How to defend the indefensible: a comment on Gottfredson (2013)

In a recent hagiography about prominent "race realist" psychologist J. Philippe Rushton (who died in 2012), Linda Gottfredson of the University of Delaware claims that "Rushton is a scholar and gentleman but it appears that his critics often act like neither" (Gottfredson 2013). She supports this claim by discussing what happened when Rushton's work was criticized in an article published in a respected journal, arguing that the resulting controversy illustrates how "mob science works to ‘‘discredit’’ valid research and enforce collective ignorance about entire bodies of evidence." Here, I argue that Gottfredson is full of shit, and that, in fact, the issue she is discussing demonstrates academia working as intended: critics noting methodological flaws in an academic's research that lead to inaccurate conclusions.

When I first read the abstract of this article, I was a little surprised, because Gottfredson seemed to be defending someone whose research had been discredited on methodological grounds, rather than ad hominem attacks of being "racist" (for a few of many examples, see Weizmann et al. 1990, 1991; Lynn 1989a, 1989b; Silverman, 1990; Zuckerman 1988, 1990; Cain & Vanderwolf 1990, Anderson 1991). So I kept reading a full copy of the paper, and this is my response to it.

The criticism in question is Lieberman (2001), an article which states the following early on: "[Samuel George] Morton collected human skulls, measured their cranial vaults, and concluded that “Caucasoids” had the largest brains and “Negroids” the smallest, with “Mongoloids” in between (Morton 1849). A century and a half later this hierarchy would be altered by J. Philippe Rushton and colleagues, placing the “Mongoloids” in the alpha position, “Caucasoids” next, and “Negroids” last. Rushton went on to correlate brain size with IQ scores, claim a Mongoloid 1 Caucasoid 1 Negroid correlation, and use variation in IQ scores to “explain” everything from civilization to barbarism (1997a). Although this view has been invalidated by a century of anthropological research and theorizing stimulated by Franz Boas (Gossett 1965, Cravens 1978), Rushton (1996) dismisses this work as no more than political correctness."

The point being made here is that race science does not conform to scientific objectivity or to the actual conclusions that are best supported by evidence, but instead to a political/social agenda of justifying oppressing a given group of people. Lieberman illustrates this in a later section of his paper, entitled "Changing Hierarchical Worldviews". He subsequently notes that in the 1970s, researchers like Richard Lynn started publishing studies arguing that, contrary to mainstream thinking at the time, Japanese people were smarter than Caucasians, and that "The publication of these theories of East Asian superiority was preceded by Japan's becoming a “world-class economic power”" (p. 72).

Here is how Gottfredson summarized what you just read (as well as a bunch of other parts of Lieberman's paper I did not discuss above):
"Lieberman opens with a question that itself damns Rushton. A century of anthropological work has invalidated Rushton’s claims, so how can he claim to find in it a ‘‘racial hierarchy’’ for intelligence and brain size? The question is thus not whether Rushton is wrong, but why and how he persists in being so wrong. The article’s first section (‘‘Changing Hierarchical Worldviews’’) justifies the premise, and the second (‘‘Abusing Anthropological Research’’) explains the ‘‘paradox’’ of how Rushton and other ‘‘scientific racists’’ could claim to be doing science when they draw evidence from the very fields that disavow racism and the concept of race (p. 74). To justify his premise, Lieberman describes 19th century research on cranial size and its social context which, he says, was the need by Caucasians to justify their domination and exploitation of other races. He discredits that research and Rushton’s own primarily by appealing to authority in 20th century anthropology: Franz Boas’s theorizing (no link between culture and genes), official statements on race from the UN and the American Anthropological Association (no biological races, no meaningful innate racial differences), Gould’s critiques of research on intelligence, brain size, and heritability (none is valid), and anthropology’s ‘‘disavowal’’ of ‘‘hierarchical’’ and ‘‘racist’’ thinking. Lieberman also draws from stock concerns, long-since resolved, about possible methodological flaws in twin research, brain research, behavior genetics, and mental testing. He says nothing about the explosion of research in the 1990s using the Y chromosome and mtDNA to trace the evolution of human lineages as they migrated across the globe. He says nothing about Rushton’s many other 3-way results on ‘‘reproductive behavior, sex hormones, twinning rate, speed of physical maturation, personality, family stability, lawabidingness, and social organization’’ (p. 74), except to summarily dismiss them as a faulty ‘‘blizzard of data’’ (p. 78)."

So her first main point seems to be that Lieberman is relying on logical fallacies to criticize Rushton, the first of these apparently being appeal to authority by citing the extensive work of Boas, the AAA, and Gould. Well, what Lieberman cited to support this claim on the first page of his paper were two sources: a 1965 book (Gossett 1965) and a 1978 book (Cravens 1978). So he's not just invoking Boas's name as a talisman of some sort to dismiss inconvenient arguments, but instead is citing evidence to support them. It is notable that, in contrast, Gottfredson does not cite any sources to support her claim about how the concerns Lieberman raises are "long-since resolved". I would also like to note that while Lieberman may not have addressed every trait Rushton studied, other researchers have; for instance, Weizmann et al. (1990) concluded that "The predictions that Rushton derives from the r/K model are arbitrary, and these predictions are supported by selective citation and misrepresentation of the research literature and by the use of unreliable sources. Changes in human life-history traits are so rapid that there is no need to posit genetic selection to explain intergroup variation." And with regard to Rushton's results on race differences in reproductive traits, Lynn (1989a) noted the following four problems in a paper Rushton co-authored on the subject in 1987: "First, they did not explain why natural selection would have favored different reproductive strategies for different races. 
Second, their data on race differences are of questionable validity because their literature review was selective and their original analyses were based on self-reports. 
Third, they provided no evidence that these race differences had significant effects on reproduction or that sexual restraint is a K characteristic. 
Finally, they did not adequately rule out environmental explanations for their data."

Back to Gottfredson (p. 4): "Zeroing in on Rushton’s IQ, race, and brain size analyses, Lieberman details his own list of 6 major ‘‘errors.’’ Briefly, Rushton ‘‘uses ‘race’ despite decades of findings that invalidate it,’’ his conclusions about racial differences in cranial capacity are ‘‘contradicted by evolutionary anthropology,’’ he did not account for environmental factors that surely influence cranial capacity and intelligence, his measurements tell us nothing because they are confounded or the differences they reveal are trivial in size, he cannot claim to ‘‘explain’’ a vast array of human behaviors because some of his measures and concepts may be faulty, and his ‘‘principle of aggregation’’ (e.g., grouping diverse populations into ‘‘races,’’ averaging results from different studies) is invalid. ‘"

She then dismisses the first two of these errors as mere appeals to authority. That being said, there is compelling evidence against the biological or genetic existence of race (Templeton, 2013; Long & Kittles, 2009). Moreover, though it may have been phrased as an appeal to authority, the assertion that Rushton's conclusions of race differences in cranial capacity is inconsistent with anthropological evidence is not an appeal to authority, nor is it any other logical fallacy. As Lieberman goes on to note, Rushton took his cranial capacity measurements from a 1984 study without mentioning that this study showed no significant relationship between race and cranial capacity, but did find that "climate variables were strongly correlated with cranial variation" (Lieberman 2001, p. 76). Other research supports this claim, and also suggests that cranial size varies with the development status of a country, not with race (Cernovsky 1993). Gottfredson is, therefore, being disingenuous in not mentioning the sources on which Lieberman's argument is based.

The next two claims Gottfredson makes are that Lieberman's 3rd and 4th points "disallow drawing conclusions until an infinite regress of alternatives has been considered" (p. 4). Those points are, respectively, that 3) "[Rushton] did not account for environmental factors that surely influence cranial capacity and intelligence" and 4) "his measurements tell us nothing because they are confounded or the differences they reveal are trivial in size" (Gottfredson, p. 4). In fact, though, these are both valid criticisms of Rushton's work in this area. Environmental factors like climate and development status have been discussed above, but there are other confounders as well that call his findings into doubt. As Lieberman noted, the following factors had been previously identified as affecting brain size, and therefore needing to be controlled for if you want to determine whether there's a real difference between groups: "sex, body size, age at death, early-life nutrition, early-life environment, source of sample, occupational group, cause of death, lapse of time after death, temperature after death, anatomical level of severance, presence or absence of cerebral spinal fluid, presence or absence of meninges, and presence or absence of blood vessels" (p. 77). Also, Rushton's claims about race differences in brain size are contradicted by some of the studies he cites (Cernovsky 1995).

So what about Lieberman's last two points? To recap, they were that "[Rushton] cannot claim to ‘‘explain’’ a vast array of human behaviors because some of his measures and concepts may be faulty, and his ‘‘principle of aggregation’’ (e.g., grouping diverse populations into ‘‘races,’’ averaging results from different studies) is invalid" (Gottfredson, p. 4). Gottfredson claims that "the last two [of Lieberman's arguments] demand uniformly perfect data and measurement before concluding anything from a body of evidence" (pp. 4-5). Again, not so: All that Lieberman pointed out was that Rushton did not consider yet another confounding factor: nutrition. This is explained in Lieberman's fifth point, which was that "Rushton does not relate environment, nutrition, cranial size, and IQ" (p. 78). Gottfredson, weirdly, ignores this point while going after his sixth point, namely that "Rushton claims to “explain” a vast array of human behavior" (p. 79). 

What's wrong with him doing this? I'll give you a hint: it's not that his data and measurements aren't perfect. The problems include that he's glossing over differences that contradict his theory by lumping different groups together into racial categories by only using averages. Rushton's doing so also ignores the larger differences in brain size across gender than across "race", and the genders of his skulls were unknown (Cernovsky 1993). Lieberman also notes that "Rushton ranks “races” on a number of variables including cultural achievements, personality traits, marital stability, law-abidingness, mental health, and administrative capacity. These variables are usually not defined, and each represents reified aggregations of diverse behaviors that vary in their causation" (p. 79).

Naturally, Gottfredson couldn't resist ending her paper with this gem: "...it is Gould’s work on cranial capacity, not Rushton’s, that we now learn was fudged and falsified (Lewis et al., 2011)—just as Rushton said it was." Lewis et al., of course, is a 2011 study of Morton's skulls which concluded that "Morton did not manipulate data to support his preconceptions, contra Gould." Of course, she omits criticism of Lewis et al. from sources like Weisberg & Paul (2016) who wrote that "We take no issue with Lewis et al.’s remeasurements, but argue that these measurements are not and cannot be evidence for their conclusion". Other researchers have argued that Lewis et al.'s defense of Morton is inadequate, and that "Gould was right to reject Morton's analysis as inappropriate and misleading, but wrong to believe that a more appropriate analysis was available" (Kaplan et al. 2015).

Addendum: Rushton's research on international variations in crime, which Gottfredson briefly mentions in her paper, is criticized by Neapolitan (1998) for not controlling for non-race factors.
Sources
The paper this post is about (Gottfredson 2013) is here: http://www1.udel.edu/educ/gottfredson/reprints/2012RaceandRushton.pdf
Anderson 1991: http://philipperushton.net/wp-content/uploads/2015/02/race-r-k-theory-rushton-anderson-canadian-psychology-1-1991.pdf
Cain & Vanderwolf 1990: http://www.sciencedirect.com/science/article/pii/019188699090185T
Cernovsky 1993: http://philipperushton.net/wp-content/uploads/2015/02/iq-race-brain-size-rushton-cernovsky-j-of-black-psychology-1993.pdf
Cernovsky 1995: http://journals.sagepub.com/doi/abs/10.1177/002193479502500602?journalCode=jbsa
Cravens 1978: The triumph of evolution: American scientists and the heredity-environment controversy 1900–1941.
Gossett 1965: Race: The history of an idea in America.
Kaplan et al. 2015: http://www.sciencedirect.com/science/article/pii/S1369848615000035
Lewis et al. 2011: http://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.1001071
Long & Kittles 2009: http://www.bioone.org/doi/abs/10.3378/027.081.0621
Lynn 1989a: http://www.sciencedirect.com/science/article/pii/0092656689900299
Lynn 1989b: http://www.sciencedirect.com/science/article/pii/0092656689900317
Neapolitan 1998: http://onlinelibrary.wiley.com/doi/10.1111/j.1745-9125.1998.tb01243.x/full
Silverman 1990: http://www.sciencedirect.com/science/article/pii/016230959090002N
Templeton 2013: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3737365/
Weisberg & Paul 2016: http://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.1002444
Weizmann et al. 1990: http://psycnet.apa.org/journals/cap/31/1/1/
Weizmann et al. 1991: http://psycnet.apa.org/index.cfm?fa=buy.optionToBuy&id=1991-20580-001
Zuckerman 1988: http://www.sciencedirect.com/science/article/pii/0191886988901365
Zuckerman 1990: http://psycnet.apa.org/psycinfo/1991-12414-001


http://journals.sagepub.com/doi/abs/10.1177/00957984930193004
http://journals.sagepub.com/doi/abs/10.2466/pms.109.3.733-736?journalCode=pmsb
https://link.springer.com/article/10.1007/BF02088001