RT list: Minimalism, Jaszczolt's POL, anaphora

Please help me falsify the following section from my paper on semantic
minimalism:

copyright Alessandro Capone
9. One final note
In this section I will sketch a solution to some of the problems
discussed, while I cannot go into the details of such a solution. It
appears to me that semantic Minimalism seeks to establish an asymmetry
between subject and predicate positions. Subject positions are invariably
subject to pragmatic intrusion, and there is nothing one can do about it.
Furthermore, pragmatic intrusion provides a constituent of the
proposition, whether minimal or at the level of speech act theory. Without
such a constituent there can be no minimal proposition, so Cappelen &
Lepore cannot really provide minimal semantics without such constituents.
Since subject positions are usually positions for reference assignment and
reference assignment is pragmatic (having to resort to a number of
contextual clues), there is no expectation that the subject position can
provide a constituent of thought without referential resolution; and there
is no expectation that there can be a minimal thought without referential
resolution and the assignment of a constituent to the subject position.
The predicate position, instead, has got a different status. It is true
that a predicate is not immune to pragmatic enrichment; but it is not as
dependent on pragmatic intrusion as the subject. Many cases have been
provided to try to show that predicates cannot really furnish minimal
truth-conditions. However, I am not persuaded that these cases are really
against Cappelen and Lepore. Consider the case of ‘The lemon is
yellow’. This may well require pragmatic intrusion, but only at the
level of the subject. We may enrich the proposition up to ‘The lemon’s
peel is yellow’. Since the pragmatic intrusion is required at the level
of the subject, the predicate is not affected. Could we extend a similar
treatment to ‘John is ready’? Presumably, we need to transfer the
pragmatic intrusion from the predicate to the subject. One way to do so is
through an implicit apposition constituent: ‘John [who must take the
exam] is ready’. We could extend this position further by positing a
null prepositional phrase as sub-categorized by ‘ready’. We thus
obtain ‘John [who must take the exam] is ready [for it]. And now we have
obtained a considerable advantage. While I previously thought that
‘ready’ subcategorizes the constituent ‘for that’, which is
fundamentally deictic, the prepositional phrase ‘for it’ is
anaphoric. This means that the constituent [who must take the exam] is to
all effects part of what is said, given that it is indispensible for
anaphoric resolution. This could explain, presumably, why ‘Mary is
ready’ cannot really mean Mary is ready for that, for that and for that.
The enrichment process is constrained and part of its constraints is that
one should make sense of a sentence by enriching the subject first and one
can enrich the predicate by anaphoric resolution through materials
provided as apposition in the subject. Furthermore, one cannot say ‘Mary
is ready and not ready’ meaning ‘Mary is ready for that and not ready
for that’. This presumably follows from the fact that the prepositional
phrase ‘For it’ is anaphoric and thus it would be a contradiction if
the first occurrence of ‘it’ and the second occurrence of it both
referred back to the same constituent. ‘Mary [who must take an exam] is
both ready [for it] and not ready [for it].
If I am right about these data, then we have bumped into the deep question
of constraints to pragmatic enrichment and where they come from. But now
we must ask the following questions. Do all predicates follow the same
pattern as ‘ready’? Why is it that predicates can be enriched only
through anaphoric resolution, while subjects can be enriched through
provision of apposition constituents (that is to say in a really free
way)? Why is that indexicality cannot occur in predicates (unless use is
made of explicit deictic pronominals), while anaphoric pronominals are
tolerated? In reply to the first question, it appears that things are
very much the same with ‘happy’. We cannot easily say ‘Mary was
happy but unhappy’, meaning ‘Mary was happy about this but unhappy
about that’, as we have the same anaphoric pattern as before: ‘Mary
[who found her jewel] was happy about it but unhappy about it. It is also
clear that somehow the anaphoric pronominal incorporates a reference to
the event of finding Mary’s jewel, which compels the semanticist to
incorporate the event in the apposition close through a device such as
‘Mary, who was such that there was an event of finding her jewel…’.
Now I address the question of why pragmatic intrusion in the predicate
appears to be subordinate to pragmatic intrusion in the subject and why it
can occur through anaphora and not through deixis. A tentative answer has
to do with a principle which has been brought to my attention by Jaszczolt
(1999): Do not multiply levels of interpretation. This seems to correspond
precisely with Cappelen and Lepore’s notion of semantic minimalism, in
which intrusion is granted for subjects but not or only minimally for
predicates. In other words, POL (Parsimony of levels of interpretation)
compels us to minimize the loci of pragmatic intrusion and to utilize an
already existing and necessary locus. Since the subject is a necessary
locus of pragmatic intrusion, reference being necessarily a pragmatic
process and the subject being the locus for an NP that is also a
referential position, pragmatic intrusion must occur primarily in the
subject and if pragmatic intrusion occurs in a predicate, then this
process must be subordinated syntactically to the subject position.
Subordination allows to keep the level of pragmatic intrusion in the
predicate to a minimum (anaphoric resolution), while the pragmatic
intrusion occurring in the subject is more radical.
Received on Wed Aug 24 12:02:42 2011