.
Regarding Schaffer's quote, I don't want to give the impression that this entire series is directed at Schaffer and Heggie in particular. It is not. They have simply given expression to ambiguities that are, I think, quite widespread in writings about the symmetry principle. In the coming weeks I hope to delve into the history of the principle in order to back up this claim and find out where the ambiguities entered the literature.
On Aiton, I hope it is clear from my posts that I am not insisting that we omit social factors from our histories. We can include them or exclude them without necessarily violating the symmetry principle. However I do want to stress that we can *exclude* them without violating the principle, since the symmetry principle is very often associated with the inclusion of social factors. I take it this is your point in invoking Aiton on theories of gravity.
Walking and chewing gum at the same time is indeed something that we historians should strive for!
"The symmetry principle means we should not confine ourselves to the main diagonal." Nicely put.
Your diciness principle seems plausible to me, though it probably does not hold across the board. Here the ambiguity of the term "bad reasons" comes into play. If we mean "reasons that the actors themselves considered bad," actors will do their best to hide such reasons, hence the detection of them will always involve a lot of guesswork. But if we mean "reasons that we consider bad today," then they may have been respectable reasons for some past scientists and therefore easy to find in the published oeuvre of those scientists.
It's worth adding that can be plenty of guesswork in the case of technical, internal history. Did Faraday get his metaphysics from the writings of the Croatian natural philosopher Roger Boscovich? This may be considered an internalist question, but it has generated as much speculation as the question of the religious origins of Faraday's theories. (Apologies to Faraday specialists if this is too crude a simplification).
Regarding point (5), I agree that it is a non sequitar to go from fallibilism to skepticism. My point in (5), and in the corresponding post (http://bit.ly/1rJRYqS), is that the symmetry principle appears to lead directly to skepticism (not indirectly via fallibilism). Ultimately I reject this argument for skepticism, but I think it is worth taking seriously -- more seriously than the simple conflation of fallibilism with skepticism.