David Rose wrote to David Kellogg on sys-func on 27 May 2022 at 7:00:
Where Halliday in the 1970s saw grammar emerging ontogenetically between phonology and semantics, with the appearance of metafunctions, another perspective has the content plane bifurcating into LG and DS to enable extended exchanges.Rose, D. (2006). A systemic functional approach to language evolution. Cambridge Archaeological Journal, 16(1), 73-96.
Blogger Comments:
[1] To be clear, for Halliday (1975), in ontogenesis a child moves from a two-level system to a three-level system, where lexicogrammar "interpolates" between meaning and sounding, and it is this that creates the semantics and phonology of language. Halliday (2004 [1975]: 28):
The paper suggests a sociosemantic interpretation of language development, based on the intensive study of one child, Nigel, from 9 months to 21/2 years. Nigel first developed (Phase I) a two-level system, having sounds and meanings but no words or structures, in which the meanings derived from the elementary social functions of interaction with others, satisfaction of needs and the like. This continued to expand for six to nine months, at which time the child entered the stage of transition to the adult language (Phase 11, corresponding to what is generally taken as the starting-point). This was characterised by the interpolation of a lexicogrammatical level between meaning and sound, and by the mastery of the principle of dialogue, the adoption and assignment of speech roles. It was also marked by a generalisation of the initial set of social functions to form a basic opposition between “language as learning” and “language as doing”.The transition was considered complete when the child had effectively replaced his original two-level system by a three-level one and moved horn monologue into dialogue; he then entered the adult system (Phase 111). He could now build up the meaning potential of the adult language, and would continue to do so all his life.
[2] This is seriously misleading. The false implication here is that "extended exchanges" are not already a feature of Halliday's model. This is countered by the quote above. This also relates to Martin's conceit that only his model accounts for 'meaning beyond the clause'. In fact, for Halliday, 'meaning beyond the clause' is realised by the systems of cohesion, and it is these grammatical systems that Martin (1992) misunderstands and rebrands as his own systems of discourse semantics. The evidence for this conclusion is presented here.
