re interrogative verbs + interrogative relators + indefiniteordinals

[Matthew Dryer]

I'm not sure how many readers of this list are still attempting to follow
this discussion between David Gil and me, but since I think David's last
response still misses what I see as the crucial point, I am not inclined to
leave the discussion where it is.

My point is that if one considers responses to "how-many" questions of the
form "more than I expected", then David's claim is probably vacuous since
there are probably an unlimited number of responses to any conceivable
class of questions.

David said earlier "I would not rule out the possibility of an interrogative
preposition whose answers, say, could range over a potentially open set
of spatial or temporal relationships.  (So if such creatures really
don't exist, that would need further explanation.)  But I would exclude
the possibility of an interrogative preposition whose well-formed
answers could only range over a closed set of prepositions such as is
characteristic of English and many other languages."

First, since there are, at least as far as we know, no instances of
interrogative adpositions in any language with a closed class of
adpositions, we surely need a principle to explain this independently of
the set of possible answers; in other words we need something to rule out
both possibilities David describes.  Second, if we include answers like
"more than I expected", or in answer to a hypothetical case with an
interrogative adposition, answers of the form "behind the house, but not as
directly behind the house as I expected him to be", then it is not clear
that ALL questions in ALL languages don't allow an unlimited set of
possible answers.  In other words, David's generalization "If a language
has interrogative prepositions, then there will be an unlimited set of
possible answers to such questions" is of the form "If a language has
property P, then a language has property Q", where there are no attested
languages with property P and no attested instances of languages lacking
property Q.  Surely, there is something suspicious about hypotheses of this
form.

I think that in attempting to salvage part of his original suggestion,
David has actually thrown out the baby with the bathwater, that whatever
generalizations are relevant MUST make reference to the syntactic category
of the interrogative word itself and not just the properties of the set of
possible answers.  I would thus predict that while it is possible that we
will not find an interrogative adposition in a language in which
adpositions are a closed class, we are more likely to find an interrogative
word that covers "adpositional meanings" in a language in which
adpositional meanings are expressed by verbs.  In fact, Riau Indonesian on
David's analysis would appear to be an instance of just that.  In fact, I
would also suggest that we are more likely to find interrogative verbs in
languages with weak category distinctions than in languages with strong
category distinctions.

I also think that there is more going on here than just the openness of
classes.  It is my impression, having examined "wh"-questions in over 500
languages, that single words meaning "how many" are much more common than
interrogative verbs, despite the fact that verbs are generally a more open
category than numerals.

Matthew Dryer