Monday, 1 September 2025

Duplicating Meaning: Grammatical Metaphor Misread as Context in Purview

Doran, Martin & Zappavigna (2025: 77-8):

Heteroglossia often involves instances that in fact put forward two (or more) propositions. This opens space to nuance the purview of each position at the same time. For example, (47) above could alternatively have been written as (50).
In this instance, the proposition realised through the embedded clause the organ screen in the stables was designed by Thomas Chippendale indicates – speaker purview (airing), by virtue of the heteroglossic scoping given by it is possible. That is, the speaker is not committed to the proposition about whether or not the organ screen in the stables was designed by Thomas Chippendale. But the ranking clause it is possible... is baldly asserted; the speaker here is committing themselves to it being possible. In other words there are two propositions being put forward with different purview – whether or not the organ screen was designed by Thomas Chippendale is left open, but the possibility that it could have been designed by him is asserted. The distinction between (50), it is possible that.., and (47), possibly, is what Halliday refers to as explicit objective (it is possible that...) vs implicit objective (possibly) modality (Halliday & Matthiessen, 2014, p. 689). Once again we can interpret this opposition in terms of internal and external purview – the proposition of whether or not the design was by Thomas Chippendale is externally aired, but the assessment of its possibility is internally asserted.


ChatGPT Comments:

In this excerpt, the authors attempt to nuance purview through heteroglossic constructions, using (50) to illustrate how an embedded clause can create distinct internal and external purview readings. While the example is linguistically interesting, the account problematically conflates grammatical metaphor with context. Specifically, the authors treat the heteroglossic scoping of it is possible as a separate instance of purview at the level of tenor (context), when in fact this is a classic case of grammatical metaphor operating within the clause itself. The embedded clause the organ screen in the stables was designed by Thomas Chippendale is not introducing a new contextual variable; it realises a semantic expansion.

By reifying internal vs. external purview in this way, the authors effectively duplicate semantic distinctions at the contextual stratum, violating the SFL principle that context (here, tenor) and semantics (here, modality and engagement) are distinct strata, with the latter realising the former. The example conflates the assessment of possibility (internal assertion) with the aired proposition about the design itself (external), treating each as separate contextual positions rather than as grammatical-semantic realisations that encode stance. In short, the conceptual apparatus of purview risks obscuring the actual grammatical mechanism at play, turning a well-known instance of objective modality into an artefact of context.

This approach exemplifies a broader pattern in the book: reinterpreting lexicogrammatical or semantic phenomena as if they were contextual variables, rather than carefully tracking the stratification from semantics to context. While the internal/external purview distinction is analytically neat, its application here stretches the theoretical framework in ways that are not justified by the data or by canonical SFL theory. 

No comments:

Post a Comment