Body size and trees, not brains, best predict how many calls a primate makes

An analysis of vocal repertoires from 77 primate species reports that average brain size and typical social group size do not predict how many distinct calls a species produces, while the interaction between body size and habitat does ^[4][8][9]. The paper, by Adriano R. Lameira of the University of Warwick's ApeTank and published in Biology Letters, takes aim at a load-bearing plank of the argument that human language grew out of social life ^[2][3][1].

The target is the social brain hypothesis: the proposal that increasingly complex social lives selected for larger brains, which in turn produced more advanced communication ^[1]. Lameira says his study was prompted by a 2000s-era paper by McComb and Semple, also in Biology Letters, that cemented the idea that social and vocal complexity evolve together ^[11]. His stated objection is not to that paper's reasoning but to its inputs: descriptions of primate vocal repertoires have expanded considerably in the two decades since, and his aim was to reassess sociality against a larger and newer dataset while controlling for physical and ecological effects ^[12][13].

The dataset, which the write-up describes as one of the most comprehensive assembled to date, comes from studies Lameira collated over years plus an extensive web search ^[16][14]. For each species he compiled repertoire size, defined as the number of distinct vocal signals produced, along with average body and brain size, typical social group size, and ecological lifestyle coded as ground-dwelling, arboreal or mixed ^[5][6]. The model controlled for phylogenetic distance, so that closely related species are not counted as independent evidence ^[7].

The result is a null for both social predictors. Neither brain size nor group size predicted the richness of a species' calls, which Lameira reads as evidence that being bigger-brained and more social does not translate into more signals ^[8]. What did explain repertoire size was the interaction of body size with habitat, which he summarises as body size and arboreal lifestyles ^[9][17]. He goes further on sociality, saying the analysis "offers hints that the benefits of group size taper off and that past a threshold number, newly added members become a burden and add a 'social tax' on communication" ^[10]. That threshold claim is a suggestion drawn from the same data, not a separate test, and should be read as such.

Two limits are worth stating plainly. The published account does not report effect sizes, model comparisons or the direction of the body-size-by-habitat interaction, so the strength of the positive finding cannot be judged from it ^[18]. And the dependent variable is a count assembled from separate published descriptions by different observers using different field methods ^[5][14]; a repertoire of "12 calls" is a decision about lumping and splitting before it is a measurement. Neither point rescues the social predictors, which failed on the same data that produced the positive result.

The context is that this is not an isolated dissent. The write-up notes recent work suggesting ecology mattered more to language evolution than originally predicted, and that more social members of a group do not always communicate in richer ways ^[15].

What to watch: whether the species-level repertoire counts and the model are released for independent re-coding, whether the body-size-by-habitat interaction survives when acoustic and respiratory constraints on large arboreal callers are modelled explicitly, and whether anyone tests the "social tax" threshold as a prediction rather than an interpretation ^[10].