168 Michael Fortescue
(Chomsky 1993). Having shaken off the ‘superfluous’ baggage of transfor-
mations and the deep-versus-surface structure distinction and thus returned
towards the central nature of the linguistic sign as binding form and content,
minimalists continue their search for ever more economic generalizations as
to the universals of language patterning. Which is fine, as long as neither
they nor we persist in the mistake of thinking this has anything much to do
with real psychological processes. We, as functionalists, however, should not
be complacent in our belief that we, by default, as it were, must be the
guardians of the ‘true’ key to Process. If we are to do so, we must first strip
our process view of language of all its hidden pattern trimmings, just as some
generativists (though perhaps not Chomsky himself) have been willing to
abandon the process, or strategy, interpretation of their evolving models. As
regards the process interpretation of the FG model in particular, I propose
that we limit our representations of sentence meaning to simple layered
structures that take experiential ‘prehensions’ as the baseline.^21
As regards the specific question I started out with, how to handle Focus
in Nootka and where in general it should be positioned in the FG model (on
some higher pragmatic/illocutionary level or in a separate interfacing dis-
course module), I can only propose that Focus should be part of both the
pattern and the process representations of utterances in Nootka, but that
this does not necessarily hold for all languages. (Nor, as mentioned above,
is Focus in the process sense of cumulative etic triggers necessarily equiva-
lent to the corresponding pattern sense of emic coding choice.) In some
languages, like Yukagir, Focus is highly grammaticalized within the pat-
terning of its morphosyntax, and in others it is not grammaticalized at all,
focality being borne solely by the ‘uncoded’ or analogue ‘etics’ of its pros-
ody (Nootka is somewhere in between in this respect).
As to where in the model Focus should be introduced, it looks as if it
needs to be admitted on process representations at the very outset, i.e. at
the outset of ‘concrescences’ that produce determinate utterances from un-
differentiated, holistic ‘intentions’ and work towards greater and greater
differentiation en route towards their final ‘satisfaction’ (= matching of in-
tention and output). Using the standard formalism (that of Dik 1989), this
actually means locating Focus at the lowest predicational level (albeit per-
colated down from the highest, pragmatic level), since choices there may
have an immediate influence on choice of structure at ensuing levels. It
should be borne in mind that process representations are by necessity lan-
guage-specific.
As regards pattern interpretations or representations (i.e. in the tradi-
tional FG model), it really does not matter very much, since ‘everything