Are Music and Language Homologues?

I have argued previously that music and language are homologous functions that evolved from a common ancestor that embodied their shared features, something that I have called the “musilanguage” system.1 According to this model, the shared/ parallel features of music and language evolved before their distinct, domainspecific features. These parallel features include use of a limited pool of discrete building blocks, arranged combinatorially, to generate structured phrases, modulated by expressive phrasing mechanisms. In thinking about the evolutionary relationship between music and language, it is useful to make a distinction between three types of features and to consider models for their respective brain localizations (see FIG. 1): (1) shared features, (2) parallel features, and (3) distinct features. Shared features are those that are identical between music and language, and would include the general processes of vocalization as well as affective prosody, that is, the expression of emotional states in music or language. Parallel features are those that are analogous (but not identical) between music and language, and would include the features of discreteness, combinatoriality, phrase formation, and phrasing mentioned above. Finally, the distinct features are those that are specific to each domain and are therefore neither shared nor parallel, and would include music’s use of isometric rhythms and pitch blends and language’s use of words and propositional syntax. I want to propose a model for how these three types of features could be instantiated in the modern brain starting from a bilaterally symmetric “musilanguage” system in the early hominid brain: (1) shared features are mediated by shared modules, (2) parallel features are mediated by duplicate modules, and (3) distinct features are mediated by diverse neural areas whose arrangements are not predictable a priori. The idea of shared modules implies that during the divergence of music and language from the “musilanguage” precursor, both functions came to adopt the same neural areas for the same functions. By contrast, the idea of duplicate modules suggests that during the divergence process, parallel functions developed as specializations in either the left or right hemisphere, where homologous functions came to occupy more-or-less corresponding positions in the two hemispheres. Finally, local-