FDG and language production 185complex forms that mimic the complexity of lexical formulations (techni-
cally speaking, to be more specific, ...).
Complex acts differ from simplex acts in consisting of one or more
subacts: in (2) above, A 2 (expressed as it’s on fire) and A 3 (expressed as my
hair) are complex acts. One of these subacts carries the pragmatic function
Focus that characterizes the act as a whole. The subacts in question are
subacts of reference (R) or of ascription (T): both reference and ascription
are thus seen as activities carried out by the speaker in interaction. If a
complex act contains only one subact, this subact then necessarily carries
the Focus function (as in A 3 ). Where a complex act contains more than one
subact (as in A 2 ), one of these carries the Focus function.
All the acts and subacts discussed here are initiated and organized at a
pre-linguistic, conceptual level. Whereas simplex acts are sent directly to
the expression component, complex acts, which I take to be definitional
of the more sophisticated communication systems of the human being,
have a more complex history. All complex acts involve the representa-
tional component, and may also involve the lexicon, although not
necessarily so. Thus both (4a) and (4b) result from complex acts, each in-
volving a subact of reference and a subact of ascription, but only (4a)
invokes lexical material:
(4) a. The president smokes.
b. He does.
Hengeveld (this volume) points out that the mapping relation between
the interactional and representational components is not one-to-one. Al-
though there are default correlations between an ascriptive subact and a
predicate (T ↔ f), and between a referential subact and a term (R ↔ p, e,
x), other possibilities are available: ‘predicate nominals’ thus reflect the
mapping of an ascriptive subact to a term (T ↔ x).
A striking characteristic of Hengeveld’s proposal is that the valency
structure rightly posited for the representational component, i.e. (f 1 ) (x 1 ), is
also present in the interactional component: within the communicated con-
tent C, we find the representation ... (T 1 ) (R 1 ) ..., which is explained as
involving ‘the speaker who ascribes properties to entities’. If the interac-
tional component is to be viewed as reflecting production processes carried
out in real time, however, an alternative view suggests itself. Let us assume
that the Focus of each act corresponds to the temporally first cognitive ele-
ment activated in the preparation of an utterance. After all, it is the very
communication of that Focus that justifies saying anything at all. Indeed, in