Why Mass, Presence, And Association Cannot Be Principles Of Instantiation

Doran, Martin & Zappavigna (2025: 18):

Instantiation is typically considered in terms of the coupling of choices as conditioned by varying probabilities in different texts, text types and diatypes. 

But recognising the concepts of mass, presence, and association (i.e., status/contact) as metafunctionally diversified syndromes of choice offers a perspective on instantiation that can help explain why certain choices co-occur in certain situations. That is, mass, presence and association can be considered principles of instantiation – principles underpinning the co-selection and arrangement of features across strata and metafunctions. This can help us move away from relatively ad hoc explanations as to why particular choices are taken up in particular texts or situations and move us into considering different texts and the various domains they enter into in terms of a multidimensional set of principles. 

For example, scientific writing tends to involve significant interlocking networks of activity, taxonomy and property, but relatively little evaluative language in comparison to other disciplines (e.g. Halliday & Martin, 1993). We could explain this in terms of science’s aim for very strong ideational mass (technicality) but relatively weak interpersonal mass (iconisation). We could also describe the fact that it regularly aims to link theory to data as illustrating a wide range of ideational presence (iconicity).

Reviewer Comments:

[1] Here the authors misrepresent their own misunderstanding of instantiation as the "typical" view. As a process, instantiation is the selection of features and the activation of their realisation statements in logogenesis. At the system pole of the cline of instantiation, every feature of system has a probability of instantiation relative to other features in the system, and at the instance pole, texts vary by the relative frequency of feature selection.

The midway point of variation on the cline can be viewed from the system pole as register ("diatype"), or from the instance pole as text type, with registers varying in terms of the probability of feature selection, and text types varying in terms of the frequency of feature selection.

Importantly, at ontogenetic and phylogenetic timescales, the probabilities of feature selection in the system are altered by the frequencies of feature selection in the instances.

[2] To be clear, even if mass, presence and association provided a theoretically valid model of context, they could not "be considered as principles underpinning the co-selection and arrangement of features across strata and metafunctions" simply because instantiation is not an inter-stratal relation. Instantiation is the relation between system and instance at each stratum. There is no instantiation relation between context and language. Context and language are related by realisation, elaborating identity, so there is no causal (enhancing) relation between them.

[3] This is misleading, because this is simply the imposition of terms that do not provide explanation. That is,

• it does not explain why science writing has less evaluation, it just labels it as strongly technical;
• it does not explain why science writing has less evaluation, it just labels it as weakly iconised.

[4] As demonstrated here, the authors' notion of iconicity misunderstands a congruent relation between ideational lexicogrammar and ideational semantics as an iconic relation between the ideational content of language (text) and the ideational dimension of context ("what it is talking about").