“I think it right…thus publicly to read my recantation”

by Greg  Mayer

In an article in the March issue of The American Naturalist I find echoes of what I regard as one of the most stirring and admirable episodes in the history of science. In the article, entitled “Aspect diversity in moths revisited”, my colleague Bob Ricklefs tests a theory put forward by our late friend and teacher Stan Rand over 40 years ago. Stan had noticed while watching moths at lights in Brazil that

Not only were there a fantastic number of species but they were strikingly different form one another and these differences were accentuated by the variety of positions taken by resting animals.

Stan proposed that the great variety of color, form, and posture in the moths, which he called “aspect diversity”, was due to natural selection by vertebrate predators.  Being fairly smart, vertebrates learn the colors and shapes of their prey, and thus there is an advantage to appearing different from the other prey: the rarer your particular appearance is, the less likely it is that vertebrate predators will learn to seek you out.  Stan also proposed that aspect diversity should be greater in species-rich tropical areas.

In 1975, Bob and a colleague had made a tropical-temperate comparison which supported the second part of Stan’s proposal: aspect diversity was higher in the tropics.  So far this is interesting, but hardly stirring: hypotheses get tested all the time. The 1975 study was based on only three study sites.  Since then, Bob has continued to collect data as his field work allowed (he has worked on all sorts of things, but mostly birds, and moths are a bit of a side line for him), and he now has data on 18 sites.  Guess what: he and Stan were wrong!  In his new paper he reports that

Comparable analyses of 15 additional samples from Ecuador to Canada fail to support Rickelfs and O’Rourke’s original result…. [Their result] was an unfortunate consequence of the particular sites sampled…. In the complete data set… [aspect diversity] was unrelated to…tropical versus temperate latitude

He was wrong, and he forthrightly declares it.  This is exactly what we expect a scientist to do, and what good scientists do all the time: your views should accord with the data, and not vice versa.

It is a little unusual though that someone gets to so decisively refute his own prior work.  This is what reminded me of the incident alluded to above: the great British geologist Adam Sedgwick‘s recantation of the theory of a universal flood, a theory which he had done very much to promote, and which had appealed to him as an Anglican clergymen, as well as a geologist. This theory, which accorded so nicely with the Biblical Noachian flood, had been quite popular amongst English geologists, but the evidence turned against it, and it was realized that the supposed flood deposits were in fact records of the coming and going of a series of immense glaciers (the Ice Ages, as we call them now, are as wondrous and wonderful as any imagined flood).  The flood theory’s death knell was sounded by Sedgwick himself, who, in his 1831 presidential address to the Geological Society of London, disavowed the theory:

Having been myself a believer, and, to the best of my power, a propagator of what I now regard as a philosophic heresy [i.e. the flood or diluvian theory], and having more than once been quoted for opinions I do not now maintain, I think it right, as one of my last acts before I quit this Chair, thus publicly to read my recantation…

We ought, indeed, to have paused before we first adopted the diluvian theory, and referred all our old superficial gravel to the action of the Mosaic flood… In classing together distant unknown formations under one name; in giving them a simultaneous origin, and in determining their date, not by the organic remains we had discovered, but by those we expected hypothetically hereafter to discover, in them; we have given one more example of the passion with which the mind fastens upon general conclusions, and of the readiness with which it leaves the consideration of unconnected truths.

(I think I first heard of Sedgwick’s recantation from Stephen Jay Gould, and the above quotation is taken from his excellent treatment of the whole episode in The Atlantic.) The flood theory, contra creationists, has not been a live scientific theory since, and it was an Anglican priest who did it in, before Darwin (who was a student of Sedgwick’s) even set out on the Beagle voyage. One of the hallmarks of science is the tentativeness with which its results are held, and this is how both Sedgwick and Bob Ricklefs held their results.

I find it pleasing not just that Bob should be treading in such admirable footsteps, but that he should do so in a work he has dedicated to the memory of Stan Rand, in which he tests one of Stan’s ideas. Stan was, like me, a student of the great herpetologist E.E. Williams; it was Stan who actually taught me how to catch lizards. He reveled in the diversity and complexity of the tropical biota, and jokingly enunciated what we called ‘Rand’s Law of Inductive Generalization’: “The diversity of the tropics forces us to generalize from a single observation.” But, of course, one, or even three observations, isn’t enough, and maybe not even 18. Stan would be pleased.