186 J. Lachlan Mackenzie
appropriate circumstances, determined by such matters as urgency, or fa-
miliarity between the interlocutors, the Focus is all that need be
communicated. In the holophrastic speech of infants, too, it is focal infor-
mation that is expressed; it is only as they develop a ‘theory of mind’, i.e. a
social competence, that they come also to express non-focal (typically topi-
cal) information (cf. Levelt 1999: 85). In this view, then, the cognitive
assembly of an act commences with the identification of a Focus: this is the
first (and possibly only) subact.
Associated with any complex act is a fundamental stance: either a giv-
ing stance, where the speaker is offering information or goods/services to
the addressee, or a demanding stance, in which the speaker is eliciting in-
formation or goods/services from the addressee (cf. Halliday 1994: 68–69).
The stance will be determined by the speaker’s initial motivation for lin-
guistic communication, and it is this stance towards the Focus that
determines the basic ‘illocutionary’ status of the act as a whole. Thus a giv-
ing stance correlates with an assertion, a demanding-information stance
with a question and a demanding-goods/services stance with an order or
request. Rather than associating the illocution with the entire act, as
Hengeveld does, I would therefore prefer to indicate the speaker’s stance
on the subact in Focus.
This subact may be supported by further subacts, in line with the
speaker’s awareness of her participation in the joint project of communica-
tion with the addressee. Here she draws upon the sources of knowledge
available in the communicative context (Dik’s ‘pragmatic information’,
1997a: 10) to establish a framework in which her Focus will affect the ad-
dressee in the desired way. The more purely cognitive relations between
the supportive subacts and the Focus subact are familiar from existing
work in FG: the supportive subacts may be Topics of various kinds (Given,
Resumed, etc.) or possibly Settings; other subacts may be inspired by the
desire to reflect the (power) relationship between the interlocutors, such as
those that result in indirect formulations, attitudinal adverbials, etc.
The representation that suggests itself will thus differ in various ways
from that proposed by Hengeveld (this volume) for the interactional com-
ponent. The representation will be incremental, i.e. the left-to-right
ordering of acts and subacts will be designed to reflect actual sequence in
cognition. An outline suggestion for the representation of (1) above would
appear as follows:
(5) (M 1 : (express A 1 : SHOCK)Foc, (assert A 2 : (SA2:1: (T: FIRE)Foc), (SA2:2: (R:
<unspecified, nonhuman>)Top), (assert A 3 : SA3:1: (R: HAIR)Foc)