208 Peter Harder
of something smaller than sentences. As suggested in Section 1 above, the
problem might also be that from the point of view of online production,
there is no guarantee that the relevant units will match grammatical units
point by point.
- Final remarks
As pointed out in Hengeveld (this volume), there are many problems that
need to be solved in order to have a fully operational Functional Discourse
Grammar. The suggestions I make raise the additional problem of how pre-
cisely to individuate the discourse categories that I would like to place in a
more spectacular position in the interpersonal tier; pending that, it will al-
ways be unclear how much empirical substance there is in the interfaces of
the model. However, problems of that kind need to be faced by anyone
who would like to relate the categories of the code with categories of inter-
action. In spite of the problems, I see the three-tier model as a clear step
forward in the quest to integrate discourse and grammar from a functional
point of view.
The main point I have defended is that if the interface potential of the
model is to be fully realized, it would be an advantage to have a level at
which linguistically represented content is fully specified, to be interfaced
with an interpersonal level focusing more explicitly on interrelations be-
tween discourse elements. In cases where discourse and clausal organiza-
tion go hand in hand, this will mean that information is systematically
duplicated at the two levels – but rather than being a disadvantage, this is
an intuitively obvious way of marking iconicity between discourse and
grammar when it occurs. And in cases where there are discrepancies be-
tween structuring of linguistic content and structuring of discourse, these
can be read straight off the model.
References:
Berman, Ruth A. and Slobin, Dan I.
1994 Relating Events in Narrative: A Cross-linguistic Developmental
Study. Hillsdale NJ: Erlbaum.
Bolinger, Dwight
1968 Postposed phrases: An English rule for Romance subjunctive. Cana-
dian Journal of Linguistics 14: 3-30.