Saturday, January 05, 2008

 

Testing for Context Sensitivity (Part 1)


There's been some further discussion on the thread about Stanley's motivations concerning Cappelen and Lepore's stance on issues about context sensitivity. Here I'd like to discuss one of their tests for context dependence put forward in Insensitive Semantics, since it strikes me as deeply problematic.

Here's their Inter-Contextual Disquotation Test:

Say one wants to test whether some expression e is context sensitive. First, one choses a sentence S containing e, but no other context sensitive expressions (2004: 105fn12). Then S, and hence e is context sensitive only if there is a true utterance of the following:

(ICD) There are (or can be) false utterances of ^S^ even though S (Where ^^ should be read as corner quotes).
(2004: 105)

Cappelen and Lepore are explicit that past and future utterances can be quantified over here (see 2004: 205fn10) Problem then is that S might feature a context invariant expression which denotes a property that an object might possess at one time, but not at another. For example, the following is supposed to show that the first-person pronoun is context sensitive:

(2c) There is a false utterance of 'I'm hungry' even though I am hungry. (2004: 106)

But I don't think this shows anything at all about the context sensitivity of 'I'. For (2c) could be true even on the bizarre assumption that 'I' invariantly denotes me (i.e. Aidan McGlynn), since my being hungry now does nothing to prevent there from being false past, future or merely possible utterances of 'I'm hungry'.

Objection: The schema should really have been understood as this:

(ICD') There are (or can be) false utterances of ^S^ at t even though S at t.

So the idea is there can be false utterances of a sentence S co-temporally with S being the case. That's the real mark of context sensitivity.

Response: This obviously won't work, since sometimes a context dependence expression is sensitive to the time of utterance. For example, take Cappelen and Lepore's discussion of 'now':

'(2d) There is a false utterance of 'Tom is leaving now' even though Tom is leaving now.

Suppose Tom is leaving now. Then, obviously, any utterance of 'Tom is leaving now' made at times other than now, say, a few days into the future when Tom isn't leaving, suffices to establish that the test utterance of (2d) expresses a truth."
(2004: 106)

By the way, notice that my original point could equally have been made with this example - one could repeat just the same reasoning as Cappelen and Lepore offer even if (2d) hadn't contained 'now':

Suppose Tom is leaving. Then, obviously, any utterance of 'Tom is leaving' made at other times, say, a few days into the future when Tom isn't leaving, suffices to establish that the test utterance of (2d) expresses a truth.

Unless I'm missing something, the ICD test is pretty hopeless insofar as the aim was to find a necessary condition for an expression to count as context sensitive.

(By the way, if anyone knows any literature where this or a related point is made about this test, I'd be very interested to know. I'm still catching up on the literature generated by the book).


Labels: , ,


Comments:
Hey Aidan,

I'm sorry, but I'm a little confused by this, which is probably due to inattentive reading. Is the problemthat we can't select a sentence S, such that for target expression e, e is the only context sensitive element in S? So, in your example (1c), the problem is that you've got two context sensitive expressions (the tense and the first person pronoun). If this is the problem can't we neutralize this problem, something along the lines of your (ICD'). Maybe (ICD*) would work?

(ICD*) Even though S is the case, there are false utterances of ^S(e)^ such that every expression ^e'^ in S(e) other than ^e^ receives the interpretation it would in my context.

I don't think that your objection to (ICD') that some expressions (tenses, 'now') are sensitive to the time of utterance would cary over to (ICD*). When it comes time to investigate the tensed elements, we just don't hold fixed their interpretation. Does this help us to apply their test?

Take care,
Bryan
 
Hey Bryan,

That wasn't what I had in mind, so I should clarify things. (ICD*) is actually what Cappelen and Lepore really want. In the post I referenced a footnote on the point that we should ensure that S contains no other context sensitive expressions, but in actual fact that footnote explains that instead we might 'restrict the domain of 'There are utterances' so that additional context sensitive expressions take the same semantic values in the imagined contexts as in the context of use'. So (ICD*) is just a more explicit version of what they had in mind.

So my suggestion wasn't the one you mention, namely that we can't select sentences which have e as their only potentially context sensitive contained expression. Rather, the thought was that it's *too easy* to construct such a sentence S that will validate (ICD), even when the target expression e is context invariant (even by C&L's lights). So, to take another example, even supposing that 'Aidan Neil McGlynn' just picks out me in every context of utterance, there is a true utterance of:

There could be a false utterance of 'ANM is hungry' even though ANM is hungry.

The problem in this example isn't supposed to be that 'hungry' might be context sensitive. Assume it isn't. The problem's just that, as a matter of non-linguistic fact, I am hungry at some times and not at others. And that's enough for the above instance of (ICD) to be true, since we just select an utterance of 'ANM is hungry' made at one of the times when I'm sated.

I'm not sure how the observation that we don't hold fixed the interpretation of tensed elements when those are under evaluation is supposed to help meet this point. My objection is that the test is too slack - the worry is that there's virtually no expression that can't be made to pass this supposed necessary condition for being context dependent. I'm claiming, to take the above example again, that even if 'Aidan Neil McGlynn' unambiguously referred to me in every context of utterance, and 'hungry' picks out the same property in every context, the following would still be true:

There are false utterances of 'ANM is hungry' even though ANM is hungry.

That because C&L allow that this existential can be made true by past, future, and merely possible utterances, and so just pick an utterance made at a time at which I happen not to be hungry. As far as I can tell, (ICD*) doesn't block this argument.

The point made in the discussion of the 'now' case was that C&L have to include utterances made at other times in the domain of 'there are utterances', or else the test wrongly gives back the result that 'now' is invariant. I don't see how letting the interpretation of 'now' vary changes that. Let's look at the relevant instance of (ICD) (and let's assume that the chosen sentence is otherwise free of context sensitive expressions):

There is a false utterances of 'Tom is leaving now' even though Tom is leaving now.

This can't be true, surely, if it cannot be made true by utterances made at other times (when Tom isn't leaving). And that holds without us having to hold fixed the interpretation of 'now'.
 
Hey Aidan,

Your discussion of ICD is certainly
interesting. But as I have interpreted C&L, they are not trying to offer a necessary condition of when an expression is context sensitive. They are just showing data on language use, and they are trying to explain this data. Their point with the ICD, as I have understood it, is to show the difference between expression in the "basic set" and predicate expressions such as "red", "ready", "weighs 80 kg" etc. It's not intended to be a critereon.

See their replies in
Mind and Language 21
and
Philosophy and phenomenological research 73
 
Huh, I'll need to look at those references. But they're utterly explicit in the book that they're trying to offer a necessary condition:

'e is context sensitive only if there is a true utterance of an instance of the following schema for Inter-Contextual Disquotation...'
(2004: 105)

'Unless e passes this ICD test, it is *not* context sensitive.'
(Ibid, emphasis in original)

I'm not sure how to begin to make sense of these quotes if they are not to be read as saying that passing the ICD test is a necessary condition for an expression to be context-sensitive. Any later suggestion otherwise has to be a revision of their stance in the book.
 
Hmm, I think that I'm still confused.

You say,

"I'm claiming, to take the above example again, that even if 'Aidan Neil McGlynn' unambiguously referred to me in every context of utterance, and 'hungry' picks out the same property in every context, the following would still be true:

There are false utterances of 'ANM is hungry' even though ANM is hungry.

That because C&L allow that this existential can be made true by past, future, and merely possible utterances, and so just pick an utterance made at a time at which I happen not to be hungry. As far as I can tell, (ICD*) doesn't block this argument."

So, as I interpret you, the test goes as follows. Target expression e is context sensitive if and only if we can construct a true instance of (ICD*).

You then seem to want to say that 'ANM' passes the test because your example is an instance of(ICD*).

But, I can't see why it's an instance of (ICD*), since there is an expression (or context sensitive element), the tense, such that this element receives a different interpretation in the quoted context than in the disquoted context.

In other words, the following seems true to me if we suppose that 'ANM' is invariant:

ANM is hungry and there is no utterance of 'ANM is hungry' expressing a false proposition such that the context of utterance fixes the same values for all elements contained in this sentence (including tense), with the possible exception of 'ANM'.

So, yeah, I don't see how the example shows how the test (properly understood) is too slack.
Am I missing something here?

It does open an ugly can of worms though concerning isolating the other context sensitive elements of the sentence, which I do think will ultimately be a problem for any use C&L want to make of the test.
 
Aidan,

I don't see what the problem is. As you say C&L give ICD as a necessary condition. But they do not suggest that it is sufficient.

Your examples of Fa then ~Fa only show ICD is not a sufficient condition, but this is not C&L's claim. Do you have an example of a putative context sensitive term that fails ICD? That would put pressure on ICD.
 
Lee - Let's suppose I'm right (which I'm actually far from sure about just now, given Bryan's comments but bracket that for just now). So suppose that every expression passes the test. Then what's the point? What's the point of a test for context sensitivity that context sensitive expressions and context invariant expressions alike pass? I'm not sure why you think such a point, were it to be established, wouldn't put pressure on the test. It's meant to be a useful tool in determining whether an expression is context sensitive; that's all I was challenging.

Two further points. C&L may avoid claiming that the test offers a sufficient condition, but i. they do claim that passing it offers 'evidence' of context sensitivity (as Stellan noted earlier), and ii. they often write as it they really do want to treat it as a sufficient condition - notice, for example, that 'she' is said to have been 'established to be context sensitive' just on the grounds that it passes the test (2004: 106). So while I don't want to attribute the sufficiency claim to them, in practice the seem willing to treat the test as offering a sufficient condition.
 
Post a Comment



<< Home

This page is powered by Blogger. Isn't yours?