84 Carlos Inchaurralde
- Conclusion
We have examined here, from a cognitive perspective, some questions
which are relevant to the role of cognition in the Functional Discourse
Grammar model proposed by Hengeveld (this volume). His model assigns
a very important role to the cognitive component, since cognition provides
input to all three levels proposed. However, cognition itself remains rather
mysterious to the model. The view adopted here is that this cannot be oth-
erwise, since cognition is situated outside the main core of the model.
Nevertheless, some assumptions can be made about the role of cognition,
particularly with regard to the various inputs from the cognitive compo-
nent. The cognitive component contains not only general knowledge about
the world, but also knowledge about relevant communicative parameters
and about linguistic competence itself. We have argued that these are dif-
ferent kinds of information. Some are linked to more central cognitive
systems in the mind, while some others are not if we accept the evidence
supporting the modular approach proposed by Fodor (1983). A computer
metaphor has also been put forward that helps conceptualize the workings
of the model in a more understandable way, and cognition has its place in
it. However, we must realize that while cognition, as a general capacity,
cannot be ignored within the Functional Discourse Grammar model, it is
wise to assume that giving a very detailed account of its internal workings
would go beyond the scope of this grammatical approach and would lead
us into matters of a fully psychological import. And this would take us
away from our main object of study, the linguistic message.
Notes
- Note that I am referring to modularity of the mind and not to modularity
within the FG model. - Other properties of modular systems (cf. Fodor 1983) are that they have a
compulsory operation, which is automatic and independent of voluntary
processes; they work at a speed faster than other cognitive processes; their
output representations are incomplete and need to be completed by central
systems, which have more complete information; they are innate, since they
have a pattern of ontogenetic development; they are realized by means of a
fixed neural structure, that is, they tend to be clearly localized in certain
fixed areas of the brain; moreover, their alterations are highly specific and
selective, by contrast with what happens in central systems. - I give some examples in Inchaurralde (2000: 98–100).