An adequate account of language evolution must reconcile the propositions that language is unique but has precursors. The paper of Bellman and Goldberg and that of Tzeng and Wang each neglects one of these propositions. We suggest that the uniqueness of language lies primarily in its dualistic structure which has a frame-content mode of organization: at the phonological level, consonant and vowel elements are inserted into syllabic frames; and at the morphological level, stem forms of content words are inserted into syntactic frames. We suggest that the morphological level evolved from the phonological level and that the frame-content mode of organization in phonology had a precursor in the form of bimanual coordination in which the nonpreferred (frame) hand holds an object operated on by the preferred hand (content). It is argued that lateralization of cortical function evolved first for bimanual coordination, then for language. Old World monkey hand preferences may be consistent with both the putative left-hemisphere specialization for bimanual coordination and the human right-hemisphere specialization for spatial functions.