The Mismeasure of Science: Stephen Jay Gould versus Samuel George Morton on Skulls and Bias

Gould on Morton, Redux: What can the debate reveal about the limits of data?

“…Michael has since published a four part blog post on issues related to Gould, Lewis et al, and Morton, and, while we do not agree with all of his claims, we again find his views to be much more subtle and well-considered than those expressed in Lewis et al (2011), DeGusta & Lewis (2011), and the various authors of Lewis et al (2011) in interviews. 6 Other writers have explored possible reasons for the inclusion of, and the prominence given to, the skull re-measurements by Lewis et al (2011), focusing on the problematic way in which the skulls were "acquired," and the continued controversy over arguments for the repatriation of, especially, native American remains. See, e.g.…”

“…Since not every skull included in the 1839 estimate was remeasured, Gould was unable to determine what, precisely, the seed-based average from that remeasured subset of the 1839 samples was. He argued that it was unlikely to be 5 Some authors cited by Lewis et al (2011) criticize Gould for failing to take Michael's work seriously, but their focus is entirely on Michael's re-measurements, ignoring Michael's much more important contributiondthe recognition that the assumptions Gould made in his statistical (re)analysis of the skull volume data were no better justified than those of Morton, and that no statistical analysis would (or even could), in this case, provide the "right" answer. Michael has since published a four part blog post on issues related to Gould, Lewis et al, and Morton, and, while we do not agree with all of his claims, we again find his views to be much more subtle and well-considered than those expressed in Lewis et al (2011), DeGusta & Lewis (2011), and the various authors of Lewis et al (2011) in interviews.…”

“…He argued that it was unlikely to be 5 Some authors cited by Lewis et al (2011) criticize Gould for failing to take Michael's work seriously, but their focus is entirely on Michael's re-measurements, ignoring Michael's much more important contributiondthe recognition that the assumptions Gould made in his statistical (re)analysis of the skull volume data were no better justified than those of Morton, and that no statistical analysis would (or even could), in this case, provide the "right" answer. Michael has since published a four part blog post on issues related to Gould, Lewis et al, and Morton, and, while we do not agree with all of his claims, we again find his views to be much more subtle and well-considered than those expressed in Lewis et al (2011), DeGusta & Lewis (2011), and the various authors of Lewis et al (2011) in interviews. 6 Other writers have explored possible reasons for the inclusion of, and the prominence given to, the skull re-measurements by Lewis et al (2011), focusing on the problematic way in which the skulls were "acquired," and the continued controversy over arguments for the repatriation of, especially, native American remains.…”

“…Gould (1978) famously argued that Morton's work (Morton, 1849) reflected unconscious manipulation to fit racist stereotypes. Lewis et al (2011) posthumously rehabilitated Morton, arguing that it was Gould who fudged the results. Lewis et al's article received significant press, both popular and scientific.…”

Introduction: Genomics and philosophy of race

“…As a second theme of our special issue, a number of authors analysed some of these controversies. Kaplan, Pigliucci, and Banta examine the dispute between Morton (1849) and Gould (1978) over the skull sizes of different folk-races, arguing (contra Lewis et al, 2011) that Gould was right to reject Morton's analysis as inappropriate and misleading, but that he was wrong to believe that a more appropriate analysis was available. Edge and Rosenberg weigh in on the so-called "Lewontin-Edwards debate" Winther, 2014), a disagreement over the implications of human genetic diversity for human ancestry, by modelling the mapping between genotype and phenotype and showing that when it comes to differences in neutral phenotypes across populations, Lewontin-type single-locus results (Lewontin, 1972) are more informative than Edwards-type multi-locus classification (Edwards, 2003).…”