A couple of weeks ago, Eric wrote about books that "that try to get away with systematically ignoring existing scholarship and alternative views," suggesting (on the basis of Colin Howson's review in NDPR) that Michael Strevens's Tychomancy might be an example, alongside some others that have been discussed here. We offered Strevens space to respond to Howson's review. Here is what he says:
"I highly recommend the beginning and end of Colin Howson's review of my book Tychomancy. The beginning summarizes the book's aims nicely, and as for the end, what author would not like to hear that their book is "far from being without merit", if perhaps not exactly in that phrasing?
Then there is the middle 80%.
Tychomancy formulates a set of rules for inferring physical probabilities from non-statistical facts such as physical symmetry (e.g., the 1/2 probability of heads from the symmetry of a fair coin) and makes three claims about these "rules of equidynamics":
1. Psychological: they are inherent in every human's mind.
2. Historical: they have been used (usually implicitly) to make important discoveries by scientists such as Maxwell and Darwin.
3. Epistemological: the rules are reliable (though not infallible).
Reading Howson's review, you would think that the purpose of the book is rather to make mathematical or physical claims, contributing to a body of literature about the probabilistic-looking behavior of things such as tossed coins, roulette wheels, particles bouncing around boxes, and even animals bouncing around ecosystems—the literature on "the method of arbitrary functions". Tychomancy in fact makes no attempt to contribute to this technical literature. It rather uses it—extensively—to explain the reliability of the equidynamic rules.
Howson also complains about my "failure to reference properly" the arbitrary functions literature. You might think, reading this, that there are items missing from my list of citations. But there are not. Everything Howson mentions is in my bibliography. What I do not do is to credit particular results to particular thinkers. So, for example, Howson complains that I don't attribute a certain independence result to Hopf. (Actually, Reichenbach scooped Hopf.) But the presentation of the results on independence (which go well beyond Hopf) is a part of a long section—almost a chapter—that is prefaced with an attribution to Poincaré, Reichenbach, and Hopf. I have failed to pinpoint Hopf's precise contribution to the project, but so what? The book is neither a history of nor a contribution to the arbitrary functions literature. It applies the mathematical and physical findings making up that body of work; what's apt, then, is a careful explanation of these findings, together with a summary identification of the major contributors—which I provide.
In the low point of his review, Howson cites a passage from p. 219 about scientists'—physicists' and biologists'—failing to notice the role of physical intuition about probability in model building, and interprets it as a self-aggrandizing claim to have made an outsized contribution to the arbitrary functions literature. The passage in question has absolutely nothing to do with that literature or with my contributions to it. (In a piece of equally hair-raising hermeneutic acrobatics, Howson interprets a discussion of humans' susceptibility to the gambler's fallacy (p. 54) as a denial that we are capable of learning from experience. Not the opposite of what I say—just something else entirely. He is a little too eager, I think, to find problems.)
Tychomancy is about rules of equidynamic inference actually used by ordinary humans, not about even better or more precise rules that we might have used but do not (interesting though that would be). A failure to keep this distinction in mind lies behind some of Howson's other complaints. He chides me for failing to present the interesting work of Diaconis and others on tossed coins, though as he himself remarks, the results are totally unexpected. Their very unexpectedness shows that they are not predicted by our equidynamic rules, so are not a proper part of my subject matter. (That said, I couldn't resist describing Diaconis's work in a footnote.)
Another example: I get into trouble for saying that "causal independence implies stochastic independence". But I don't say that; nor do I believe it, except (as related in the book) for a notion of causal independence so extreme as to have no practical application. What I do is to attribute to ordinary humans a heuristic principle defeasibly warranting under certain circumstances an assumption of stochastic independence given causal independence; I then claim that, though the principle is not infallible, it is reliable in the circumstances in which it is typically applied.
Only one of Howson's criticisms takes aim directly at the theses of the book: he opines that the equidynamic rules I attribute to ordinary humans are too complex; it is implausible that we have such fancy stuff in our heads. It seems crazy to me to decide what might or might not explain our patterns of thought on an a priori basis. My equidynamic rules, though not stupidly simple, might very well be the simplest things capable of accounting for the considerable sophistication and complexity of our inferences about physical probability; certainly, Howson offers no rivals.
In any case, I have a more immediate complaint about Howson's complexity objection: in presenting my thesis to the reader, he indiscriminately mixes vocabulary that appears in my formulation of the rules themselves with far more technical vocabulary that appears only in my explanation of the reliability of the rules. To take one example, reading Howson you would think that my equidynamic rules attribute to the average person thoughts about the dynamic property I call microconstancy. But in fact, none of the rules mentions microconstancy; the notion is used only in the vindication of the rules. So Howson makes the rules appear to be more complicated and more technical than they actually are.
OK; out of space. Read the review; but please, also read the book!
Recent Comments