Misunderstanding SFL's Architecture Of Language

Doran, Martin & Herrington (2024: 212):

SFL has been concerned with the relation between language and context throughout its 60-odd year history. This has produced a range of models and highlighted a breadth of phenomena that must be accounted for if we are to develop a truly social semiotics. For much of the history of SFL, there have been only three variables used to model context: field, tenor, and mode. But over the last few decades, our theoretical architecture has expanded. The context plane has been expanded to include genre (Martin and Rose 2008). Our understanding of the system of language in relation to the instance has become more nuanced through the development of the cline of instantiation (Halliday 1991a; Matthiessen 1993). And our modelling of how resources in language get differentially distributed across society (allocation) and how people use language to build community (affiliation) has opened up via the scale of individuation (Martin 2010). Until these dimensions were elaborated, field, tenor, and mode were expected to do all of the heavy lifting when it comes to our modelling of context. But as we noted above, recognising the dimensions of instantiation and individuation in addition to realisation allows us to distribute the work. In doing so, we are able to make theoretically clearer distinctions between aspects of ‘context’, while holding onto the context/metafunction resonance Halliday proposed to underpin SFL modelling.

Reviewer Comments:

[1] Trivially, the first decade of this 60-odd year history was not SFL, but Scale-&-Category Grammar, which, following Firth, gave equal weight to the paradigmatic and syntagmatic axes, and did not include metafunctions or system networks.

[2] To be clear, 'social semiotics' specifies a class of semiotics, as opposed to, say, somatic semiotics.

[3] As previously explained, SFL models the contextual aspects of Martin's genre as mode. Martin has merely relocated these distinctions to a higher stratum, thereby creating theoretical inconsistencies. 

[4] To be clear, instantiation and individuation are not models of context. They are dimensions of all strata. For example, at the level of context, the field tenor and mode of the culture is instantiated as the field tenor and mode of situations. The authors' model of context does not include these two poles on the cline of instantiation, because Martin (1992) misconstrued context potential as a stratum of genre, and context instance as a stratum of register, such that potential is realised by instance.

[5] This continues the authors' misunderstanding of 'context-metafunction resonance' as requiring that contextual parameters only specify linguistic realisations in their counterpart metafunction, as previously explained.