kinds of semantic ill-formedness/anomaly

[Darryl McAdams]

what varieties of semantic ill-formedness or anomaly are there that the field aims to address in some form? i can think of only a handful, but i'm sure there are more:

1. presupposition failures

   a. nothing satisfies the presupposition (e.g. "the book" in absence of a book)

   b. too many things satisfy the presupposition (e.g. "the book" in presence of multiple books)

2. type mismatches (do these even exist, when the syntax is correct???)

what other kinds can you think of?

[Darryl McAdams]

add to this what usually gets called "selectional" restrictions, where the sort of thing in question can't really be described in the fashion that's used. e.g. the colorless green ideas example. these might be a kind of type mismatch at some level, in certain theories

[Todd Snider]

I guess it depends how broadly you construe "anomaly". And also of course where one decides to draw {the | a} line between semantics and pragmatics.

One might include: failure (in context) to resolve an anaphor or ambiguous scope; accommodation (information mismatch, fixed); (ir)relevance and (un)informativity; and others, I'm sure.

[Darryl McAdams]

definitely also a question of drawing a line between syntax and semantics as well, especially regarding scope ambiguities.

what did you have in mind with (ir)relevance and (un)informativity?

[Todd Snider]

(1)  A: Did you hear? Kevin got a new puppy!

       B: I have a purple umbrella.

B's response in (1) is problematic, certainly for reasons of (ir)relevance. Something's off: one might even call it anomalous. It's not really ill-formed, but it's problematic.

(2) A: I heard they elected a new pope yesterday. What do you know about him?

      B: He's Catholic.

B's response in (2) is relevant, but it's not informative. If A knows what the word pope means, then A knows he's Catholic. B's response is problematic, but not necessarily ill-formed. But there's a mismatch, here, in terms of expectations (about what B should say) and facts (about what B actually said).

[Darryl McAdams]

aha ok, strong gricean things. yes, those are good. very hard tho. hmm.

[Emiel van Miltenburg]

There's also 'anomalies' in the patterning of meaning. An example from Morzycki (2009):

1. a. John is a big idiot
    b. John is a small idiot

(1a) has two readings: (i) a 'degree' reading where John can be any size, but he does have a high degree of idiocy, (i) a physical size reading: John is a physically big person who is also an idiot. (1b) can ONLY mean John is physically small, and on top of that he is also an idiot. No degree reading here (#John has a low degree of idiocy).

[Darryl McAdams]

would you really want to call this an anomaly tho, or just a difference pattern? by anomaly i mean whatever those things are that we indicate by #. certainly we might want to say #1b at the intended low-degree-of-idiocy reading. but is this even really an anomaly, or is it merely "thats not what that means"? i think there's an important difference between #1b and other anomalous sentences in that regard.

[Emiel van Miltenburg]

Taking the dictionary meaning of anomaly ("something that deviates from what is standard, normal, or expected" according to the Spotlight dictionary on my Mac), I'd say that this really is an anomaly. Saying "that's not what that means" is taking the easy way out. It is unexpected that the data are the way they are, and that requires an explanation.

You can explain away everything by saying "that's just not what it means" or "that's just not how those words are used." Not accusing you of doing so, but I fail to see the argument in favor of a distinction.

[Darryl McAdams]

i'm not saying it doesn't need explanation, just that it's not of the same sort of thing as a presupposition failure. there's a big difference between presupposition failures on the one hand, and this on the other. usually we use # to indicate a sort of funniness in a sentence -- we know what it kinda means, or know what it means completely, but something's off, and it just can't be said because something went wrong. this case, on the other hand, is not like that. this case is more, analogous to "John is tall" not meaning "Frank is short" -- there's nothing wrong with it, you can say it all you want, it just doesn't mean this other thing, and that's only "anomalous" on inspection/reflection.