These Pretzels

I meant to post a whole bunch of things about the Carleton Semantics Conference, but it looks like that won’t happen. At the risk of sounding perfunctory, let me note at the start that it was an excellent conference. The organisers, lead by Rob Stainton and Stephen Davis, did an excellent job. The conference was in honour of Ernie Lepore, and I think it must have been pretty cool for Ernie to have that many good people turn up, and that many good papers presented, at a conference in his honour.

For the first time I wished I had a laptop with a wireless network connection so I could blog papers as they were being done. Of course, that would have meant publishing before checking what was for attribution or not. At a conference you’d think everything was on the record, but some participants were clearly off-the-record at all times. (And now that I’ve reached the stage where I pay attention to what’s on the record and not, I think I’ve clearly been paying too much attention to this blogging thing.) But that didn’t happen, so now I have to try writing some things up after the event.

Michael Glanzberg presented the paper he and Susanna Siegel wrote on complex demonstratives. The full paper is available online, and it’s well worth reading.

Presupposition and Policing in Complex Demonstratives

They defend the following theory. Any sentence containing a complex demonstrative that F fails to express a proposition if the object “appropriated” by the speaker’s demonstrative is not F. Call such complex demonstratives, and the sentences containing them unfulfilled. The primary argument for this thesis involves the following two tests. If the two tests are failed, they say an utterance fails to express a proposition.

Echo Assessment

An echo-assessment of a sentence S that has already been spoken in a context is a repetition of S (perhaps correcting for occurrences of indexicals like I and you), preceded by either yes or no. In responding to an utterance of [an utterance of an unfulfilled sentence], speakers (who do not have any misapprehension that there is a demonstrated object) are strongly unwilling to make echo-assessments. (pg. 18)

Indirect Speech Reports
If a sentence does not express a proposition, say Glanzberg and Siegel, speakers will not report it homophonically. So by this test, if A’s utterance does not express a proposition because ‘Sidd Finch’ is non-referring, then B’s report should be inappropriate.

A: Sidd Finch signed for the Mets
B: A said that Sidd Finch signed for the Mets.

The hypothesis then is that utterances containing unfulfilled complex demonstratives pass these tests, and so fail to express propositions. Let’s test this with a particular example. (Not, I should add, Glanzberg and Siegel’s – they keep a slightly higher class of example.) Kramer is snacking on some things in a small bowl on the bar. Sadly for Kramer, the bowl is an ashtray, and he’s snacking on cigarette butts. He thinks, somehow, that he’s eating pretzels. Naturally, he’s not feeling too good after a while, so he utters (1).

(1) These pretzels are making me thirsty.

We can accept that the things Kramer is eating are making him thirsty. And he intends to denote the things he’s eating by ‘these pretzels’. But could we say either (2) or (3)?

(2) Yes, those pretzels are making him thirsty.
(3) No, those pretzels are not making him thirsty.

Glanzberg and Siegel say no, and this sounds reasonable. The indirect speech report test asks whether (4) would be a decent report of Kramer’s utterance.

(4) Kramer said those pretzels are making him thirsty.

I think (4)’s a little bad, so it looks like (1) passes that test as well. Does that mean that (1) fails to express a proposition? No, because the tests return too many false positives to be comfortable with such a conclusion. Jeff King and Mike Fara came up with nice examples of false positives, but I won’t repeat their examples. Here’s one example to add to what they said. Consider Kripke’s example of the person who misidentifies Jones (who he can see in the distance) as Smith, and says

(5) Smith is hunting squirrels.

Assume Jones really is hunting squirrels. Would we happily echo-assess (5)? I think not. Would we happily report this by saying He said that Smith is hunting squirrels? Again, I think not. Such a report sounds exactly as bad as (4) to me. But (5) clearly expresses a proposition, so I think we have a false positive here.

I do think Glanzberg and Siegel’s tests test for something though, just not what they say it is. What’s important in Kripke’s example is that speaker meaning comes apart from semantic meaning. I conjecture that whenever that happens Glanzberg and Siegel’s tests will be passed. So what their information shows is that the semantic meaning of (1) is not the same as its speaker meaning. You might think this is a boring conclusion, since the speaker is obviously confused about something. But it might not have been obvious that the fact the speaker was confused about was semantically relevant, and I think this is some evidence that it is.

Here’s another test for whether a sentence expresses a proposition.

T-schema test
An utterance of S does not express a proposition in a context if an utterance of “S” is true iff S would have been defective in that context. (Assume I have corner quotes here if that’s the kind of thing you worry about.)

This returns the nice results that lexically defective utterances (those containing non-words) do not express propositions, and that utterances containing racist epithets do not either. (You might not think the latter result is nice, but I think the analysis of racist epithets as embodying false presuppositions is attractive, and it suggests sentences containing these terms fail to express propositions.) It also does not say that (5) does not express a proposition, which is nice. But what does it say about complex demonstratives. Well, that’s hard to tell, because as far as I can (6), said by Kramer in the relevant context, is clearly marginal.

(6) “These pretzels are making me thirsty” is true iff these pretzels are making me thirsty.

If my judgment is out here, and (6) is just bad, then I’m inclined to think Glanzberg and Siegel are correct.

Here’s a final question. Glanzberg and Siegel only discuss singular complex demonstratives. But the issue they are interested in arises for plural complex demonstratives as well. How does their thesis apply in those cases? If none of the things Kramer intends to denote are pretzels, then his complex demonstrative is clearly unfulfilled. But what if we change the case so there are no cigarette butts in the bowl, but a mix of pretzels and other snacks, but Kramer still clearly intends to denote all the things he’s eating? What if there are a majority of pretzels? A large majority? This isn’t, perhaps, a question about complex demonstratives as such as about generics, but I think it’s an interesting issue.