查看文章

Models of coarticulation typically handle local interactions between temporally overlapping consonants and vowels either through advance planning or algorithmic calculation of weighted averages, or some combination of these (see Farnetani and Recasens, 2010). Here we propose a third option, where the intrinsic biomechanics of the human body resolve local interactions in a feed-forward model. Gick et al.(2012) describe a model in which muscular responses to articulatory and aerodynamic aspects of speech production interact. In this model, facial muscle actions associated with achieving articulatory goals (eg, closing the lips) activate simultaneously with those responsible for containing intraoral air pressure (eg, stiffening lips and surrounding perioral structures). These simultaneous activations are resolved via the body’s natural biomechanics, with no additional calculation or planning, as demonstrated in a …