Recent work in semantic and syntactic typology reinforces the idea that most of the structure of languages and especially the patterned variation across them is cultural rather than innate. This leaves somewhat unexplained why humans have languages of a kind that other species don’t. The explanation, I’ll argue, is that there is a rich underlying universal infrastructure of communicational abilities that must fundamentally affect the way languages are organized. I’ll review two aspects of this infrastructure: turn-taking and speech act coding and explore how this communication perspective might have substantial consequences for how we think about language structure.