Thomas Wier wrote:
> From:    Herman Miller <[log in to unmask]>
>>Isn't that just a problem with trying to read more into a sentence than
>>the words imply? You can't tell if all of the dogs were seen at the same
>>time and place or one after the other in different places. Why would you
>>expect to be able to tell if each man saw the exact same three dogs, if
>>this is unspecified? If it's a significant fact, it can be expressed as
>>"every man saw the three dogs", or to be extra clear, "every man saw the
>>same three dogs". English has many problems, but trying to find one in
>>"every man saw three dogs" when there are much more radical problems in
>>other areas seems a bit strange. 
> Well, it isn't just a trivial problem.  Of course context will usually 
> clarify which of the two scopal readings is possible, but not always.

Why just two? They could just happen to be the same three dogs, or one 
or more of the dogs could be different and the others the same, or they 
could all be different in each case. I don't see how leaving out this 
particular detail is any different from leaving out any other detail.

> That's a crucial point, because it indicates that pragmatic implicatures
> are formally distinct from the truth-conditional semantic values a given
> utterance may have.

I'm afraid I don't have the slightest clue what an "implicature" might be.

   And besides that, the goal of linguistics is to
> describe what it means to "know" a language, and without any such context,
> English speakers will generally readily agree on these judgements as 
> something they know, and they will not accept other possible judgements
> (such that one or the other quantifier obligatorily takes wide scope).
> The point I was trying to make, in re idioms, was that languages differ
> on precisely such points:  languages with nominal scrambling like Japanese
> IIRC do not get the same scope ambiguities that English does.

Could you give an example from Japanese (or some other language if 
you're more familiar with one) to illustrate what you're trying to get 
at here? I'm not familiar with this issue as it applies to Japanese. I 
think it could be interesting to include a requirement like this in one 
of my nonhuman languages (which is why it caught my attention), but I'm 
not clear how it would work out in practice.

(There we go with one of those weird English idioms again, I mean "it's 
not clear to me" :-) And "there we go" also deserves mention in a list 
of idioms....)

  So, trying
> to define what is an idiom, and what isn't, is not a straightforward 
> enterprise, since there is a real sense in which any difference between
> languages is "idiomatic".  We cannot just reduce the set of idioms to 
> semantically noncompositional constructions like "kick the bucket", 
> since there are other kinds of purely structural noncompositionality.

I think there's a distinction between "idiomatic" and "an idiom". Idioms 
are expressions that can't be translated directly, but those aren't the 
only idiomatic differences between languages. Your other example of 
"can't seem to find" is a better one. It seems to me that "every man saw 
three dogs" only includes the case of "every man saw the exact same 
three dogs" as one possible meaning because that's one of many possible 
configurations of three dogs, not from some peculiarity of English. I 
wouldn't go so far as to consider that as one of exactly two alternative 

(Actually, now that I think about it, I'm not sure that "every man saw 
three dogs" is really all that likely in the case where each man saw the 
exact same set of three dogs; "every man saw _the_ three dogs" seems 
more natural in that case. The "the" would only be left out if it didn't 
matter that they happened to be the same three dogs....)