JASs Invited Reviews

Paleobiological inferences on general cognitive abilities (intelligence) in fossil hominoids strongly rely on relative brain size or encephalization, computed by means of allometric residuals, quotients or constants. This has been criticized on the basis that it presumably fails to reflect the higher intelligence of great apes, and absolute brain size has been favored instead. Many problems of encephalization metrics stem from the decrease of allometric slopes towards lower taxonomic level, thus making it difficult to determine at what level encephalization metrics have biological meaning. Here, the hypothesis that encephalization can be used as a good neuroanatomical proxy for intelligence is tested at two different taxonomic levels. A significant correlation is found between intelligence and encephalization only at a lower taxonomic level, i.e. on the basis of a low allometric slope, irrespective of whether species data or independent contrasts are employed. This indicates that higher-level slopes, resulting from encephalization grade shifts between subgroups (including hylobatids vs. great apes), do not reflect functional equivalence, whereas lower-level metrics can be employed as a paleobiological proxy for intelligence. Thus, in accordance to intelligence rankings, lower-level metrics indicate that great apes are more encephalized than both monkeys and hylobatids. Regarding fossil taxa, encephalization increased during hominin evolution (particularly in Homo), but during the Miocene a significant shift towards higher encephalization (and inferred enhanced cognitive abilities) must have been also involved in the emergence of the great-apeand-human clade (Hominidae). This is confirmed by the modern great-ape-like degree of encephalization displayed by the fossil great ape Hispanopithecus, which contrasts with the rather hylobatid-like degree of the stem hominoid Proconsul. The similarly low encephalization of Oreopithecus might result from secondary reduction under insularity conditions, but the australopith-like degree of encephalization of Homo floresiensis seems incompatible with the cognitive abilities inferred from the stone tools attributed to this taxon.