Dear Ira , dear all,
I was wondering whether your concern is specifically about the scope of
negation over quantified expressions (all X are not Y -> not all X are Y
or -> no X is Y), or if you were considering the scope of negation in a
wider range of linguistic configurations. Looking at the preservation of
the presuppositional existential quantification (once again!), there are
good examples, I think, which illustrates the pragmatic accomodation of
the scope of negation.
For instance, if you take (1):
(1) Paul's Ferrari isn't in the parking lot.
This can just mean the propositional form:
(1') Paul has a Ferrari and it is not located in the parking lot now
or implicate:
(1'') There is no Ferrari belonging to Paul.
(1') preserves the existential presupposition and has a narrow scope on
the predicate. (1') can very well be seen as a first case of pragmatic
enrichment from a logical form which doesn't determinates negative scope.
(1'') has a wide scope negation and defeats the existential presupposition.
(1'') can in fact be analysed as follows:
explicit premiss: there is no Ferrari belonging to Paul in the parking lot
implicit premiss: if Paul had a Ferrari, it would be in the parking lot
conclusion: Paul doesn't posess a Ferrari
The difference between (1) and Russell's king of France is that it's
quite easy to find a context where (1) could be used to communicate
(1'') as an implicature.
For instance, if Paul pretends to be rich and powerful, the speaker may
want to mock him (you can see this as a case of irony proper, but I
think it's not necessary). A higher-level implicature would then be
something like "Paul is presumptuous / arrogant / pretending to be rich
and powerful..." etc.
I would say that the accomodation of the scope of negation is typically
illustrated with this, since it is necessary that (1) is understood
first as (1') , (1'') being derived optionaly on the basis of (1').
This is just a type of possible analysis. Maybe it is erroneous. And
it's just a data on the basis of which we cannot simply generalize.
Anyway it is very possible to hold that there is no such think like a
scope-specification at the semantic level (Robyn's position vs
Burton-Roberts' position) and, when looking at pragmatic enrichment,
some pragmatic enrichments do depend on other pragmatic enrichments. In
some cases, just like this one, it really seems that the wide-scope
interpretation is indeed a second-level enrichment.
It is very tempting to assume that the "default" interpretation - the
most natual one, the first tempted when the context does not constraints
it strongly - is the narrow-scope one, since, very simply, not only you
can anayse scope-widening as a pragmatic process, but in general, we do
interpret negative utterance as having narrow-scope, and this looks
pretty much like a prefered routine of negative utterance understanding.
In short I would follow your line of investigation.
My answers to your questions are:
a) is likely to be true
b) is likely to be true, but it should be added that, probably, the
narrow scope is not itself part of the linguistic encoding (but
semantics in general disagree with this). It's a first level of
pragmatic enrichment at the explicature level.
The remaining question is the whether in all cases of negative
sentences, we do absolutely need to transit via a representation with
narrow scope. There is no absolutely convincing argument for that
generalization.
I think it's Larry Horn who said that negation is the Rasputin of
propositional calculus.
All the best,
Louis
Nicholas Allott wrote:
> Hello all,
>
> There are some interesting developmental data in the literature (from
> Musolino, Lidz and company) and which my lab has largely replicated
> showing that young children (4 year olds) are more likely to accept the
> narrow scope reading of negation in sentences such as:
>
> All the children are not in the pool.
>
> That is, children interpret this as "No children are in the pool."
> Adults,
> on the other hand, treat this with wide scope as "Not all the children
> are
> in the pool." Robyn Carston, who has been generous in talking this over
> with me, has argued that the linguistic code gives you an unspecified
> scope and it's up to pragmatics to fix its scope.
>
> Now, given the data, I'm ready to go further and argue that the narrow
> scope is the initial interpretation and that wide scope occurs through
> some form of pragmatic loosening. It makes sense to me in light of our
> other developmental data showing how children's initial semantic
> interpretations are prominent before pragmatic (narrowing) influences
> kick
> in. The only difference with the narrow-cum-wide scope adjustment is
> that
> in this case it is surface form that prompts the initial narrow scope
> reading. My questions for the list are as follows:
>
> Is there any reason not to suppose that a) only the narrow scope reading
> is the initial reading and that b)the wide scope reading is the result of
> a pragmatic, and not linguistic-encoding, process?
>
> Ira
>
>
--Louis de Saussure <http://www.louisdesaussure.tk>
This archive was generated by hypermail 2b29 : Wed Feb 16 2005 - 11:20:14 GMT