Dr. James Thompson has been blogging about a very important new paper by scholars Michael A. Woodley, Mateo A. Penaherrera, Heitor B.F. Fernandes, David Becker, and James R. Flynn.  It discusses the 20th century increase in brain size and what, if any relevance it may have to the 20th century rise in IQ test performance (The Flynn effect).    The paper cites a study where brains of all 20-50-year-olds (excluding pathological cases) were autopsied in the London Hospital over multiple decades.  Men and women born in 1860 had brains that averaged 1372 g and 1242 g respectively, and those born in 1940 averaged 1424 g and 1265 g respectively.  That’s a gain of 52 g for men in 60 years, and a corresponding gain of 23 g for women.

The authors also cite some brain weight studies from Germany done in various years and conclude from the trend line that mean male brain weight measured in 1880 was likely 1342.63 g while mean brain weight reported in 1979 was likely 1415.79 g, a difference of 73.16 g in 99 years.  The corresponding figures for German women were 1218.59 g in 1880 and 1270.86 g in 1979, a gain of 52.27 g.

To convert these gains to standard deviation units, the authors cite a data-set published by Ho et al (1980) which found that white men and women (in the U.S., and presumably in other Western countries too) have a brain weight standard deviations (SDs) of 130 g and 125 g respectively.  From here they conclude that the brain size of UK males increased by 0.4 SD in 80 years and for UK females, the increase was 0.18 SD.  Similarly, in Germany male brain weight increased by 0.56 SD in 99 years, and for German females, the brain weight increase was 0.42 SD.  Averaging the rate of gains of all four groups together (which the paper does not do or endorse), the study implies brain size has been increased in Western countries by about 0.21 SD per half-century.

Pumpkin Person responds:  The Ho et (1980) standard deviations are not ideal in my humble opinion because that data set included the brains of people ranging in age from 20 to 80.  Such a wide range of ages will inflate the standard deviation to an unreasonable degree because of both age and cohort differences in brain size.  Since a major purpose of the paper is to show birth cohort gains in brain size in SD units, one needs to know the brain size standard deviation from a sample of people where everyone has the same age and birth year.  Such data may not be accessible, but a good substitute is to look at cranial capacity data from the U.S. military because enlisted army personnel tend to all be young adults and since cranial capacity reflects peak brain size, age related declines in brain size are removed as a source of variance.  Based on the standard error and sample sizes in table 1 of this paper, enlisted Caucasoid males have a cranial capacity standard deviation of about 91 cm3 and enlisted Caucasoid females have a cranial capacity standard deviation of about 90 cm3.

To express the brain size gains in the paper using these leaner standard deviations, we must convert autopsied brain weight into volume.  Scholar J.P. Rushton has explained how to do so.  One must first divide by 1.09 because brain weight increases by 9% post-mortem so the in-vivo weight is 9% lower.  Then, converting brain weight to brain volume is simply a matter of multiplying by 1.036.

When these simple calculations are done on the brain weights discussed in the paper, we see that for British cohorts born in 1860 to 1940 (who died between 20 and 50), cranial capacity increased from 1304 cm3 to 1353 cm3 respectively in men (an increase of 49 cm3) and from 1180 cm3 to 1202 cm3 respectively in women (an increase of 22 cm3).  Meanwhile, German brain volume from 1880 to 1979 increased from 1276 cm3 to 1346 cm3 in men (a 70 cm3 increase) and 1158 cm3 to 1208 cm3 in women (a 50 cm3 increase). Dividing these gains by the above mentioned cranial capacity standard deviations for Caucasoid men and women respectively, gives gains of 0.54 SD and 0.24 SD for British men and women over an 80 year span respectively, and gains of 0.77 SD and 0.56 SD respectively for German men and women over a 99 span respectively. 

Averaging the rate of gains of all four groups together, it implies brain size has been increased in Western countries by about 0.29 SD per half-century –somewhat larger than the 0.21 SD gain per half-century implied by the paper, but still much smaller than the 1 SD gain in brain size per half-century assumed by both scholar Richard Lynn (1990) and scholar Arthur Jensen (The g Factor 1998, pg 326), making this paper potentially ground-breaking.  Implications for the Flynn effect (the 1 SD rise in IQ test performance per half-century) will be discussed in the next post.