Misrepresenting Halliday On Register And Its Relation To Context

Doran, Martin & Zappavigna (2025: 9, 28n):

In the model used in this book, register is the cover term for the stratum comprising field, tenor and mode (Figure 1.2).

This contrasts with Halliday's use of the term register to refer to the skewing of probabilities in semantic systems inside language by field, tenor and mode (e.g. Halliday, 1991b, p. 48; 2006). Both concepts – field, tenor, and mode (what in this book we call register), and the language patterns that arise due to specific choices in field, tenor, and mode (what Halliday calls register) – occur within this model. The difference with Halliday in this regard is purely terminological.¹

¹ As this terminological distinction has often led to confusion, it would perhaps be useful to use Gregory’s (1967) suggestion of ‘diatype’ for the skewing of probabilities in the systems of language by choices in field, tenor and mode (i.e. Halliday’s register), and leave register as the cover term for field, tenor and mode. It is of course worth emphasising that there is a substantial difference in the model of social context presumed here and that of Halliday’s in that this model divides context into register (field/tenor/mode) and genre as distinct strata, whereas Halliday’s only has field/tenor/mode.

Reviewer Comments:

[1] This is misleading, since it misrepresents Halliday. Cf. Halliday (2005 [1991]: 60):

On the contrary; it is the probabilistic model of lexicogrammar that enables us to explain register variation. Register variation can be defined as the skewing of (some of) these overall probabilities, in the environment of some specific configuration of field, tenor and mode. It is variation in the tendency to select certain meanings rather than others, realising variation in the situation type.

That is, the skewing of probabilities defines register variation, not register, and it is not caused by field, tenor and mode of a situation type, because the realisation of one stratum in another is not a causal relation. Less importantly, in this instance, Halliday was describing lexicogrammar, not semantics.

[2] This is very misleading, because the difference is far more than merely terminological. For example, a consequence of renaming context after a variety of language is that context is no longer culture instantiated as situations of situation types. In fact, in Martin's model, which in terms of strata, distinguishes context from language, an instance of context is deemed to be a text, an instance of language. This contradiction alone is sufficient to invalidate Martin's model.

[3] To be clear, this qualification in an endnote, nineteen pages later, flatly contradicts the authors' claim in the text that the difference between the models is 'purely terminological'.