Monday, 17 November 2025

The Demographic Strawman and the Erasure of Context

Doran, Martin & Zappavigna (2025: 177-8):

A third complementary perspective on tenor that needs to be brought into the picture involves the social roles and relations that affect the way we talk (Halliday & Matthiessen, 2014) – at a relatively local level (e.g., the way we talk as a mother-child pair) or at a much more global level (e.g,. the way we talk in relation to our economic class). In some discussions of tenor in SFL, there is a suggestion that we can simply ‘wire-in’ various demographic features and suggest there is a relatively straight realisational path from, say, ‘mother’ and ‘daughter’ to the language being spoken. But what we have tried to show through this book is that being a mother involves a wide range and diverse set of options. In other words, how a mother means is immensely varied and contested and constantly undergoing change. It intersects with the genres, situations, communities and feelings we have at all times.


ChatGPT Comments:

Doran, Martin & Zappavigna (2025: 177–8) introduce what they call a “third complementary perspective” on tenor. This is the construction of a strawman to justify their deviation from the canonical model.

1. They mischaracterise the Hallidayan position to defend their own

The claim that SFL suggests you can “wire-in” demographic labels (“mother,” “daughter,” “economic class”) and get a deterministic linguistic output is not found in Halliday & Matthiessen or any serious SFL scholarship.

What Halliday actually says (and what their citation pretends to align with) is that tenor is the configuration of social roles and relations in a situation — not a demographic variable, and not a fixed identity category.

The “mother–child pair” is not a demographic; it is a social relation in a situation type.
They replace situational roles with demographic categories and then call the resulting caricature “SFL”.

This inversion allows them to “correct” a model that never existed.


2. They continue the slippage between contextual relations and semantic realisations

They say:

“being a mother involves a wide range and diverse set of options… how a mother means is immensely varied and contested”

But this is not a claim about tenor.
It is a claim about interpersonal meaning, genre, and community-specific practices.

Tenor is not:

  • “how a mother means”

  • “the feelings we have at all times”

  • “options available to mothers”

Tenor is the situational role configuration being enacted through meaning — not the semantic choices themselves and not the sociocultural variability of identity categories.

By grounding tenor in the variability of identity performance, they collapse context into discourse semantics yet again.


3. The final sentence completes the conflation

“It intersects with the genres, situations, communities and feelings we have at all times.”

This is precisely backwards.

  • Tenor does not “intersect with” genre;
    in their model, genre is realised by tenor (alongside field and mode).

  • Tenor does not intersect with feelings;
    feelings are realised by appraisal systems in discourse semantics.

  • Tenor does not intersect with communities;
    communities constrain situation types, which condition tenor.

Their formulation treats tenor as a free-floating sociocultural affective force, drifting across strata and systems, touching whatever seems relevant to identity or meaning.


Summary

This passage exemplifies how the authors protect their “resource” model by attacking a strawman version of SFL: they caricature tenor as demographic determinism, replace contextual relations with identity-performance semantics, and end by dissolving tenor into an amorphous blend of genre, community, and feelings — all while citing Halliday to legitimise a model that directly contradicts his. 

No comments:

Post a Comment