How humans stand out in frontal lobe scaling

All studies from the last 15 y of collecting data on primate frontal lobe volumes share one central result: allometric scaling of the human frontal lobe is not larger than expected. Recent studies collecting data on the prefrontal cortex confirm this result, but additionally indicate human prefrontal white matter specialization.

A recent publication by Barton and Venditti (1) reanalyzed data from previous studies, arguing that a common phylogenetic framework will clarify perceived confusion in the literature. The authors present scaling results from replicate analyses of datasets that have previously been analyzed using appropriate phylogenetic methods, and from phylogenetic analyses of datasets that have not. The study by Barton and Venditti, however, fails to emphasize that for the latter datasets, the new analyses indicate that the authors are controlling for a phylogenetic signal that does not exist (λ < 0.01), which explains why their scaling results mirror those presented in the primary work. Based on a variable-rates approach, Barton and Venditti further report that human relative frontal gray matter, and prefrontal white and gray matter volume, are not larger than expected, which are also results already established in the primary work (2, 3).

Barton and Venditti (1) report one unique result: using data based on gross anatomical delineation criteria, larger than expected human (left) prefrontal white to gray matter is suggested to be a result of smaller than expected nonprefrontal white matter or (left) prefrontal gray matter. In Fig. 1, I provide evidence that this suggestion is not supported by data based on cytoarchitectonic delineation criteria [considered more valid than gross anatomical criteria (4)].

Fig. 1.

The 95% confidence intervals of phylogenetic generalized least-squares regressions (cf. 1). Thick lines indicate nonhuman allometry, thin lines nonape allometry. (A) Human left prefrontal white relative to gray matter is shown to be larger than expected; B and C show this is not explained by smaller than expected nonprefrontal white matter or left prefrontal gray matter (contrary to ref. 1). Data are available from the departmental Web site of J.B.S. (http://www.stonybrook.edu/anthro/staff/jsmaers.shtml) or on request. White circles indicate human, gray indicate ape, black indicate monkey values.

Barton and Venditti further state that previous work (3) indicated that “white matter volume relative to grey matter volume in the left (but not right) PFC [prefrontal cortex] was found to be significantly greater in apes than in monkeys” (1). This statement is incorrect. The work referred to [Smaers et al. (3)] indeed states that “residuals of the regression ... are elevated in apes and humans compared to monkeys” (considering results based on nonhuman allometry). However, this does not refer to a significant result with respect to apes but rather to a trend in which apes, contrary to monkeys, consistently have positive residuals, with only human values significantly different from others (Fig. 1A). The reanalysis by Barton and Venditti simply confirms the results presented in Smaers et al. (3), but fails to recognize that apes consistently indicate positive residuals. Additional analyses comparing ape values to the allometric trend of monkeys alone (nonape allometry) confirms the importance of this “ape trend” by indicating larger than expected values for several species (Fig. 1A).

Results presented in Barton and Venditti (1) duplicate those of the corresponding primary work, but the conclusion that no aspects of the human frontal lobe are larger than expected (controlling for allometry) is not upheld by all data (Fig. 1). The suggestion that future work should focus less on the frontal lobe in isolation and more on distributed networks also mirrors what has been suggested in the corresponding primary work. Indeed, research on the evolution of distributed systems—and the contribution of the frontal lobe within them—has already advanced beyond what Barton and Venditti highlight (5).