LAUSR

In their comprehensive review article, Gong, Shuai and Wu argue that theorizing about the foundations of language requires taking evolution into account—in particular, the cultural evolution of linguistic structure. We wholeheartedly concur with this point [2,9]. In this commentary, we further emphasize two features of language that support this account: linguistic diversity and individual variation. These are critical to bringing theories of language in line with evolutionary considerations. Variation fuels evolution—without it, no adaptive changes are possible (though other types of change may occur due to non-adaptive processes such as random drift). Fortunately, as noted by Evans and Levinson [12], human language is perhaps unique among animal communication systems in its sheer diversity. There are currently between 6,000 and 8,000 different languages spoken in the world [12,15,21], and these vary substantially along every imaginable linguistic feature. However, this diversity is not randomly distributed. On the contrary, there are statistical trends that govern these patterns [7,12,20], be it geographical [14], morphological [4], lexical [25,17], phonological [11,13], among others. As the target article shows, there has been substantial progress towards understanding the mechanisms that allow for language to emerge. However, there has been significantly less progress in determining both the general processes that allow languages to be diverse, and the specific processes that produce the patterns detectable in that diversity. One of the better-studied processes that could drive diversity is the adaptation of language to the cultural community that speaks it. On a timescale of generations, a language culturally evolves to be learnable and usable [6,8]. However, different communities of language users are not necessarily homogeneous. They may have systematic differences in their processing and production mechanisms. Some of the relationships between these differences and language diversity have been explored. For example, Dediu et al. [11] hypothesize that slight but systematic differences in the vocal tract could underlie some of the phonological patterns such as the appearance and maintenance of click consonants.