Friday, June 25, 2004

Have a Little Peppered Moth with Your Science....

I have just recently finished reading Judith Hooper's Of Moths and Men, which receives my high recommendation. In saying this, it is worth noting that I have become very picky about pop-sci books. Picky? Cynical is the better word. This is especially true of pop-biology, which is the bane of all good reason. This is sad, since it does not have to be this way. Pop-physics, for instance, even when it exhibits all the flaws of pop-sci (e.g., a scientist's notion of what good literary writing is * shudder *), tends to be reasonably good. Most pop-sci suffers from an inability to reason properly, but pop-biology takes this sometimes to bizarre levels. Note to all people inclined to write scientific popularizations: Keep it simple, straightforward, and let the charm of the science (and scientists) do the real work in adding charm to the writing.

Hooper's book follows this advice quite well. There are a few weaknesses. For instance, her discussion of "scientific creationism" lacked a genuine insight into the dispute, and her appeals to falsifiability are too uncritical. Nonetheless, the weaknesses are astoundingly few. What chiefly impressed me was that the book gets its primary interest from the (hint of) scandal of the peppered moth, but does so with considerable sympathy for all those involved in it (this is particularly true of Kettlewell, who despite being the center of the (possible) scandal is almost the hero, both flawed and admirable, of the book). This was the right way to go, and Hooper deserves kudos for it. It enables her to show, to a degree rare in such books, the course of science as a genuinely human endeavor, with all our foibles, charms, faults, and ingenuities thrown into the mix. What we often call 'science' is usually just its residue in books and articles; Hooper goes some way toward introducing people to the real nature of science. Some way; there is much more to be done.

-> Here is a much more critical review. It seems to me they have missed the point of the work, since Hooper bends over backward to avoid alleging fraud. The book, in fact, ends with a praise of Kettlewell's work, despite the problems with the standard peppered moth account that arose out of it. Hooper, in fact, does not argue that Kettlewell cheated; she argues that there are signs that Kettlewell tweaked the experiment as he tweaked most of the things he did: he wanted his experiment to work, so he nudged it a bit in an attempt to get a clearer result. One thing that comes out of the book is Kettlewell's innocence: it is impossible by the end of the book, assuming the reader actually has reading skills, to conclude that its message is that Kettlewell committed fraud. It is worth reading, however, the Matt Young essay to which the Wikipedia article links, since it adds information that suggests the matter is more complicated than Hooper suggests. Since much of Hooper's book is devoted to showing how the matter is much more complicated than the standard account allows, this is not the knock-down rebuttal Young seems to think it is. And the rational conclusion for the paper would be not, as Young seems to think, "The peppered moth properly remains a valid paradigm -- no, an icon -- of evolution." First, because Young does not deal with all of Hooper's questions, and second, because Young's own argument does not show that Kettlewell was right but only that if other factors have the effect he thinks they might have, Kettlewell could be right. The conclusion Young should draw from this is, I suggest, the following: We have reason to think the problems attributed to the peppered moth experiment may not be problems; further research is needed to clarify the matter. (A good example of the way Young's argument, almost right, goes wrong, is in the moonlight issue, which I found to be a fascinating discussion. Young's argument on this point relies wholly on his statistical analysis of the Kettlewell case, and requires that he conclude of the actual research done on the effect of moonlight on recapture rates, "I do not think that the conclusion of Clarke and colleagues is necessarily pertinent" - not a strong conclusion. Young is reasonable in concluding it, but it is not strong enough to support the conclusions with which Young finishes the paper. Young should have instead concluded, perhaps, by noting that there is good reason to think moonlight might affect recapture rates in a way that deals with this particular issue, and pointing out how Kettlewell could be supported on this by actual experiment.)