On Martin's Distinction Between Grammar And Cohesion

Bateman (1998: 5):

The existence of a distinct, more abstract stratum of semantic description leaves Martin free to explore meanings whose realisations span both structural (i.e., lexicogrammatical) and nonstructural (i.e., ‘cohesive’) realisations, and to consider these as available, discourse motivated alternatives that may be systematically related. It provides a principled basis for capturing precisely those commonalities and regularities that fall outside the remit of grammar.

Blogger Comments:

[1] The implication here is misleading.  The stratification of content in SFL theory precedes Martin's (1992) proposal, and can even be found explicitly stated in the work that Martin takes as his point of departure: Cohesion In English (Halliday & Hasan 1976: 5). 

[2] In SFL theory, the distinction between structural and nonstructural realisations is a distinction within the resources of the textual metafunction at the level of lexicogrammar.  Martin's misrepresentation of this as a distinction between lexicogrammar and cohesion serves his purpose of relocating cohesion from grammar to semantics, as explained here.

[3] As demonstrated in previous posts, Martin's argument for proposing a discourse semantic stratum is invalid — as summarised here — and the model is inconsistent in terms of the principles of metafunction and stratification.  As such, it is entirely false of Bateman to claim that "it provides a principled basis for capturing precisely those commonalities and regularities that fall outside the remit of grammar."