166 Michael Fortescue
(with the adverbial satellite, or predicate restrictor if one follows Henge-
veld’s approach, being focused later). This can hardly be called psycho-
logically plausible.
Recall what was suggested above as regards two-tiered thetic proposi-
tions: these too need to be indicated as such at the initial phase of a process
representation.^18 Note also that term variables have been eradicated from
(21b). Moreover, the semantic role Actor has replaced Ag, since these roles
are, according to Nakayama (1997: 90ff.), restricted to Actor and Under-
goer in Nootka. The role Actor is required by the predicative construction
(an intransitive declarative clause) rather than being closely tied to the
lexical predicate itself. The Actor is no doubt the default argument associ-
ated with mamu:k (the default function of which is in turn verbal), but
other arguments, e.g. of place, could optionally have been indicated instead
of the Actor (nominal arguments are not usually marked morphologically
for semantic or syntactic role).
As to how the Nootka sentences we have looked at relate to Hannay’s
specific ‘message management’ types, one could say that (1–5) and (19–
20) at least would correspond to his REACTION mode (presupposing a
discourse context that requires Focus on the first non-term element of the
utterance). (6a) and (6b) would presumably also be in that mode, but with
contrastive Focus on a core term (one of the few constructions in Nootka
where Focus does not lie on the initial constituent). (7–9) could be seen as
reflecting the same mode, but perhaps NEUTRAL is more fitting, since the
initial position of the element concerned is lexically determined, there be-
ing no choice as to what is regarded as the sentential predicate; note that
Focus here has to be indicated in the lexicon as adhering to these words,
since they automatically trigger initial positioning. (10–11) and (13–14)
probably do reflect a choice and thus go with (1–5), but (15–17) are again
probably NEUTRAL, since the positioning of modifiers before heads is
obligatory. (12) and (18), finally, are not relevant here. Is this very helpful?
Certainly not as useful as for English, since Focus in the broad sense of
‘newsworthiness’ is so strongly associated with initial P1 position in
Nootka that it virtually always adheres to that position (except in the Con-
trastive Focus construction), and there is virtually nothing to choose
between – only what particular item in the message is to fill it.^19 In general,
it looks as if the structure of Nootka is so strongly determined by dis-
course-pragmatic factors that there is little to be found that parallels the
complex interaction between such ‘triggers’ and the often conflicting re-
quirements of sentence-internal syntax displayed by English.