Abstract: The investigation of the neural basis and evolution of language abilities is best pursued as a search for language adaptations rather than as a search for the language faculty. The species uniqueness of language functions is contrasted with the conserved homologies linking human brain structures to anthropoid primate brain structures, and the failure to find species-specific neuroanatomical or genetic correlates of linguistically-defined innate features of language (e.g. universal grammar). Comparisons to animal call systems demonstrate minimal anatomical overlap between language systems these vocal functions, and yet extensive overlap with the anatomical substrates of gestural language production, suggesting that language evolution did not proceed by progressive elaboration from nonhuman vocal communication. Although there are no unambiguous gross neuroanatomical dishomologies distinguishing human brains that would suggest a role in language processing, there are clear allometric deviations of quantitative traits, including both gross brain size and deviant scaling of internal structural relationships in human brains, that suggest plausible roles in language processing. Evidence of correlated changes in patterns of axonal connections also implicate the extensive allometric deviations of human brains with language adaptations. One of the most likely correlates of allometrically-related connection change related to language evolution involves the existence of direct cortical projections to the nucleus ambiguous (the laryngeal control nucleus of the brainstem), which are likely absent in other mammals. This enables humans to have articulate control over the viscero-motor lung and larynx control systems and to couple this with articulate control of the skeletal-motor tongue, facial, and jaw muscles. Tracer studies and physiological recording studies of the macaque monkey ventral premotor and prefrontal cortex provide evidence of extensive homology of connectivity, suggesting that the circuits associated with these cortical areas were recruited for language processing during human evolution. Also cells in adjacent macaque premotor cortex that differentially fire with respect to self-initiated and other-initiated grasping behaviors. This suggests that the human homologue to this or nearby areas might be relevant to the mimicry necessary to acquire language. Genetic studies of human language adaptations have identified a gene, FOXP2, that is damaged in an inherited language deficit that affects automatizion of speech and syntactic processes. It turns out to be a highly conserved gene regulating forebrain basal ganglia development in embryogenesis. The human version of the gene contains two unique point mutations, neither of which is implicated in the language disorder. The functional difference produced by these changes are not known, but appear to have spread quickly in the early human population. The homologues to this gene in other species also play roles in vocal behavior. This genetic change is probably only one of a great many that contribute the adaptation for language. These sources of comparative functional and anatomical information argue against saltationist scenarios that hypothesize a sudden recent appearance of language abilities, and instead suggest that many diverse adaptations converged to make language possible.