No Two Alike

by Judith Rich Harris

  48. Sayers, 1931/1968.

  49. Staples, 2003.

  CHAPTER 2: THAT DAMN RECTANGLE

  1. Harris, 1998a, p. 48.

  2. E.g., Gladwell, 1998.

  3. Cheever, 1991. Cheever’s mother supported their family by running a gift shop; he described his father as an “unemployed shoe salesman” (p. 215).

  4. E.g., Harris and Liebert, 1991.

  5. My epiphany is described in Harris, 1998a, pp. 264–265. The textbook I was working on at the time was never completed; I backed out of the project when I realized I could no longer give the publishers what they were expecting.

  6. Sayers, 1931/1968.

  7. To calculate the variance, subtract the mean (average) score from each individual score and square each of these deviations from the mean. The variance is the mean of these squared deviations.

  8. More precisely, heritability is defined as the proportion of the variance in the measured outcome—the variance in a personality trait, for example—that can be attributed to variations in genes.

  9. Bouchard and Loehlin, 2001.

  10. Falbo and Polit, 1986.

  11. Minor differences between children who did or did not go to day-care centers have turned up in some studies, but the differences have not been consistent from one study to the next. Some studies (e.g., NICHD Early Child Care Research Network, 2003; Sagi, Koren-Karie, Gini, Ziv, and Joels, 2002) show negative effects, while others (e.g., Andersson, 1992; Harrison and Ungerer, 2002) show positive effects. Similar inconsistencies are found in the research on the only child.

  12. Golombok, Cook, Bish, and Murray, 1995; Golombok, MacCullum, Goodman, and Rutter, 2002; van Balen, 1998.

  13. Begley, 1998, p. 56

  14. Vandermeer, 2004, p. 473.

  15. Ehrlich, 2000.

  16. Bouchard and Loehlin, 2001; Lake, Eaves, Maes, Heath, and Martin, 2000.

  17. E.g., Hoffman, 1991

  18. Eisenberg, Cumberland, and Spinrad, 1998.

  19. Feinstein, 1985, p. 409.

  20. Feinstein, 1985, p. 408.

  21. Estimates of the proportion of the variance attributable to shared environment cannot go below zero, which is one reason why they are seldom exactly zero. Any imprecision in the estimate will put it above zero, since it can’t wobble in the other direction.

  22. E.g., Bateson, 2002, p. 2212.

  23. Heritability does not have to be squared to get the proportion of the variance accounted for. Heritability is the estimated proportion of variance accounted for by variations in genes.

  24. Laleh and Ladan Bijani: see chapter 1. Chang and Eng: Ehrlich, 2000.

  25. E.g., Greenspan, 1997; Joseph, 2002.

  26. E.g., Kamin, 1974.

  27. Reiss, 2000.

  28. A correlation shouldn’t be confused with a percentage. A correlation of .50 between identical twins doesn’t mean that half of the twins are exactly alike and the other half are entirely different. It means that, on average, identical twins are fairly similar to each other in personality, but not extremely similar.

  29. Twins: Bouchard, Lykken, McGue, Segal, and Tellegen, 1990; Plomin, DeFries, McClearn, and Rutter, 1997. Adoptive siblings: Bouchard and Loehlin, 2001.

  30. Note that correlations can be either positive or negative, whereas heritability can only be a positive number (or zero). If adoptive siblings were less alike in personality than two adoptees picked at random, then the correlation for adoptive siblings would be negative.

  31. Reiss, 2000, p. 142.

  32. Plomin and Daniels, 1987.

  33. Pinker, 2002.

  34. Steinmetz, Herzog, Huang, and Hackländer, 1994.

  35. Diabetes: Bach, 2002. Ear infections: Casselbrant, Mandel, Fall et al., 1999.

  36. Andreasen, 1999; Bach, 2002.

  37. Kondo, Schutte, Richardson et al., 2002.

  38. Pinker, 2002.

  39. Dickens, 1860–1861, chapter 9.

  40. See Harris, 1998a, chapter 8; Harris, 2004a.

  41. Caspi and Roberts, 2001.

  CHAPTER 3: MONKEY BUSINESS

  1. Harris, 1995.

  2. J. R. Harris, “Don’t Blame Your Parents: The Nurture Assumption on Trial.” Paper presented at the annual meeting of the American Psychological Association, San Francisco, August 15, 1998 (available on the Gifts of Speech website, http://gos.sbc.edu/).

  3. APA Monitor, December 1998.

  4. Phone conversation with M. Bristol-Power, May 16, 2002.

  5. E.g., Suzuki, Griffiths, Miller, and Lewontin, 1989. My thanks to Kevin Rich, the artist who made these drawings.

  6. Maccoby, 2002, p. 42.

  7. Maccoby, 2002, p. 42.

  8. Plomin and Daniels, 1987.

  9. Rare in nature: Rowe, 2001. Mutant fruit flies: Suzuki et al., 1989.

  10. Caspi, Sugden, Moffitt et al., 2003.

  11. Criminal behavior: Mednick, Gabrielli, and Hutchings, 1987. Schizophrenia: Tienari, Wynne, Moring et al., 1994.

  12. IQ: Capron and Duyme, 1989. Criminal behavior: Gottfredson and Hirschi, 1990. Schizophrenia: Andreasen, 1999. Neighborhoods: Sampson, Morenoff, and Gannon-Rowley, 2002; see also Harris, 1998a, pp. 297–299.

  13. S. J. Suomi, “Parents, Peers, and the Process of Socialization in Primates.” Paper presented at the NICHD conference on Parenting and the Child’s World, Bethesda, MD, August 2, 1999.

  14. Harris, 1998a, p. 153.

  15. McCrae and Costa, 1988; Myers, 1992.

  16. Collins, Maccoby, Steinberg, Hetherington, and Bornstein, 2000; Rowe, 1994.

  17. Harris, 1998a. For my reply to Collins et al. (2000), and to a similar article by Vandell (2000), see Harris, 2000b.

  18. Collins et al., 2000, p. 224.

  19. Suomi, 1997. The title of the book and the names of the editors are given incorrectly in the reference list of Collins et al., 2000.

  20. Tey, 1951/1977, p. 84.

  21. Suomi, 1999.

  22. E-mail from W. A. Collins, February 18, 2000.

  23. E-mail from S. J. Suomi, March 1, 2000.

  24. E-mail from S. J. Suomi, March 1, 2000.

  25. Suomi, 1987, pp. 405, 406.

  26. Suomi, 1987, p. 414.

  27. Suomi, 1991, p. 48.

  28. Suomi, 1987, pp. 408, 410, 412, 413; Suomi, 1991, pp. 50, 51, 52, 54.

  29. E-mail from E. E. Maccoby, July 17, 2000.

  30. In the last few years, several papers by Suomi and his colleagues have been published that contain terms like “rearing condition” in their titles (e.g., Bastian, Sponberg, Suomi, and Higley, 2003). The studies described in these papers involve differences between peer-reared and mother-reared monkeys, not differences between monkeys reared by good versus bad mothers. Although I accept that peer-rearing—an unnatural condition that could never occur in the wild—might have deleterious long-term effects on a monkey’s social development, there is another possible explanation for the subtle deficits found in these animals: nutritional deficiencies. The mother-reared animals are breast-fed; the peer-reared animals are bottle-fed. A study from Suomi’s own laboratory (Champoux, Hibbeln, Shannon et al., 2002) showed that standard commercial monkey formula lacks certain fatty acids that are important in brain development, and that baby monkeys bottle-fed on the standard formula had weaker “orienting and motor skills” (p. 273). Adult humans who are deficient in these nutrients were found to have higher levels of a hormone associated with fear and anxiety (Hibbeln, Bissette, Umhau, and George, 2004).

  31. Suomi, 2002.

  32. Maestripieri, 2003, p. 321.

  33. Maestripieri, 2003, p. 322.

  34. Kagan, 1994.

  35. Begley, 1998, p. 56.

  36. Kagan, 1994; Arcus, 1991.

  37. Arcus, 2001, pp. 48, 50, 52–55. See also Kagan, 2003. In this article on biological and environmental influences on temperament and personality, Kagan doesn’t mention Arcus’s research at all. “It has proven difficult,” he reports, “to demonstrate that experiences of the infant years determine pro
files during childhood or adolescence” (p. 12).

  38. Rowe, 2002b.

  39. Bouchard and Loehlin, 2001.

  40. Quartz and Sejnowski, 2002, p. 129.

  41. E-mails from T. J. Sejnowski, June 18 and June 20, 2003.

  42. K. Wright, 1997, p. 78.

  43. B. Harris, 1979.

  44. Bellew, ca. 1955, p. 16.

  45. Collins, Maccoby, Steinberg, Hetherington, and Bornstein, 2001.

  46. Downloaded December 19, 2004, from http://www.psych.umn.edu/courses/ psy1001/study/Wk6and7LrnObj.pdf.

  47. Collins et al., 2000, p. 228.

  48. Maccoby, 2002, p. 42.

  49. Letter from D. Alexander, December 1, 1999.

  50. National Institute of Child Health and Human Development, 2001, pp. 9, 14 (also available on the NICHD website, http://www.nichd.nih.gov/publications/ pubs/parenting/index.cfm).

  51. Maccoby, 1992, p. 1008; see Harris, 1998a, pp. 8–9.

  52. Kagan and Moss, 1962.

  53. McCrae and Costa, 1994, made this discovery.

  54. Kagan and Moss, 1983.

  55. E-mail from G. Fein, June 29, 1996.

  56. Waring, 1996, p. 76.

  57. F. Farley, quoted in Begley, 1998, p. 54.

  58. Caspi, Sugden, Moffitt et al., 2003.

  59. Caspi, McClay, Moffitt et al., 2002.

  60. The criteria used by the researchers to categorize a subject as maltreated were not given in the report published in Science but are available online (http://www.sciencemag.org/cgi/content/full/297/5582/851/DC1). A subject was classified as “maltreated” if he was exposed to one unfavorable experience in childhood, and as “severely maltreated” if he was exposed to two or more unfavorable experiences. The unfavorable experiences included: two or more changes in primary caregiver before the age of ten; a mother who was observed to behave in an unhelpful or indifferent manner to the child at age three; and parents whose responses to a questionnaire on their child-rearing methods were in the top 10 percent in regard to harshness of discipline.

  61. Knutson, 1995.

  CHAPTER 4: BIRTH ORDER AND OTHER ENVIRONMENTAL DIFFERENCES WITHIN THE FAMILY

  1. Rowe, The Limits of Family Influence, 1994. David Rowe, whose earlier article titled “As the Twig Is Bent?” (1990) was an important influence on my thinking, died in 2003, at the age of fifty-three, of liver cancer. His death was a personal loss (David was an e-mail friend and colleague) as well as a great loss to science.

  2. Plomin and Daniels, 1987, p. 1.

  3. Plomin, Asbury, and Dunn, 2001.

  4. Turkheimer and Waldron, 2000, p. 78.

  5. Harris, 1998a, p. 27.

  6. Stattin and Kerr, 2000.

  7. Reiss, 2000, pp. 406–407.

  8. Quoted in A. M. Paul, 1998, p. 46.

  9. Plomin et al., 2001, p. 231.

  10. D. Grady (2003, September 13), “Surgeons meet conjoined twins: One quiet, the other with a grin.” New York Times (http://www.nytimes.com/2003/ 09/13/nyregion/13TWIN.html).

  11. There is an exception to the rule that differences between firstborns and laterborns are nonbiological. Blanchard (2001) has reported that the incidence of homosexuality is higher among laterborn males with older brothers than among firstborn males or laterborn males with older sisters. Blanchard believes that this effect is due to biological factors in the prenatal environment: he hypothesizes that the mother’s immune system becomes sensitized to an antigen produced by male fetuses and produces antibodies to it.

  12. The concept of dethronement comes from Alfred Adler (1927).

  13. Cheever, 1991, pp. 106–107.

  14. Daly and Wilson, 1988.

  15. Daly and Wilson, 1988; R. Wright, 1994.

  16. Trivers, 1985, p. 156.

  17. Eibl-Eibesfeldt, 1989; LeVine and LeVine, 1963; Romney and Romney, 1963.

  18. Jenkins, Rasbash, and O’Connor, 2003.

  19. British parents: Dunn and Plomin, 1990, p. 75. Americans: McHale, Crouter, McGuire, and Updegraff, 1995.

  20. Ernst and Angst, 1983; Hoffman, 1991.

  21. Sulloway, 1996, pp. xiv, 79. (Italics omitted.)

  22. Sulloway, 1996, pp. xv, 60.

  23. E.g., Dick and Rose, 2002, p. 70; Plomin, 1989, p. 109.

  24. Adoptive siblings: Bouchard and Loehlin, 2001. Twins: Bouchard, Lykken, McGue, Segal, and Tellegen, 1990.

  25. McCartney, M. J. Harris, and Bernieri, 1990.

  26. Quoted in Boynton, 1996, p. 74. (Italics omitted.)

  27. Plomin and Caspi, 1999, p. 254; Schachter and Stone, 1985; Simonoff, Pickles, Hervas, Silberg, Rutter, and Eaves, 1998; Spinath and Angleitner, 1998.

  28. E.g., E. M. Miller, 1997.

  29. Segal, 1999, p. 54.

  30. Sulloway, 1996, p. 373.

  31. Dawkins, 1989; Hamilton, 1964.

  32. Kin recognition: Pfennig and Sherman, 1995. Polygamous communities: Jankowiak and Diderich, 2000. Twin competition and grief: Segal, 1999, 2002; Segal and Hershberger, 1999.

  33. Sulloway, 1996, p. xv.

  34. Trivers, 1985, p. 156.

  35. Ernst and Angst, 1983.

  36. Sulloway, 1996, p. 21.

  37. Edwards, 1992, p. 302.

  38. Goodall, 1986, pp. 173, 116.

  39. Shatz and Gelman, 1973.

  40. Dunn, 1985.

  41. Dishion, Duncan, Eddy, Fagot, and Fetrow, 1994.

  42. Abramovitch, Corter, Pepler, and Stanhope, 1986, p. 228.

  43. Deater-Deckard and Plomin, 1999. The sibling pairs in this study were identified as older child, younger child, rather than firstborn, laterborn. However, by a conservative estimate, at least 80 percent of the older siblings in this study were in fact firstborns (e-mail from K. Deater-Deckard, September 9, 2003).

  44. Siblings who fight: East and Rook, 1992; Stocker and Dunn, 1990. Only children: Falbo and Polit, 1986.

  45. Turkheimer and Waldron, 2000; Reiss, 2000, p. 407.

  46. Sulloway, 1999, p. 192.

  47. Sulloway, 1996, p. 474, n. 78; 1999, p. 192. For a reply to these criticisms, see Jefferson, Herbst, and McCrae, 1998.

  48. Sulloway, 1996, p. xiii.

  49. Behavior with peers: Abramovitch et al., 1986; Deater-Deckard and Plomin, 1999. Academic achievement: Blake, 1989; Ernst and Angst, 1983; McCall, 1992.

  50. Paulhus, Trapnell, and Chen, 1999.

  51. Drugs and sex: Ernst and Angst, 1983; Rodgers and Rowe, 1988. Delinquency: Plomin and Caspi, 1999; Rowe, 1994. Early maturing girls: Stattin and Magnusson, 1990; D. M. Wilson, Killen, Hayward et al., 1994.

  52. Falbo, 1997, p. 939.

  53. Ernst and Angst, 1983; Sulloway, 1996, p. 73.

  54. My effort to replicate Sulloway’s meta-analysis is described in appendix 1 of The Nurture Assumption; my tally was posted in August 2000 on The Nurture Assumption website (http://xchar.home.att.net/tna/birth-order/). For a more detailed review, see J. R. Harris (2002), “The Mystery of Born to Rebel: Sulloway’s Re-Analysis of Old Birth Order Data” (http://xchar.home.att.net/tna/birth-order/methods.htm).

  55. Science reviewer: Modell, 1997. J. Angst is quoted in Horgan, 1999, p. 192.

  56. The results of Sulloway’s meta-analysis of Ernst and Angst’s data were first published in the journal Psychological Inquiry, in a short commentary (Sulloway, 1995) on a target article by Buss (1995). No list of the 196 findings was included. Commentaries are handled differently from regular journal articles; they are not ordinarily subjected to peer review. A commentary I wrote for the same journal (Harris, 1998b) was accepted without peer review.

  57. Park, 2000, p. 27; Townsend, 2000a/2004, pp. 144, 143. Sulloway’s letters to Townsend are quoted in Townsend, p. 144, and in Goldsmith, 2004, p. 14.

  58. Townsend, 2000a/2004, pp. 147–153. In March 2004, after Townsend’s lists of studies and findings from Ernst and Angst (1983) were published, Sulloway posted a list of his own on his website (http://www.sulloway.org/metaanalysis.html). What he posted, however, was not the long-awaited list of 196 findings (72
positives, 110 nulls, and 14 negatives) reported on pp. 72–73 of Born to Rebel. Instead there are 230 “scorable outcomes” based on “more than a nine hundred individual findings [sic].” These outcomes include “57 full confirmations, 42 partial confirmations, 112 nulls, 17 partially opposed outcomes, and 2 fully opposed outcomes” (downloaded September 28, 2004). The original 196 findings are presumably included in this list but perhaps re-coded in a different way; they are not identifiable. See Johnson, 2000/2004, p. 226.

  59. Sulloway, 1996, pp. 39–41, 332. See Townsend, 2000a/2004, p. 140, for a list of the points that have changed.

  60. Goldsmith, 2004, p. 18; Townsend, 2000b/2004.

  61. Johnson, 2000/2004.

  62. Harris, 2000/2004.

  63. Sulloway’s letter is quoted in Johnson, 2000/2004, pp. 212–213. (Johnson’s italics omitted.)

  64. Quoted in Johnson, 2000/2004, p. 213. Sulloway supplied Johnson with the wording for a lengthy “editorial forewarning,” which Johnson has printed not above Townsend’s article but in his editorial explaining the delay in publication.

  65. Johnson, 2000/2004, pp. 217–218, 222.

  66. Johnson, 2000/2004, pp. 218–219, 222. I’ve mentioned only one of Sulloway’s accusations; for the others, see Johnson.

  67. Sulloway’s letter to the president of Lake Superior State University is quoted in Johnson, 2000/2004, p. 222.

  68. Ritter, 2004, p. 17.

  69. Johnson, 2000/2004, p. 223.

  70. Johnson, 2000/2004, pp. 223–224.

  71. Townsend, 2000a/2004, p. 142; Sulloway, 2000/2004, p. 183; p. 198, n. 2.

  72. Johnson, 2000/2004, pp. 224–225. The use of odds ratios instead of risk ratios has been criticized by medical researchers (e.g., Davies, Crombie, and Tavakoli, 1998; Deeks, 1998) because odds ratios can give a misleadingly inflated impression of effect sizes. For example, Sulloway (1996, p. 51) reported that in one sample 54 out of 63 laterborns and 9 out of 20 firstborns supported liberal theories. The risk ratio is (54/63)/(9/20) = .86/.45 = 1.91; thus the “risk” of supporting a liberal theory was about twice as high for the laterborns in the sample. But Sulloway instead computed the odds ratio: (54/9)/(9/11) = 6.00/.82 = 7.32, which, he said, “equals 7.3 to 1 in favor of laterborn adoption.”

  73. Johnson, 2000/2004, p. 228.

  74. Sulloway, 1996, pp. 266–267.