76 Carlos Inchaurralde
Also from psychology come the now well-established assumptions of
so-called ‘schema’ theories (Bartlett 1932; Rumelhart and Ortony 1977;
Rumelhart 1980), according to which everything we know about the world
is stored in large mental knowledge structures. Different labels have been
used to refer to these structures, depending on the type of information and
the way it is stored: ‘frames’, ‘scenarios’, ‘scripts’, ‘situations’, ‘idealised
cognitive models’, and so on (Minsky 1975; Schank and Abelson 1977;
Sanford and Garrod 1981; Fillmore 1982; Lakoff 1987, etc.) are all equally
useful in understanding how people use vocabulary in context. For in-
stance, Wierzbicka (1994, 1997) has pointed to the existence of different
‘scripts’ for interaction in different cultures and has drawn attention to how
words have uses and meanings that are highly dependent on culturally-
shared knowledge structures. All this seems to imply that it is necessary to
leave behind us the fixed-entry approach, provided by a dictionary perspec-
tive on the lexicon, and adopt some sort of distributed, encyclopaedic view.
However, at the same time, from an FG perspective, we do need fixed
entries in the Fund, where all predicate frames and terms are placed at our
disposal for the building of predications, and this is the place where the
cognitive component can act as input. The cognitive component is the in-
terface between our mental encyclopaedia and the lexicon that is available
to us. The cognitive component provides a ‘lexicopaedia’, in the form of a
fixed-entry list with points of access to a whole semantic network (cf. In-
chaurralde 2000). We cannot deny the evidence that shows us how
individual lexical units are a psycholinguistic reality,^3 which gives psycho-
logical support to the existence of a lexicon in the Fund; but at the same
time it is important that we admit that there exists semantic information of
an encyclopaedic nature, and this is a natural and simple way of incorporat-
ing it in the FDG model.
2.2. Long-term knowledge and the communicative component
Knowledge of the relevant parameters that can affect the coding of the in-
terpersonal level should also be part of the long-term information held by
the cognitive component, especially if we bear in mind that this is the first
level in Hengeveld’s model. There are elements in the communicative con-
text which, once they become available to the cognitive component, are
relevant in further processing, and this fact is even more apparent in lan-
guages with a strong grammaticalization of pragmatic phenomena that deal
with the social relationship between the interlocutors in the communicative
situation. Japanese, for instance, attaches a lot of importance to the per-