David Rose Misrepresenting How Martin's Approach to SFL Differs From Halliday's and Hasan's

David Rose wrote to Sysfling on 17/9/23 at 8:53:

Returning briefly to Attia’s question...

How does Martin Approach to SFL is Different from Halliday and Hasan?

To understand its [Halliday and Hasan’s Cohesion in English] relation to Martin’s work in SFL, the place to start is his English Text. ... ET extends the descriptions in CiE...CiE describes structures of discourse, which ET extends to systems and structures

CiE focuses on textual structuring, termed ‘cohesion’, which ET extends to structures across all metafunctions 

CiE describes roles of grammar features in discourse structures, which ET extends as features in discourse semantic systems realised as features in grammar systems 

CiE’s relates textual structuring of discourse to the contextual variables of mode and field as ‘coherence’, which ET extends by describing systems of mode, field and tenor that are realised by discourse semantic structures in each metafunction.

Blogger Comments:

[1] To be clear, the question was about the different approaches to SFL, but Rose has reduced it to a question about the differences between cohesion (Halliday & Hasan 1976) and discourse semantics (Martin 1992).

[2] To be clear, Martin (1992) takes the original, groundbreaking work of Halliday & Hasan (1976) as his source material and rebrands their lexicogrammatical systems of cohesion as his own systems of discourse semantics.

[3] This is misleading, because it is untrue. Halliday & Hasan (1976) is concerned with non-structural lexicogrammatical resources of cohesion. 

[4] This is very misleading. Of Martin's 49 discourse semantic system networks, across four metafunctions, only 4 specify structural realisations, and all of these are confined to the interpersonal metafunction.

[5] To be clear, this is true, and in doing so, Martin creates theoretical inconsistencies by rebranding the textual lexicogrammatical systems of cohesive conjunction and lexical cohesion as ideational discourse semantic systems of ideation (experiential) and conjunction/connexion (logical).

[6] This is misleading, because it is untrue. Halliday & Hasan (1976) is concerned with non-structural lexicogrammatical resources of cohesion, not with discourse structures.

[7] This is misleading, because it is untrue. Not one of Martin's 49 discourse semantic system networks, across four metafunctions, specifies how discourse semantic features are realised as features in grammar systems.

[8] This is misleading, because it is untrue. Halliday & Hasan (1976) has no discussion of coherence. The notion of coherence first appeared in Hasan (1984), and it does not involve relating text structure to field and/or mode. See, for example, Hasan (1989/1985: 93-4) for the linguistic correlates of variation in coherence.

[9] This is seriously misleading in several ways. First, since Halliday & Hasan (1976) has no discussion of coherence, and coherence does not involve relating text structure to field and/or mode, Martin (1992) did not extend that model. Second, mode, field and tenor (style) were first formulated by Halliday in Halliday et al (1964), not Martin (1992). Third, Martin (1992) does not specify how mode, field and tenor are realised by discourse semantic systems, let alone structures.

∞

See also

The Practice Of Public Bluffing In The SFL Community.

David Rose Negatively Appreciating The Deployment Of Logic

David Rose Positively Judging 'Being Wrong'