C. J. Everett

All of the world's spoken languages make consistent use of a relatively narrow set of contrastive speech sounds—phonemes. Here, we argue that phonemes constitute cognitive tools, supporting, guiding, and extending speaker cognitive capacities. We outline commonalities between phonemes and other cognitive tools, which include tendencies in their usage based on biological constraints, their extensive variation across cultural lineages, their criticality to the efficient transmission of information, and their importance in the scaffolding of various cognitive capacities. Studies of the commonalities of phonological systems reveal the physical and biological underpinnings upon which the tools of phonemes are necessarily predicated, and the constraints on their cross-cultural refinement. Our view complements cognitive linguistic perspectives on other cognitive tools, and on human perception and consciousness more broadly, by emphasizing the sounds of speech themselves.

In this chapter we consider the theories of embodied cognition and extended mind with respect to the human ability to engage in numerical cognition. Such an enquiry requires first distinguishing between our innate number sense and the sort of numerical reasoning that is unique to humans. We provide anthropological and linguistic research to defend the thesis that places the body at the center of our development of numerical reasoning. We then draw on archaeological research to suggest a rough date for when ancient humans first were able to represent numerical information beyond the body and in enduring material artifacts. We conclude by briefly describing how these capacities for embodied and extended numerical cognition shaped our world.

Recent research has suggested that the Pirahã, an Amazonian tribe with a number-less language, are able to match quantities > 3 if the matching task does not require recall or spatial transposition. This finding contravenes previous work among the Pirahã. In this study, we re-tested the Pirahãs’ performance in the crucial one-to-one matching task utilized in the two previous studies on their numerical cognition, as well as in control tasks requiring recall and mental transposition. We also conducted a novel quantity recognition task. Speakers were unable to consistently match quantities > 3, even when no recall or transposition was involved. We provide a plausible motivation for the disparate results previously obtained among the Pirahã. Our findings are consistent with the suggestion that the exact recognition of quantities > 3 requires number terminology