Computational challenges of evolving the language-ready brain

Abstract

Abstract A theory of evolving the language-ready brain requires a theory of what it is that evolved. We offer the TCG (Template Construction Grammar) model of comprehension and production of utterances to exhibit hypotheses on how utterances may link to “what language is about.” A key subsystem of TCG is the SemRep system for semantic representation of a visual scene. We offer an account of how it may have evolved as an expansion of the ventral pathway supporting the planning of manual actions, complemented by a dorsal pathway for articulation. The Mirror System Hypothesis (MSH) claims that early Homo sapiens had protolanguage but not language and that cultural evolution yielded the social structures within which children could indeed acquire language. The article poses the challenge of understanding how a brain system could be innately specified that could develop into a TCG-like form, posing a range of questions for future research.