26.01.2013 Views

Presuppositions in Spoken Discourse

Presuppositions in Spoken Discourse

Presuppositions in Spoken Discourse

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

Chapter 2<br />

Julia doesn’t have a brother from the set of worlds where Julia bakes cookies,<br />

rather than remov<strong>in</strong>g it first from all the world, and then tak<strong>in</strong>g only the worlds<br />

where Julia bakes cookies.<br />

In some examples the presupposition is related to the <strong>in</strong>formation <strong>in</strong> the<br />

antecedent of the conditional. For these examples some researchers have argued<br />

(Beaver, 1992, to appear) that a conditional presupposition is actually the correct<br />

prediction, and they have argued that the b<strong>in</strong>d<strong>in</strong>g theory fails <strong>in</strong> that it can’t<br />

account for these conditional presuppositions, for example:<br />

(59) If Fred is a scuba diver, then he will br<strong>in</strong>g his wet suit. 15<br />

Here it seems that it is more likely that we want a conditionalized presupposition,<br />

that is that Fred has a wet suit is cont<strong>in</strong>gent on him be<strong>in</strong>g a scuba diver. However,<br />

these counterexamples all seem to share an <strong>in</strong>tuition that there is a potential<br />

entailment relationship or an anaphoric relationship between the antecedent and<br />

the presupposition, and that this relationship makes the presupposition already<br />

satisfied <strong>in</strong> the antecedent, prevent<strong>in</strong>g the presupposition from be<strong>in</strong>g globally<br />

accommodated.<br />

Intermediate accommodation is also a problem for the satisfaction theory<br />

because <strong>in</strong> this theory there is no mechanism for representation. For example,<br />

consider how updat<strong>in</strong>g proceeds with a conditional of the form S 1 → S 2{p}. If the<br />

presupposition needs to be accommodated <strong>in</strong> the antecedent of the conditional we<br />

run <strong>in</strong>to problems. Accord<strong>in</strong>g to the update rules given <strong>in</strong> (55), c+[ S 1 → S 2{p}] is<br />

the same as c – ( c+[ S 1 ] – c+[ S 1 ] [ S 2{p}]). First, the context is updated with<br />

the <strong>in</strong>formation <strong>in</strong> S 1 , i.e. c+[ S 1]. This results <strong>in</strong> a context we can call c’. Now<br />

our expression is c – ( c’–c’+[ S 2{p}]) and our next step is to update c’ with [<br />

S 2{p}], i.e. c’+[ S 2{p}]. But first we have to make sure that the presupposition p is<br />

satisfied. Because we began with an empty context, we will need to accommodate<br />

p. 16 Local accommodation would be possible because we could revise the second c’<br />

to one that satisfies p. But there is no context <strong>in</strong> our current expression that can be<br />

revised to result <strong>in</strong> <strong>in</strong>termediate accommodation. For <strong>in</strong>termediate accommodation<br />

we need to “back up” to the second c <strong>in</strong> our orig<strong>in</strong>al expression, i.e.: c –<br />

(c+p+[ S 1 ] – (c+[ S 1 ][ S 2{p}]) ) But because contextual update proceeds<br />

<strong>in</strong>crementally we will never know if we need to <strong>in</strong>termediately accommodate until<br />

after we have made a contextual revision that prevents <strong>in</strong>termediate<br />

accommodation. That is, at the po<strong>in</strong>t <strong>in</strong> time where we realize that we need<br />

<strong>in</strong>termediate accommodation our expression is c – ( c’–c’+[ S 2{p}]). And because<br />

we don’t keep a derivational history (which would <strong>in</strong> effect be a level of<br />

representation, someth<strong>in</strong>g CCS argues is unnecessary) we can’t get back to the<br />

15 This example is from Geurts (1999), p. 97, and he attributes it to Manfred Krifka.<br />

16 If p had been part of the <strong>in</strong>formation <strong>in</strong> S1 then p would already be satisfied.<br />

38

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!