174 Michael Fortescue
tial choices in Japanese watashi wa kono hoo ga ii ‘I like this one’, in which
watashi, which corresponds to the English subject I, is marked as Topic (or
Theme), the inner grammatical Subject is kono hoo (‘this one’) and the
predicate is ii ‘good’. The standard FG treatment must start with basic
predication this one is good and somehow attach (as for) me at a later stage,
say as Theme at the pragmatic function assignment level, even though it is
clearly central to the basic communicational intention (the propositional in-
put to be expressed). In the following French sentence, it seems even more
difficult to decide on the basic predication: j’ai ma voiture qui est en panne
‘my car has broken down’, where the Topic is actually me and the domain
the whole predication (a kind of possessive cleft). Should the possessive
predicate be included in the underlying structure or introduced at the prag-
matic level? It is evidently a matter of focus, involving the presupposition
that the speaker has a car but with the focus actually on the whole sentence
as the essential new information (Lambrecht 1994: 14). Note that thetic,
presentational propositions like this are often expressed with the help of
clefts or dummy locative subject constructions, effectively splitting the
proposition into two – in Whiteheadian terms this corresponds to expressing
the indicative and the propositional prehensions behind a judgment as if
they were separate propositions (they are indeed separate ‘speech acts’), and
it would be natural for a process model to reflect this two-tiered prehen-
sional structure at the deepest (initial) level.- Note that Nakayama’s ‘predication’ corresponds roughly to Dik’s State of
Affairs, but this is for him not a matter of the lexicon at all (Nakayama
1997: 94ff.). - The predicate-based formalism of the original model does need to be ex-
tended, however, to allow for predicate-level restrictors (for the ‘restrictive’
lexical suffixes of Nootka at least) of the kind I argued for in connection
with Koyukon and in line with Hengeveld’s independently motivated pro-
posal for predicate variables (Hengeveld 1992: 31ff.). - The distinct focus construction illustrated in (6a) and (6b) above (which
highlights an argument for essentially contrastive purposes) does at all
events need to be marked – presumably on the pragmatic-function assign-
ment level of the traditional model. - The case is parallel: THETIC or PRESENTATIVE would be part of the
outermost bracketing – reflecting the act of embedding the inner proposition
with its NewTop within a broader contextual nexus – i.e. much the same af-
fect as is achieved by Hannay’s discourse management mode of that name. - The scope of that focus is in part determined by the presence or absence of
the TEL suffix mentioned above. - Dik (1989: 8ff.) himself stressed that much of communication is implicit
and not overtly coded, and this evidently is something that can vary more