Arbib hypothesizes that evolutionary modern language significantly postdates human speciation. Why should this be so? I propose an account based on niche construction theory, in which Arbib's language-ready brain is primarily a consequence of epigenetically-driven adaptation to the biocultural niche of protolanguage and (subsequently) early language. The evolutionary adaptations grounding language evolution were initially to proto-linguistic socio-communicative and symbolic processes, later capturing and re-canalizing behavioural adaptations (such as serial and hierarchical constructive praxis) initially “targeted” to other developmental and cognitive domains. The intimate link between praxic action and symbolic action is present not only in the human brain, but also in the human biocultural complex. The confluence of praxis and symbolization has, in the time scale of sociogenesis, potentiated the invention of domain-constituting and cognition-altering symbolic cognitive artefacts that continue to transform human socio-cultural ecologies. I cite in support of this account, which differs only in some emphases from Arbib's account, my colleagues' and my research on cultural and linguistic conceptions of time in an indigenous Amazonian community.