« Blogging, with altitude | Main | Conceivability and the Alethic Gap »



"When you consider an Escher sentence holistically, it has a relatively clear and determinate meaning. However, when you consider it atomistically and try to build-up the meaning on the basis of compositional principles, everything goes to hell."

makes me think of wave/particle duality. though I'm sure others have thought of that more than I have.

Q. Pheevr

I find the idea of an "Escher construction" problematic. If there is such a template, then how is it acquired? (Granted, we need more data on how frequent Escher sentences are.) And if Escher sentences are indeed semantically well-formed, then why is it that they seem to disintegrate into hamburger upon closer inspection? (Granted, we need more data on how robust this effect is in speakers whose minds have not been warped by the study of semantics.) Other idiomatic expressions don't seem to do this, or at least not in quite the same way: people do notice that "could care less" or "head over heels" are used with illogical meanings, but both the literal and the idiomatic meanings of these expressions are much clearer than the meanings of Escher sentences.

Andy B

I disagree completely ;-).



Well, this is a toughie, and no mistake. So tough that I've nothing positive to contribute, only nitpicks at Marc's suggestion.

I think there is something ad hoc about the suggestion that we understand such sentences compositionally because of an "Escher construction" that goes with an Escher sentence. For instance, the construction that is offered to explain

More people have thought about Escher sentences than I have.

doesn't cover similar Escher sentences:

More people have thought about Escher sentences than Marc has. (no pronoun)
More people complained than I did. (intransitive verb)
More people are cat-lovers than I am. (no verb, just a predicate)
Smarter people have thought about Escher sentences than I have. (different comparative)

I assume that we want a more general construction that can generate not only the specific instance that Marc gives for his example but these others. (I assume such a construction can be given, which is why this is a nitpick.)

However, even granted such a general construction, I'm skeptical that these sentences will actually be analyzed in a way parallel to each other. For instance, I assume that Marc's account would analyze

More people complained than I did


Other people complained more than I did.

But it seems to me that it is better understood as:

More people than I complained (that is, I'm not the only one who complained)

Similarly, I note that his original sentence can also be analyzed as:

More people than I have thought about Escher sentences (I'm not the only one who has thought about them)

which more conservatively preserves the actual grammar of the sentence. Thus, we have two possible analyses for:

More X's [pred] than Y

They are

Other X's [pred] more than Y


More X's than Y [pred]

Why prefer one to the other?

I don't think it will turn out to be a matter of transitive verbs aligning with one analysis and intransitive verbs aligning with the other.

More people have murdered their spouses than O.J. has

seems more readily understood as "More people than O.J. have murdered their spouses," and

More presidents lied than Clinton did

seems more readily understood as "Other presidents have lied more (often) than Clinton did."

I think our readiness to interpret in one way rather than another has as much (or more) to do with the actual content of the sentences as with their grammatical or semantic structure. That is, it seems to me that we interpret these Escher sentences based upon what they are talking about; we pick out the relevant terms and rearrange them to make maximal sense. I think this is one reason why the sentence Pullum noticed:

More people have been to Iraq than I have

baffles. It's not the kind of Escher sentence that "goes down well"at least, it stopped me dead in my tracks when I saw it. That's because neither analysis

Other people have been to Iraq more (often) than I have
More people than (just) me have been to Iraq

in context seems like a good interpretation of what is meant. The best way to interpret that sentence, it strikes me, is

Other people have more first-hand experience of Iraq than I have had

but that's a lot of content to pull out of a theoretical semantic construction.

It doesn't seem unlikely that we have constructions like Marc proposes floating around. But it seems to me that when language goes awry, we need more resources than such templates to explain how we manage to interpret them.

Lisa Alloju

If the meaning of the sentence is pretty clear, then it isn't semantic hamburger; and if it is semantic hamburger, then its meaning isn't all that clear.

The comments to this entry are closed.

August 2011

Sun Mon Tue Wed Thu Fri Sat
  1 2 3 4 5 6
7 8 9 10 11 12 13
14 15 16 17 18 19 20
21 22 23 24 25 26 27
28 29 30 31      
Blog powered by Typepad