« The Urban Sink | Gene Expression Front Page | The history of the Jews...a very special people...sort of »
June 15, 2005

Bad science?

Howard Metzenberg has some strong words for Greg and Henry's paper. I wouldn't respond, but his objections/criticisms have been echoed in other places. I want to take a minute (literally) to address some of charges.

1) All theory, no research. W.D. Hamilton's work in the 1960s was done at his bedsit and in train stations. Biology can have theory, there is a Journal of Theoretical Biology. In Narrow Roads of Gene Land Hamilton claimed that his papers on kin selection were the most cited in the biological literature of the past generation, and they have led to an enormous field of empirical work in ethology. G.C. Williams work in the 1960s was also rather theoretical-philosophical, and stimulated others to rework their research programs on individual selectionist lines.

2) Other people's data. This happens all the time, social scientists do it, and Lewontin and Hubby's work with allozyme polymorphisms stimulated a lot of reworking of the theoretical foundations of evolutionary population genetics by the likes of Kimura (and the new field of molecular genetics).

3) "Provable." If Howard wants to wink at the Cult of Popper, he should say falsifiable, and that is something that Greg and Henry's work is.

4) Science is theoretical and empirical. But they don't have to be slapped together in the same scientist, or the same paper, or the same project.

5) I think in hindsight using the "Methods and Discussion" style format would have helped, but Greg and Henry have their own reasons I suppose.

6) Also, Howard says: 'He expects others to take on the hard work of analysis and proof, and credit his "brilliance" when they are finished.' I can't make transparent heads or tails of this, since "proof" is usually reserved for pure mathematics, and it seems one criticism of Greg and Henry's work is that it is theoretical analysis, but in any case, there is a long tradition of synthesizers who build on the work of others. Julian Huxley, and even more prominently E.O. Wilson, built their public reputations through synthetic tomes (The Modern Synthesis, Sociobiology) that distilled the scientific zeitgeist in their field to its raw essence and set the terms of debate for the next generation. Since scientific ideas are not bound by traditional property rights and reuse generally adds to the prestige of the originator of the ideas Huxley and Wilson are generally not reviled as thieves of the ideas of others.

I might have more to say later, but I think that gets to the crux of the major issues.

Update: I think there are two primary sources of the critiques of Greg and Henry's work. One some level many people have a visceral distaste for the mode of thinking, and the conclusions, that they have come to. So, they are trying to pick them apart on small (but numerous) points. Ultimately, it reminds me of the likes of Allison Gopnik simply dismissing any talk of heritability or the genetic underpinnings of behavior by noting that interactions between variables makes it difficult to parse the various factors, ergo, talk of such things is "incoherent." Of course, this would likely result in the dismissal of all forms of scholarship aside from mathematics and the most stringent of reducible physics as "incoherent." As a philosopher and psychologist I suspect Gopnik wouldn't go that far, but we know that "other fields are different." (I find it ironic how some wish to interpose the specter of epistasis into any discussion of human genetics so as to throw ink over any model that attempts predictivity, but feel free to make gross approximations for the sake of sociological models if those models support their normative values)

Then there is the media issue. This paper has gotten a lot of play for something isn't going to be in Nature or Science. But the fact is, the topic is sexy, it is about humans, it is about a "taboo" topic and it is backed by two people with reputations who have cobbled together a mathematical model that goes beyond impressionistic assertions (Henry is an NAS member). The concrete nature of what they assert, how they assert it, etc. not only makes them good targets to critique, but it is also something that will make good copy in the public press which likes nice digestable numbers and precise contentions that it can slot into a few paragraphs.

Posted by razib at 08:32 PM