Abstract

Nonhuman primates show remarkable variation in several aspects of social structure. One way to
characterize this variation in the genus Macaca is through the concept of social style, which is based
on the observation that several social traits appear to covary with one another in a linear or at least
continuous manner. In practice,macaques aremore simply characterized as fitting a four-grade scale in
which species range from extremely despotic (grade 1) to extremely tolerant (grade 4). Here,we examine
the fit of three core measures of social style—two measures of dominance gradients (hierarchical
steepness) and another closely related measure (counter-aggression)—to this scale, controlling for
phylogenetic relationships. Although raw scores for both steepness and counter-aggression correlated
with social scale in predicted directions, the distributions appeared to vary by measure. Counteraggression
appeared to vary dichotomously with scale, with grade 4 species being distinct from all
other grades. Steepness measures appeared more continuous. Species in grades 1 and 4 were distinct
from one another on all measures, but those in the intermediate grades varied inconsistently. This
confirms previous indications that covariation is more readily observable when comparing species at
the extreme ends of the scale than those in intermediate positions. When behavioral measures were
mapped onto phylogenetic trees, independent contrasts showed no significant consistent directional
changes at nodes below which there were evolutionary changes in scale. Further, contrasts were no
greater at these nodes than at neutral nodes. This suggests that correlations with the scale can be
attributed largely to species’ phylogenetic relationships. This could be due in turn to a structural
linkage of social traits based on adaptation to similar ecological conditions in the distant past, or
simply to unlinked phylogenetic closeness.