Image-schemas are recurrent, dynamic instantiations of sensory-motor activity that bring at once ?embodied? and ?mental? non-conscious, emotional, preverbal meaning to human experience. It is assumed that the origins of image-schemas can be traced as far back as early infancy, emerging in the context of communicative musicality. Every-day adult-infant multimodal interactions would thus provide the enactive scenario where these embodied structures begin to develop in cognition. The talk presents the results of microanalytical data of adult-infant interaction in contexts of communicative musicality where those structures seem to emerge, and then seeks to follow through their developmental path to adult music cognition.