Halliday & Matthiessen (2014: 46-7):
Most accounts of ‘multimodal text’ so far have probably focused on combinations of written texts and instances of ‘visual semiotic’ systems. From a developmental and evolutionary point of view, it would make more sense to start with spoken texts unfolding together with instances of other somatic semiotic systems (i.e. other semiotic systems using some aspect of the body as their expression plane; see Matthiessen, 2009a, and cf. Thibault’s, 2004, notion of the ‘signifying body’) before moving on to interpret and describe exo-somatic semiotic systems. Indeed, the protolanguages of early childhood tend to be both vocal and gestural in their expression (see Halliday, 1975, 1992d, 2004); and we can hypothesise that the same was true of protolanguages in human evolution (see Matthiessen, 2004a). … An early systemic functional contribution to the study of language and gesture is Muntigl (2004), and the systemic functional work on language and gesture has been followed up by Hood (2011).
Not insignificantly, the theoretical basis of "Hood's contribution to the study of language and gesture" is the unpublished work of this blogger (Cléirigh 2009), which distinguishes — on the basis of ontogenesis and phylogenesis — between protolinguistic, linguistic and epilinguistic gesture–&–posture systems.