The complementarity of the process and pattern
interpretations of Functional Grammar
Michael Fortescue
- Introduction
From its inception, the Functional Grammar formalism has in principle
been interpretable in two ways, either as a static system of functional
choices, or as a dynamic process model (more specifically, a production
model). Interest in the second kind of interpretation has increased with the
rapprochement between cognitive and functional linguistics in recent years,
and not everyone within the FG sphere has been as sanguine as Dik himself
in considering the formalism directly implementable as a computer model
of cognition.^1 What I want to explore in this chapter is the question of what
the consequences of interpreting FG as a process model really are.
If one wants the FG model to represent production and/or comprehen-
sion directly, one should, I would suggest, base it on underlying cognitive
processes which have an intuitive feel of plausibility about them (not, for
instance, on hypothetical neural circuits at some dehumanized ontological
level). As has been argued by Harder (1992: 313ff.), one aspect of the
model that simply does not fit a process/procedural interpretation is the
first-order predicate formalism of the lowest level of the model.^2 Taken
straight from the abstract world of logicians like Quine (1960), representa-
tions of this kind, with their variables, quantifiers and open predicates, are
frankly not obvious candidates for psychological ‘reality’. Strawson, in de-
scribing the cognitive origin of ‘perspicacious grammar’, specifically
warns against introducing variables and quantifiers as opposed to individ-
ual particulars at the basis level of the logical representation of
propositions (Strawson 1974: 117). That is because it is individual particu-
lars (and their more abstract, higher-order entities derived by ‘sub-