Michael Devitt, Cultural Evolution and the Division of Linguistic Labour

There is a general consensus that there is a division of linguistic labour and that it is important in explaining the expressive power of human language; our ability to talk about phenomena beyond the reach of our own experience. But there is disagreement between Michael Devitt and defenders of causal description theories as to how that division is sustained in a linguistic community. Causal description theorists argue that we have indirect ways of specifying the referential targets of our names and terms; Devitt (on behalf of causal theories) argues that the doxastic prerequisites for referential competence are much more minimal. It is unclear how to resolve this debate, as appeals to intuitions about particular cases have little evidential weight. This paper explores a way forward, by seeing the division of linguistic labour as a special case of cumulative cultural learning. There is to hand a rich (though highly contested) literature on the cognitive prerequisites of cumulative cultural learning; one aim of the paper is to connect these two literatures. The more substantive aim is to distinguish between the cognitive demands of the vertical and the horizontal transmission of referential competence (that is intergenerational versus within generational transmission of that competence) and to suggest that while Devitt’s minimalism is a plausible view of the requirements of vertical transmission (for these are environmentally scaffolded in various ways), something closer to causal descriptivism is more plausible for the horizontal cases.