Human Diversity

by Charles Murray

  32. E,g,, Kahneman, Slovic, and Tversky (1982) and Thaler and Sunstein (2008).

  33. Harden, Domingue, Belsky et al. (2019).

  34. “On the bottom line, a variable that has no within-family variation cannot covary with the within-family variation in another variable.” Turkheimer, D’Onofrio, Maes et al. (2005): 1226.

  35. Trzaskowski and Plomin (2015): 9.

  36. Population stratification does not change the situation. Population stratification can cause major errors in interpretation of the polygenic scores, but it’s irrelevant to backward causation.

  37. The subsequent discussion and direct quotes are based on Eric Turkheimer, “The Blueprint Metaphor,” GHA Project: Turkheimer’s Projects: Genetics and Human Agency, October 30, 2018, geneticshumanagency.org.

  38. I reviewed the evidence on G×E interactions at early ages in chapter 13, but I did not give a full account of the degree to which many psychologists believe that G×E interactions are pervasive and, once understood, will fundamentally alter our conception of the heritability of traits. Many expressions of this position are akin to heritability denial—for example, Burt and Simons (2014), discussed in note 26 for chapter 10. A knowledgeable and nuanced expression of this position, combining both theory and empirical examples, is Sauce and Matzel (2018).

  39. Methylation changes during the lifespan come into the picture here, but what we know to date does not give reason to think that they will prove to be a major confounding factor. See the discussion of epigenetics in chapter 13.

  40. I owe the idea for this search to Steven Hsu, who conducted a similar search for his blog, Pessimism of the Intellect, Optimism of the Will (infoproc.blogspot.com), May 25, 2019.

  41. Plomin (2018): 172.

  15: Reflections and Speculations

  1. Pinker (2002): 340.

  2. The concept of the mind as initially a blank surface that experience writes upon goes back to the Greeks and has a long pedigree thereafter. But Locke’s formulation in An Essay Concerning Human Understanding has been the main source for the modern era. Locke actually used the phrase “white paper” (“Let us suppose the mind to be, as we say, white paper, void of all characters, without any ideas”). Book II, chapter 1.

  3. Pinker (2002): 5.

  4. Smith (1776): vol. I, chapter 2. Smith (1979).

  5. Hamilton, Madison, and Jay (1982): No. 51.

  6. Leon Trotsky, “Socialism Will Bring Giant Advances for Mankind,” The Militant 5, no. 34 (1924): 5.

  7. Durkheim (1982): 33.

  8. Durkheim, The Rules of Sociological Method, quoted in Tooby and Cosmides (1992): 24–25.

  9. E.g., Watson (1914); Skinner (1938).

  10. Watson (1924), quoted in Pinker (2002): 19.

  11. Contrary to rumor, the Skinner box was not used to experiment with operant conditioning on Skinner’s infant daughter, though Skinner did invent an “air crib” that was intended to reduce the tasks of caring for an infant.

  12. On a personal note, my coauthor on The Bell Curve, Richard Herrnstein, was a behavioral psychologist who succeeded B. F. Skinner as the Edgar Pierce Professor of Psychology at Harvard. As you will find if you read The Bell Curve or Crime and Human Nature, coauthored with James Q. Wilson, Herrnstein was a behaviorist who did not go off the deep end.

  13. Murray (1984).

  14. Pinker (2002): viii.

  15. Hamilton (1964); Trivers (1972).

  16. Wilson (1975); Dawkins (1976).

  17. Tooby and Cosmides (1992). The book in which it appeared, Barkow, Cosmides, and Tooby (1992), is considered a foundational text for evolutionary biology.

  18. Stewart-Williams (2018): Appendix A.

  19. von Hippel and Buss (2017).

  20. For a systematic discussion of the science of evolutionary psychology, see Low (2015). For two technical book-length accounts of the sociobiology controversy presenting the arguments on both sides, see Segerstråle (2000) and Alcock (2001).

  21. Stewart-Williams (2018): 291. My characterization of the attacks on evolutionary psychology are a condensation of Steve Stewart-Williams’s presentation in Appendix 1.

  22. Stewart-Williams (2018): 292.

  23. Spiro (1954).

  24. Beit-Hallahmi (1981).

  25. Technically, it applies only to colleges and universities that accept federal funds. But you can count the ones that don’t on the fingers of one hand.

  26. For greater male attraction to team sports from childhood onward, see Lever (1978); Sandberg and Meyer-Bahlburg (1994); and Deaner, Geary, Puts et al. (2012). For cross-national data, see Deaner and Smith (2012).

  27. For rough-and-tumble play, see DiPietro (1981) and Pellegrini (2007). For sex-typical competition and exposure to testosterone in utero, see Hines and Kaufman (1994). For sex differences in aggression, see Card, Stucky, Sawalani et al. (2008). Byrnes, Miller, and Schafer (1997) is a meta-analysis of the literature on risk-taking as of 1997. Subsequent work includes Zuckerman and Kuhlman (2000) and Morrongiello and Dawber (2004). For the effects of testosterone on an increase in risk-taking in women, see van Honk, Schutter, Hermans et al. (2004).

  28. Lewontin (1970).

  29. Jencks (1979).

  30. My conjecture was inspired by Cheverud’s conjecture that genetic correlations and phenotypic correlations are similar. Cheverud (1984); Cheverud (1988). Subsequent research (e.g., Dochtermann (2011); Sodini, Kemper, Wray et al. (2018)) indicates that the conjecture is usually correct.

  31. Murray (2003): 275–83. From 1400 to 1800 CE, the only famous Jewish figure living as a Jew was Spinoza (Montaigne had a Jewish mother but was a lifelong Catholic).

  32. There are exceptions to the slow pace of change in milieu. For example, even though the sexual revolution in America was brewing throughout the twentieth century, there was a gaping discontinuity in the course of a few years in the mid-1960s.

  33. Murray (1984): x.

  34. For AA, see Kelly and Yeterian (2011); for KIPP, see Angrist, Dynarski, Kane et al. (2012); for Success Academy, see Unterman (2017).

  35. Herrnstein and Murray (1994): 535. The subsequent discussion draws directly from the concluding pages of The Bell Curve.

  36. Murray (2006).

  37. Walter Lippmann, “The Great Confusion,” New Republic, January 3, 1923: 46.

  Appendix 2: Sexual Dimorphism in Humans

  1. Fairbairn (2016): 105.

  2. Kodric-Brown and Brown (1987). For a nontechnical account, see Paco Garcia-Gonzalez, Damian Dowling, and Magdalena Nystrand, “Male, Female—Ah, What’s the Difference?,” The Conversation, March 26, 2013, theconversation.com.

  3. Frank Newport, “Americans Greatly Overestimate Percent Gay, Lesbian in U.S.,” Gallup News, May 21, 2015, galluppoll.com.

  4. I adapted this formulation from Sax (2002).

  5. This definition is taken from Blackless, Charuvastra, Derryck et al. (2000).

  6. Khalid, Oerton, Dezateux et al. (2012) found incomplete masculinization in 2 out of 33 male cases diagnosed at birth. For 44 female cases, 33 presented with virilization of female genitalia.

  7. All the percentages in the table are taken unchanged from Blackless, Charuvastra, Derryck et al. (2000) to avoid needless arguments. These additional points should be noted: Classic CAH. Blackless, Charuvastra, Derryck et al. (2000): Table 8 puts the incidence of classic CAH for females and males combined at 0.0077 percent (1 in about 13,000). A 2012 epidemiological study of classic CAH in Great Britain found a smaller incidence of 0.0055 percent for males and females combined (1 in about 18,000). Khalid, Oerton, Dezateux et al. (2012). Forty-three percent of the cases were male. Androgen insensitivity syndrome. The Blackless study puts incidence at .0084 percent (about 1 in 12,000) for all forms combined. Subsequent national epidemiological studies in Denmark (Berglund, Johannsen, Stochholm et al. (2016)), and the Netherlands (Boehmer, Brüggenwirth, van Assendelft et al. (2001)) found incidence rates of 0.0064 percent (about 1 in 15,600) and 0.0010 percent (1 in 10,000) respective
ly.

  8. Although the authors searched for incidence studies worldwide, they caution that because the bulk of the studies were from Europe and North America, their generalizations hold only for a “generic Euro-American, Caucasian population.” Blackless, Charuvastra, Derryck et al. (2000): 159. The 1.728 percent estimate of the incidence of intersexuality in Blackless was given prominence in Fausto-Sterling (2000) and attracted considerable media attention.

  9. Blackless, Charuvastra, Derryck et al. (2000): 152. The original text includes two references that are omitted here.

  10. Ranke and Saenger (2001). Hyperfemininity is such a common symptom of XXX women that the identification of a lesbian woman with Turner syndrome warranted a note in a technical journal. Fishbain and Vilasuso (1980).

  11. Otter, Schrander-Stumpel, and Curfs (2010).

  12. Sax (2002).

  13. About 7–8 percent of persons with vaginal agenesis also have AIS, but they would be classified as intersex under the AIS criterion. Foley and Morley (1992).

  14. Carmina, Dewailly, Escobar-Morreale et al. (2017).

  15. Speiser, Knochenhauer, D’ewailly et al. (2000).

  16. Moran, Azziz, Carmina et al. (2000); Witchel and Azziz (2010).

  17. Carmina, Dewailly, Escobar-Morreale et al. (2017): 12.

  18. Witchel and Azziz (2010).

  19. As in the case of hermaphrodites, classic CAH, and AIS, the incidence rates for the other departures from the Platonic ideal of sexual dimorphism are as reported in Blackless, Charuvastra, Derryck et al. (2000): Table 8. Subsequent research gives reason to think the estimate of 1.5 percent is too high. Carmina, Dewailly, Escobar-Morreale et al. (2017), a systematic literature review and analysis published 17 years after the Blackless study, studied late-onset CAH women who presented symptoms of androgen excess. They estimated a worldwide prevalence of 4.2 percent among such women. Azziz, Carmina, Dewailly et al. (2009) estimated that ∼10 percent of women are affected by hyperandrogenism. The authors declined to extrapolate an estimate of prevalence of late-onset CAH from those numbers both because of the imprecision of the 10 percent prevalence rate and because of the uncertainty about the numbers of asymptomatic women. At a minimum, it seems unlikely that the prevalence of symptomatic late-onset CAH as a percentage of the female population could reach 1.5 percent.

  20. Sax (2002): 177.

  21. Gates (2011): Fig. 1.

  22. Bailey, Vasey, Diamond et al. (2016).

  23. Sources: Frank Newport, “In U.S., Estimate of LGBT Population Rises to 4.5%,” Gallup News, May 22, 2018; Tables from the National Health Interview Survey 2015 on the survey’s website, www.cdc.gov/nchs/nhis; author’s analysis, General Social Survey 2016.

  24. Studies were limited to those that claimed nationally representative samples of adults, usually ages 18 and above, and obtained self-identified sexual orientation limited to the categories of gay/lesbian, bisexual, and heterosexual. Percentages are based on persons who chose one of those categories. I omitted studies where more than a few percent of the answers were “other” or more than a few percent of the participants refused to answer (because large numbers of such responses might indicate people who were homosexual or bisexual but didn’t want to say so). The sources were: Norway: Gates (2011); UK: Sexual Orientation dataset available at the Office for National Statistics website, www.ons.gov.uk; Canada: Canadian Community Health Survey, Cycle 2.1, available at the Statistics Canada webside, www.stat.can.gc.ca; Australia: Richters, Altman, Badcock et al. (2014): Table 1; New Zealand: Greaves, Barlow, Lee et al. (2017).

  25. Gates (2011). The Williams Institute is a think tank attached to the UCLA Law School that is devoted to law and public policy regarding sexual orientation and gender identity.

  26. Bailey, Vasey, Diamond et al. (2016): 53.

  27. Gates (2011): Fig. 4.

  28. Savin-Williams and Vrangalova (2013): 60.

  29. Savin-Williams and Vrangalova (2013): 59.

  30. Bailey, Vasey, Diamond et al. (2016): Fig. 1.

  31. Bailey (2009).

  32. Savin-Williams and Cohen (2018).

  33. Savin-Williams and Cohen (2018): 197.

  34. For an overview of scholarly opinion on the varieties of transsexualism, see Blanchard (2008) and Bailey and Blanchard (2017).

  35. Bailey, Vasey, Diamond et al. (2016).

  36. Bailey, Vasey, Diamond et al. (2016).

  37. The publication of Littman (2018) in August 2018 caused an uproar among transgender advocates. See Meredith Wadman, “New Paper Ignites Storm over Whether Teens Experience ‘Rapid Onset’ of Transgender Identity,” Science, August 30, 2018, www.sciencemag.org/news. The protests led the online journal in which it was published, PLoS ONE, to conduct a post-publication review and require Littman to issue a correction, posted on March 19, 2019, that emphasized the study was based on parents’ reports and was intended as a hypothesis-generating study. She was also required to include an expanded discussion about limitations and biases. The revisions did not lead to any substantive changes in the results reported in the original article. The reaction to Littman’s cautiously analyzed and mildly worded research could serve as a case study for my observation in the introduction that for an academician to depart openly from the orthodoxy usually carries a price.

  38. Littman (2018): 15–16 of 41.

  39. J. Michael Bailey and Ray Blanchard, “Gender Dysphoria Is Not One Thing,” 4thWaveNow, December 7, 2017, 4thwavenow.com.

  40. de Graaf, Giovanardi, Zitz et al. (2018).

  41. Zucker (2017).

  42. de Graaf, Giovanardi, Zitz et al. (2018): Fig. 1.

  43. Arcelus, Bouman, Van Den Noortgate et al. (2015).

  44. Zucker (2017).

  45. From the Encyclopedia of Surgery, available online at www.surgeryencyclopedia.com: “Reliable statistics are extremely difficult to obtain. Many sexual reassignment procedures are conducted in private facilities that are not subject to reporting requirements. Sexual reassignment surgery is often conducted outside of the United States. The number of gender reassignment procedures conducted in the United States each year is estimated at between 100 and 500. The number worldwide is estimated to be two to five times larger.”

  Appendix 3: Sex Differences in Brain Volumes and Variance

  1. The earliest was Willerman, Schultz, Rutledge et al. (1991).

  2. For a review of the 1990s literature, see Goldstein, Seidman, Horton et al. (2001).

  3. Based on the 25 unique studies listed in Ruigrok, Salimi-Khorshidi, Lai et al. (2014): Table 2.

  4. Ritchie, Cox, Shen et al. (2018).

  5. Desikan, Ségonne, Fischl et al. (2006).

  6. Caspari (1979).

  7. Holloway (1979); Epstein (1979).

  8. Willerman, Schultz, Rutledge et al. (1991).

  9. McDaniel (2005), a meta-analysis of in vivo brain volume and full-scale IQ, reported an average correlation of +.33. Pietschnig, Penke, Wicherts et al. (2015) argued for a comparatively low average of +.24.

  10. Gignac and Bates (2017) reanalyzed the 2015 set of studies in Pietschnig, Penke, Wicherts et al. (2015), restricting them to ones with samples of healthy adults and correcting for restriction of range. They reached an estimated correlation of +.31. The 2017 meta-analysis also classified studies according to their quality of measurement—“fair,” “good,” and “excellent.” The estimated correlations for these subsets were +.23, +.32, and +.39 respectively.

  11. van der Linden, Dunkel, and Madison (2017): Table A1.

  12. Perhaps the number of neurons in the cerebellum is also important. See Buckner (2013).

  13. Herculano-Houzel (2017). Comparing brain volumes of men and women does not pose the same problem. The Ruigrok meta-analysis found that the male-female ratios for the volumes of the cerebrum and cerebellum were close (9.8 percent and 8.6 percent larger in males respectively). Ruigrok, Salimi-Khorshidi, Lai et al. (2014): Table 3.

  14. See Hogan, Staff, Bunting et al. (2011), which also has a good literature
review of similar work.

  15. Reardon, Seidlitz, Vandekar et al. (2018).

  16. Van Essen (2018): 1184.

  17. Reardon, Seidlitz, Vandekar et al. (2018): Table S3.

  18. There is an active research program questioning whether this is true. Neves, Guercio, Anjos-Travassos et al. (2018) found that “whereas neuronal number is a good predictor of cognitive skills across species, it is not a predictor of cognitive, sensory or motor ability across individuals within a species, which suggests that other factors are more relevant for explaining cognitive differences between individuals of the same species.” The species in question was mice, however, so much remains to be learned.

  19. Winkler, Kochunov, Blangero et al. (2010).

  20. Ritchie, Cox, Shen et al. (2018).

  21. Sowell, Peterson, Kan et al. (2007).

  22. Gur and Gur (2017): 9.

  23. de Vries (2004): 1063.