Friday, July 22, 2005


Even I have to allow that some eternal verities have a kernel of truth. For example, the cliché that scientific results are always provisional is correct as a matter of principle since on any reasonable account of what the “empirical” in empirical science means, it doesn’t mean deduced deductively from unquestionable axioms. Formally speaking, we could perfectly well discover next Tuesday that DNA never codes for protein or learn to our surprise that due to a clerical error we were wrong all along about the atomic number of oxygen. At least in a contemporary context, however, the rhetorical force of such avowals of methodological modesty depends on a covert awareness that DNA damned well codes for protein and the atomic number of oxygen really is 8. An enormous number of scientific issues are up in the air, of course; but the guilty secret of the researchers is that they actually can and have achieved results that are materially correct. Nobody’s going to come along and overturn the periodic table. It just isn’t going to happen. I don’t think we’ve come to terms with this embarrassing fact and its many implications.

Thursday, July 21, 2005

The Fluoride of the Progressives

Robert F. Kennedy Jr. has been making the rounds of talk shows talking up the idea that autism is caused by thimerosal, a mercury compound formerly used in this country as a preservative in childhood vaccines. While some right-wingers are also upset about thimerosal, most of the concern about it comes from the left, which is peculiar in a way since suspicion about vaccination is traditionally a conservative preoccupation that reflects a programmatic unwillingness to bear even tiny personal risks for the general good. To be sure, Kennedy and Arianna Huffington and various folks at Salon and Rolling Stone claim to support vaccination and relate their attacks on thimerosal to an anti-corporate theme. Their narrative is all about how Big Pharma conspired with bunch of bought-off scientists to cover up the connection between mercury and autism. Of course if thimerosal really is a cause of autism, the motives of its critics don’t matter very much. What is the state of the evidence?

No official scientific agency in the United States and Europe has reported any link between thimerosal and autism or any other neurological problem, but a consensus developed that it should not be used in vaccines because of a general concern about mercury exposure. For this reason, thimerosal has not been used in vaccines in this country since 1999 while research into the possible effects of thimerosal has continued without turning up any very definitive results. As with many recent epidemological issues—the health effects of electro-magnetic fields, for example—the technical difficulties of defining the real risks are formidable, in part because the association between the agent and the effect is so weak. Nobody is claiming that flu shots are to autism what cigarettes are to lung cancer. Mercury is not good for the nervous system, but the toxicity of mercury depends upon its chemical form. Thimerosal delivers ethyl mercury to the body whereas environmental exposure to mercury largely takes the form of methyl mercury. The pharmacology of ethyl mercury remains to be elucidated. While methyl mercury tends to linger in the body, ethyl mercury is rapidly excreted in the feces. On the other hand, some babies may not be as capable of excreting ethyl mercury as others; and ethyl mercury from a vaccine injection comes as a single big dose while methyl mercury from food and water is ingested a little at a time. A bolus of mercury may produce a more severe effect than an incremental exposure.

The case against thimerosal also assumes that the incidence of autism has increased markedly in step with the frequency of childhood immunizations. This assumption is not obviously valid. You sometimes read that autism was unknown before the introduction of vaccines with thimerosal, but what happened in the forties was the introduction of the term autism, not the sudden appearance of withdrawn, affectless children. Many cases of what we would now call autism are described in the older medical literature, just as millions of people died from heart attacks over the years even though myocardial infarction was only defined in the 20th Century. Because the clinical definition of autism is so vague, it is perfectly possible that its increasing incidence reflects to some degree the tendency of doctors to make trendy diagnoses in borderline cases, an instance of that syndrome inflation or societal hypochondria familiar from the ADD saga. In this regard, I find it telling that follow up studies of children diagnosed as autistic in recent decades find that many of them go on to college and careers while the autistic kids of yesteryear are still banging their heads against a wall.

Because as a rule it is hard and expensive to figure things out, none of the questions about thimerosal are going to be answered quickly and anybody who issues conclusive statements about the matter is overstating his or her case. That said, I confess that I lost interest in the issue a couple of years ago because of a simple consideration. Thimerosal was removed from vaccines in Denmark in 1992 but the reported incidence of autism continued to rise sharply thereafter (Science, Vol 301, Issue 5639, 1454-1455, 12 September 2003). That finding doesn’t guarantee that the mercury in vaccines has no relationship whatsoever with autism, of course; but it certainly deflates the notion that thimerosal is the fundamental problem. The research also reminds us that a giant and crucial experiment is already underway in this country. If thimerosal really was the cause of a U.S. autism epidemic, the epidemic should be coming to an end shortly since vaccines no longer contain thimerosal. Since this vast if accidental test will much larger and more conclusive than any feasible planned study, I think it makes sense to reserve scarce research dollars to deal with other issues.

Note: for a rundown of information on thimerosal that includes links to sources on both sides of the issue, see this Wikipedia page.

Tuesday, July 19, 2005

Le Meme Chose

In Goethe’s play, Faust sells the Holy Roman Emperor on the idea of issuing paper money backed up by buried treasure since the emperor is entitled to such hordes by imperial law and the gold will turn up eventually. Naturally, the float sinks; but the brilliant absurd notion is no crazier than what various philosophers and scientists have promoted in earnest. Hume, as an impatient young man, advertised his initial epistemology as the application of Newtonian methods to the physics of ideas as if sense impressions formed little solar systems governed by simple laws of attraction. A.J. Ayer’s also proposed to cut all the Gordian knots at one stroke by assassinating traditional ethics and epistemology with a superficially plausible theory of meaning. This sort of thing can also be institutionalized. Some Sumerian grad student once got the bright idea of keeping track of what happened in Mesopotamian politics when comets appeared or the moon was eclipsed, but his program continues to this day as astrology. Learning theory in psychology can’t claim such a long run, but it is also based on an obvious methodological insight. All we’ve got to do is find the law that relates stimulus and response! Well, that’s the sort of thing that works when it works and, in any case, is always easy to explain, which counts for a lot.

Although every successful theory gets mechanically applied to every available problem, in most cases all that gets transferred is a set of buzzwords. That’s certainly true in relationship to the current fad for applying ideas drawn from evolutionary biology to discussions of the transmission and fate of ideas and institutions in human society—the meme meme. Since both living things and cultures change through time, it’s hardly surprising that the processes involved in their evolution are analogous. The trouble is, they are aren’t apparently homologous, i.e. the terms of the theory of natural selection—genes, genotypes, phenotypes, fitness—don’t match up one-to-one with the corresponding cultural terms, assuming there are corresponding terms. While genes prosper because the organisms to which they belong survive and reproduce better than organisms with other genes, the success of melodies, jokes, poems, slogans, gestures, gods, faiths, philosophies, and other cultural detritus seems to mostly depend merely on their tendency to be copied. Meanwhile, while it is possible to define what gets replicated in natural evolution—crudely speaking, stretches of nucleic acids—almost anything can and has been identified as a meme, though only a few candidates for memehood have anything like the specificity of an allele. It may be possible to model the fate of a catchy melody or an exactly repeated ritual by recourse to the mathematics of population genetics, but, as Scott Atran points out in his book In Gods We Trust, even something so apparently cut and dried as the Ten Commandments is impossible to specify. We’re not talking about a series of sounds in Hebrew or a series of letters in the KJV, after all, but the meaning of a form of words. In lieu of anything like the chemical formula that defines a gene or a protein, such entities are, as I like to say, boneless. Just think for a moment of the admirable flexibility of the Thou Shalt Not Kill provision.

Knowledge is limited by the intelligibility of its objects. Natural evolution is understandable because living things have genetic and developmental systems that drastically constrain how they change over time. They are good to think. The cultural systems that are best understood—languages, folk taxonomies, musical systems—are also apparently constrained by nature, which is to say human neural biology. Where such preexisting regularities are lacking, the will to know will simply create superstitions. I think that’s what’s happening with memology, which seems remarkably like the Intelligent Design movement in its endless production of programmatic statements and its utter lack of a research program.