On Martin's 'Explicit Stratification Of Lexicogrammar And Discourse Semantics'

Bateman (1998: 16): 

One of the few large-scale attempts to carry out a program of this kind is Fawcett’s COMMUNAL project (cf. Tucker, 1998); here, however, the rules of the game have been altered by Fawcett’s adoption of the system network to represent semantics rather than grammar, a position reminiscent of that given in, for example, Halliday (1978:132). Meaningful comparisons are also made more difficult by Fawcett’s explicit rejection of Martin’s (1987) call for a clear separation of networks where features are motivated on grammatical grounds and those where features are semantic: a position developed to its logical conclusion in English Text’s explicit stratification of lexicogrammar and discourse semantics.

Blogger Comments:

[1] For evidence that Fawcett does not understand the theoretical dimension of stratification, see the clarifying critiques here.

[2] To be clear, Halliday (1978: 131) writes:

Table 3 (p. 132) sets out the principal semantic systems arranged by function and rank, showing their functional location in the semantic system and their point of origin in the lexicogrammar.

[3] This is misleading.  As demonstrated here, Martin's systems of IDENTIFICATION, CONJUNCTION and IDEATION are rebrandings of Halliday & Hasan's (1976) lexicogrammatical systems — reference, conjunction and lexical cohesion — misunderstood, and relocated to discourse semantics.