202 Peter Harder
comes in when we look at what information is provided by the actual de-
scriptive format. In Hengeveld’s schematic version of the ‘interpersonal
level’, the information reflects the familiar grammar-oriented structuring
principle, covering participants in the discourse seen as ‘arguments’ of the
illocutionary frame. The ‘discourse act’ information quoted above is not
immediately visible; instead we see a format that is tailored to cover the
generic features of declarative sentence types.
Because of that, the interpersonal level suggested for the Spanish in-
dicative version, cf. (2) above, would seem to cover the subjunctive version
as well: an act coded by a declarative sentence with a speaker, an addressee
and a content which involves an act of reference performed by means of
the subclause (which serves as the second argument of the verb temo). The
difference, it appears, is only reflected in the substitution of ‘communi-
cated content’ on the representational level and its relation to the expres-
sion level.
This appears to me less than ideal as an illustration of the aim of inte-
grating discourse structures explicitly into the model. Ultimately, it may be
a matter of notational variants, but it seems to me that if the difference be-
tween the two readings reflects an interpersonal choice, i.e. the choice
between ‘saying what one feels’ and ‘conveying a mitigated piece of bad
news’, the interpersonal level should also be the central point at which the
difference is visible in the model.
‘Hedging’ would appear to be a good example of a subsidiary discourse
act attached to a central act with face-threatening potential (with the two
together constituting a move, according to the description given by Henge-
veld). As such, hedging should be represented in the decision tree for
‘realization of intentions’ (cf., e.g. Brown and Levinson 1987: 60, showing
the decision tree of someone who is considering a face-threatening act).
Therefore the interpersonal level should specify two different choices, per-
haps describable as follows (S = speaker, H = hearer, C = communicated
content, T = ascription, R = reference):
(a) Communicative intention: Relieve one’s mind by entering into ‘trou-
bles talk’, realized by Move: ‘report of emotional situation’, divided
into a main act (A 1 ): ‘S tell H of C = emotional state’ (T = fear, R 1 =
S) + subact (a 1 ): ‘specify object of emotion’ (= R 2 )
(b) Communicative intention: Do one’s duty towards H, realized by
Move: conveying relevant ‘bad news’, divided into a main act (A 2 ):
S tell H of C (T = ill, R = J) + subact (a 2 ): mitigate by showing sym-
pathy ‘S tell H of C (T = fear, R 1 = S)’