286 Jan Nuyts
tational dimension in language use (Hengeveld 1989) is not very well mo-
tivated. My objections (which I will not repeat here) partly boil down to an
intrinsic problem with the concept of separate strata in linguistic organiza-
tion correlating with separate language functions (whatever these are).^14
This clearly involves a departure from Dik’s (1986) view that there is a
multiple mapping from functions to forms, such that there is no one-to-one
relation between the two dimensions. I wholeheartedly endorse Dik’s origi-
nal view (cf. Section 3 above), but to go into this matter here would lead us
astray.
More relevant in the present context is that, if the layered system ap-
plies at the conceptual and not at the linguistic level, notions such as
‘speech event’ or ‘narrated event’ (which Hengeveld correlates with the in-
terpersonal and representational dimensions, respectively) are not at stake
anyway. A conceptual structure is neither a narrated nor a speech event, it
is just a coding of derivations from percepts (SoAs), including a marking
of their status (qualifications). That is, a conceptual system, including the
layered system (in a way – see Section 5 below), just represents. To illus-
trate this directly in terms of qualifications: it is hard to see why deontic
modality, i.e. whether a speaker considers something good or bad, or desir-
able or necessary, would be representational, while an evidential category
such as inference would be interpersonal. Both just participate in the con-
ceptual representation of reality, no more and no less.^15
Thus, the layered system in FG to a considerable extent integrates inter-
actional – and correlated with this, discursive – elements, not only in the
rationale offered for the organization of the system, but also in terms of
concepts and notions actually figuring in the system. But none of these be-
long in a conceptual system of layering, even if some of them do relate to
dimensions of the layered system in conceptualization. Let me explain
these cryptic statements in some more detail.
Firstly, the FG layered system fully integrates the matter of illocution-
ary force. This is justified in an account of the behaviour of qualificational
forms in the grammar, since speech act markers clearly interact with and
have properties comparable to expression devices for qualifications. But
illocutionary force as such is not a conceptual category. It is certainly not a
qualification of an SoA: it is a matter of the speaker’s plans and intentions
with an utterance vis-à-vis the hearer in the actual communicative situa-
tion, in relation to his/her deeper (non-linguistic) intentions with respect to
the world (as relevant at the conceptual level). Thus, unlike qualifications
which assess the position of an SoA in the world, illocutionary force is
only relevant in connection with the process of producing linguistic expres-