200 Peter Harder
of the aims stated in Hengeveld (this volume), even if it differs in the way
it conceives of the three strata.
Taking my point of departure in an example used by Hengeveld, I am
going to illustrate the issue with an example from English grammar, that of
so-called comment clauses.
- The hedged performative and the report of emotional state in
FDG
To demonstrate the kind of complex phenomena that are brought into the
purview of the model by FDG, Hengeveld (this volume: 15) provides an
analysis of the ambiguity between the Spanish sentences (1) and (2):
(1) (Me) temo que Juan esté enfermo
to.me I.am.afraid that Juan is.SUB ill
(2) Me temo que Juan está enfermo
to.me I.am afraid that Juan is.IND ill
Both translate into ‘I am afraid that Juan is ill’, but in the indicative version
the subclause contains the main message, which is reflected in its status on the
representational and the interpersonal levels: the content of the subclause con-
stitutes the content of a referential act made on the interpersonal level:
(A 1 :[DECL(P 1 )Sp(P 2 )Addr(C 1 : [... (R 1 ) ...] (C 1 ))] (A 1 ))
(c 2 )
(CL 1 )This illustrates the advantages of the three simultaneous levels in FDG: the
interactive choice made by the speaker simultaneously affects the morpho-
syntactic expression and the linguistic representation of the coded content,
by inserting a different content (c 2 ) as constituting the main representational
element.
In this respect the model represents an extension of the advantages of
earlier FG versions over a traditional grammatical description. From a tra-
ditional grammar point of view, the difference is simply a difference in
terms of mood, located in a paradigmatic slot in the Spanish verb; and tra-
ditional discussions about indicative and subjunctive mood were conducted
in terms of notional differences such as that between ‘real’ and ‘unreal’,
which were very imprecise and often difficult to apply to concrete cases
without a great deal of ‘semantics’ in the wrong sense of the word.