Saturday, February 25, 2006

Genius germs?   posted by agnostic @ 2/25/2006 10:00:00 PM

I began this series at my other blog before I became a poster here. You could read this on its own, but the first four parts -- I here, II here, III here, IV here -- provide the necessary background (esp. part I, first 3 paragraphs of part II, and part III, all of which are short). Briefly, the idea was to investigate whether microbes could affect human cognition in ways more subtle than rabies. Now comes the empirical support I've uncovered: a strong winter-spring birth seasonality effect on "genius," which I take to reflect early infection. (Somewhat long read.)

We present evidence that early infection likely contributes to "genius" status -- recall from part II our definition of "genius" as anyone who had an Index Score (IS) of at least 50 (from 0-100) in the inventories of Charles Murray's Human Accomplishment. The prediction is that at the highest level -- the "giants," who score in any category's 10th "decile" (i.e., IS at least 90) -- the births will be the most lopsided toward winter-spring (WS = Dec-Feb and Mar-May), when infant infection is most likely; that the top 5 deciles will show less lopsidedeness though still toward WS; and that the bottom 5 deciles will not necessarily show the pattern. More, we predict that the more abstract the field (and thus the more it requires superhuman creativity), the more pronounced the bias. We first examine the giants, then the geniuses of the most abstract fields -- Philosophy in the humanities, Music in the arts, and Math in the sciences -- and finally the geniuses of the remaining fields. In our research, we found birth month data only for Westerners, which constrains the scope of the argument w.r.t. the arts, though no non-Westerner is among the geniuses of any science category. All lists of births available by email (see my bl*gspot profile).

First, in HA 18 Westerners scored 90 or above in any field, though only 16 figures had known birth months (unknown: Aristotle & Hippocrates). Of these 16 giants, 14 are WS: Galileo, Kepler, Darwin, Newton, Einstein, Euler, Pasteur, Koch, Edison, Watt, Beethoven, Mozart, Michelangelo, Shakespeare. Among these, 10 are winter, 4 spring. Only 2 of 16 are summer-fall (SF): Lavoisier and Lyell. The prediction checks out: 87.5% are WS, and the actual value of winter births is 2.5 times the expected value of 4.

Next, the most abstract art: Music. Homo sapiens' natural mode of expression is linguistic, and we can grope our way through visual modalities such as gesture, mime, etc. But we are utterly at a loss when it comes to non-linguistic sound. Moreover, Western music emphasizes both complex melodies, which are serial, as well as complex harmonies, where notes are stacked on top of one another. Juggling these elements for various instruments in one's head, all while attempting expression in the most foreign of artistic languages, is the greatest test of artistic genius. The Western Music inventory is particularly instructive since birth months are known for all figures save a neglible number down in the 2nd decile (2D: we use XD as short-hand for the Xth decile). There are 5 geniuses (people in 6D-10D), all 5 of whom are WS: Beethoven, Mozart, Bach, Wagner, Haydn. All graphs visible here.

Turning to the most abstract science, Math, there is little role for "inevitable discovery," and the objects it studies are farther removed from the real world than in other sciences. Moreover, we have less innate / intuitive scaffolding to hoist ourselves up by when it comes to math as compared to physics or biology. There are 8 geniuses, 1 of whose birth months is unknown (Euclid in 9D). Again, the hypothesis checks out, though not as strongly as in Music -- of the known 7, 5 are WS: Euler in 10D, Newton & Gauss in 9D, Descartes & Cantor in 6D. The 2 exceptions constitute 8D: Fermat & Leibniz.

As for Philosophy, it is the only humanities field Murray included, presumably because it is (at least for now) the only one demanding genius thought, fields like history being closer to (extremly important) clerical work. Again, we only found data for Westerners. There are 4 geniuses, 1 of whose birth months is unknown -- unfortunately, the top-ranked and only figure in 10D: Aristotle. We admit this weakness. Still, the hypothesis checks out -- of the known 3, all 3 are WS: Plato in 9D, Kant in 8D, Descartes in 6D.

To sum up so far: there are 17 geniuses in the three most abstract fields, and 15 of their birth months are known. 13 of these 15 (~87%) are WS, and the 2 exceptions show up in 8D in Math. So, the percentage of WS is almost identical to that among giants, though the gross overrepresenation of winter births is gone: of these 15, only 3 (or 4 -- Plato was born in either Dec or May) are winter, which, depending on Plato's birth, is either 0.25 above or 0.75 below the expectation of 3.75.

The link to the graphs begins w/ raw number and percentage of summer-fall births in the three most abstract fields; SF are shown in order to highlight data points that falsify the hypothesis. A dashed line indicates no data points in that decile; a red numeral indicates the number of points in the decile for which data were not found. Depending on the inventory, some of the lower deciles were not examined since 1) they did not bear on the hypothesis, 2) they had larger numbers that would have required more hard labor to collect, and 3) these are the people most likely to fluctuate in and out of the inventory depending on which encyclopedias are consulted (unlike, e.g., Mozart or Newton).

Next is a graph of the other arts, Western Art and Western Literature. In Art, there are 5 geniuses, 1 of whose birth months is unknown (Titian in 6D). Of the known 4, 3 are WS: Michelangelo in 10D, Raphael in 7D, Leonardo in 6D. The 1 exception, in 7D, is one of the few people widely considered by reviewers of HA to be an epochcentric anomaly (see, e.g., this review by Denis Dutton): Picasso, who scores two deciles above Dürer, Rembrandt, Giotto, Bernini, Cezanne, & Rubens (5D). The least abstract art, Lit, we predict to be least lopsided toward WS since Murray explains that creators of Lit encyclopedias strongly consider the role the writer played in social movements, since writing is used not only for expression but for communication and persuasion. (The Art and Music inventories do not show such an effect.) E.g., for his political impact, Rousseau (in 5D) ranks two deciles above inter alia Aeschylus, Ovid, Whitman, & Proust. There are 5 geniuses, 1 of whose birth months is unknown (Homer in 6D). Of the known 4, only 1 is WS: Shakespeare, the lone figure in 10D. As the graph shows, Lit is the most SF-friendly art, as predicted.

We continue w/ a graph of the Combined Sciences. Murray only listed the top 20 figures, though his point was only to illustrate the "big fish in a small pond" effect for figures such as Lyell who dominate their small pond but don't show up in the larger pond. There are 6 geniuses, 1 of whose birth months is unknown (Aristotle in 8D). Of the 5 known, 4 are WS: Newton in 10D, Galileo in 9D, Kepler & Descartes in 6D. The 1 exception is Lavoisier in 6D. So considering the sciences as a whole, the prediction is met.

Considering each scientific pond, we turn next to the three most established sciences after Math: Physics, Chemistry, and Astronomy. In Physics, there are 9 geniuses, of whom only 4 are WS, though these include the two in 10D -- Newton & Einstein -- along w/ Galileo in 9D and Thomson in 6D. The exceptions are Rutherford & Faraday in 9D and Cavendish, Bohr, & Maxwell in 6D. In Chemistry, there are just 3 geniuses, 1 of whom is WS: Scheele in 6D. The 2 exceptions are Lavoisier in 10D and Berzelius in 7D. In Astronomy, there are 9 geniuses, 1 of whose birth months is unknown (Ptolemy in 8D). Of the 8 known, 5 are WS: Galileo & Kepler in 10D, Laplace & Copernicus in 8D, and Brahe in 6D. The 3 exceptions are Herschel in 9D, and Halley & Cassini in 6D. So, by zooming in closer on each pond, the actual value of WS is 0.5 below expectation (10 of 21 in these three fields), which weakens the hypothesis. Alternatively, sub-giant-level insight in these sciences may not require as much "outside the box" creativity as does giant-level insight in these sciences, or as does sub-giant insight in the arts.

Now we consider the two least abstract sciences (as of 1950, when Murray's survey ended): Earth Sciences and Biology. In Earth Sciences, there are 4 geniuses, 2 of whom are WS: William Smith & Agricola (Georg Bauer) in 6D. The 2 exceptions are Lyell in 10D and Hutton in 8D. This matches the prediction that the least abstract field will not show lopsidedness toward WS (as w/ Lit). Biology is another case in point, since by 1950 it was barely established as a field and not very abstract at that. There are 7 geniuses, 2 of whose birth months are unknown (Aristotle in 10D and Harvey in 6D). Of the known 5, just 2 are WS: Darwin in 10D and Linnaeus in 6D. The 3 exceptions are Lamarck in 10D, Cuvier in 9D, and Morgan in 8D.

None of the geniuses who developed biology into a mature science during the 20th Century even made it to 4D, again because Murray's encyclopedias focused on periods before 1950. For example, Darwin the Second -- Bill Hamilton -- is not included at all, while R.A. Fisher barely shows up in 1D. Though it is too early to provide a ranking of who encyclopedias 200 years from now will consider the equivalents of Newton and Rutherford, we can at least come up w/ an unordered list of newcomers and their birth seasons: Hamilton (sum), Haldane (fall), Fisher (win), Wright (win), Smith (win), Trivers (win). This is not definitive, but on the right track.

Haldane is clearly an exception, and though Hamilton's data point may appear to falsify the hypothesis of higher likelihood of infant infection among geniuses, the fuller story is revealing. Unlike every other European genius, he was born in Cairo, Egypt's most crowded, slum-ridden urban area, in Aug 1936 -- less than 4 months after Egypt had even established a Ministry of Health! More, his mother was a medical doctor, who would've been exposed to god knows what in those days and potentially have brought it home. In this sole case, we consider geographical location to be more informative than WS birth in assessing likelihood of infant infection. The geniuses in other categories show a general bias toward urban birth, but the effect is not as strong as birth month, and it is susceptible to alternative interpretations. We return to this point later.

Finally, we consider the two applied sciences: Technology and Medicine. While Technology is by definition creative and inventive, we are agnostic on whether it requires the genius of Beethoven or Newton; and Medicine is largely discovery, not creative model-building. Nevertheless, next is a graph for these two. In Technology, there are 6 geniuses, 1 of whose birth months is unknown (Archimedes in 6D). All 5 of the known are WS: Edison, Watt Leonardo, Huygens, Marconi. In Medicine, there are 8 geniuses, 3 of whose birth months are unknown (Hippocrates in 9D, Galen in 8D, and Paracelsus in 7D). Of the 5 known, all 5 are WS: Pasteur in 10D, Koch in 9D, and Ehrlich, Laennec, & McCollum in 6D. We interpret these data as not falsifying the hypothesis, though hardly a ringing endorsement of it, given the conceptual nature of the fields, their relative immaturity up to 1950, and the lacunae among the data for Medicine.

So, overall the hypothesis passes the tests for finding lopsided seasonality among genius births in abstract, creative fields, as we believe the Combined Sciences ranking better highlights scientific genius than the rankings in the separate fields which compose it. This was especially so for the mostly-winter "giants." But how do we interpret this finding? First, imagine we examined another trait w/ 0.5 probability of occuring in the general population -- say, male vs female sex. If we observed a similarly lopsided male to female ratio, we would need to account for it somehow: sex discrimination, different distributions in cognitive ability, a mix, etc. We find it implausible that social factors contribute to the seasonality of genius births: there is no evidence that WS children are encouraged more, that SF suffer Zodiac "stereotype threat," that either of these would make such a difference in magnitude anyway, and so on. In epidemiological studies, seasonality of births is typically taken to reflect the role of infection, as it cannot be easily confounded w/ other variables, unlike the effect of urban birth -- the latter could reflect selection bias for higher IQ, class structure, better access to mentors, and so forth. But the only powerful, non-magical explanation for seasonality is infection.

In principle, WS seasonality could also reflect, e.g., lack of sun exposure and thus lack of vitamin D. But unlike vitamins, microbes are alive & evolving, meaning their presence (or absence) can have either positive or negative effects, depending on whether they are mutualist or parasitic. We cannot easily conceive of how lack of vitamin D would help smart people become singular geniuses, so we find subtle microbial influence much more plausible. Indeed, a recent study done to assess seasonality of schizophrenic births also found higher cognitive development among normal WS children, as measured by various psychometric tests, though the data do not report adult IQ, which would be more noteworthy.

Now, are we saying that early infection is all there is to genius? No, because we already know from Behavior Genetics that in adulthood, the broad-sense heritability of g is ~0.7, so genes certainly play a crucial role as well, not to mention access to mentors, etc. Recall that in the lower 5 deciles of the inventories, there was no apparent seasonality, so this infection likely plays a role in a tiny minority of cases indeed -- though these are the most impressive of cases -- and again we assume the affected individuals already had a high IQ due to additive genetic effects. But since we do not know exactly what the germs are, we cannot tell what effect they would have on an average or below-average intellect; in principle, it could go either way. Therefore if the germs were identified, administering them in the hopes of turning one's child into the next Mozart would almost certainly fail, since presumably many more individuals were infected in addition to Mozart, Newton, et al. Now, if the parents had good reason to believe their child's IQ would already be quite high, the prospect would be more promising.

Where, then, does this leave us as far as exploring the cells in the "brain germ" matrix outlined in part III? We are utterly clueless as to the route, aside from knowing that it must begin early after birth, and we are also unsure of its impact of reproductive fitness. We could not easily locate data for average family size in the times and places that produced the "giants," but here are the numbers of children sired by each of the 14 WS giants: 0 (Newton, Beethoven, Michelangelo), 1 (Koch), 3 (Galileo, Kepler, Einstein, Shakespeare), 5 (Pasteur), 6 (Edison, Watt, Mozart), 10 (Darwin), and 13 for the man whose genetic output was second only to his mathematical (Euler). How these actual values compare to the expected values given the time & place in which they flourished, we leave open for now.

So if early infection is one piece of the puzzle behind Galilean excellence, might better hygiene play a role in the decline of the per capita *rate* of accomplishment that Murray wrestles w/ in HA (Ch. 21), which accelerated downward after 1800? Part of his argument is that the secularization of Europe left each generation after ~1800 w/ less motivation to pursue their calling in life. As in our discussion in Part IV of Judith Rich Harris' personality model, we don't discount social influences such as the ones Murray mentions. However, in Murray's own list of "Significant Events" (Ch. 9) for Medicine, he has boldfaced the entry of 1796 to underscore its importance: "Edward Jenner systematizes vaccination for smallpox, founding immunology" (p.194; original emphasis).

During the 19th Century, the scientists Paul Ewald calls "the microbe hunters" in his book Plague Time began searching out infectious causes for diseases and proposing cures or preventative measures, including Semmelweis' efforts to introduce rigorous hygiene among doctors who were delivering newborns in order to cut down neonatal mortality rates, not to mention Pasteur's establishment of the germ theory of disease and Koch's formulation of Koch's Postulates to determine infectious origin. Conversely, we interpret the increasing rate of accomplishment up to and shortly after the Renaissance to reflect in part the increasingly frequent exposure to microbes as a result of urbanization. Similar reasoning suggests a partial reason for why advanced civilizations produce more geniuses than hunter-gatherers.

The early immunological efforts and their present-day descendents have surely improved the quality of life for the average person born in Western nations. Yet they might also have contributed to the decline in the rate of genius-level accomplishment. As elsewhere, science can only illuminate a trade-off -- if our interpreation is correct, in this case between level of public health and rate of genius-level excellence -- and the value judgment of where to resolve the trade-off is ultimately up to the individual or the society, not scientists. More big thinkers are better than fewer, ceteris paribus, but few will accept a larger percentage of geniuses if it requires diminishing the effect of public health on quality of life for the average person. Still, we feel a certain optimism is in order: after all, the Industrial and Information Revolutions took off when the rate of accomplishment was declining, suggesting that what matters on a day-to-day level -- e.g., having electricity so you can listen to a Chopin CD -- is more a function of the raw number of big thinkers rather than their proportion of the general population.

Epilogue: Psychology. I can hardly drone on about genius cognition w/o mentioning the geniuses of Psychology. Now, the field is far too immature to show up in Murray's survey, but I came up w/ the tiny handful of psychologists I estimate will compose the top 5 deciles in the 200th anniversary edition of HA. There are two main groups in psychology: those that deal w/ universals and those that deal w/ individual differences. From the former, I'd wager that Francis Galton (win) will make it. From the latter, I'd wager that William James (win) will make it, and perhaps the founders of the two fields of cognitive psychology that are best understood: language pioneer Noam Chomsky (win) and vision pioneer David Marr (win). Just as I assume Cuvier will fall from the top 5 deciles in Biology once Hamilton & Fisher work their way into the encyclopedias, I assume the following will drop from their (even now decreasing) Deity status: Freud (win), Skinner (spr), & Piaget (sum).