I've been too busy with other stuff to follow this thread carefully
(sorry, Sally!!), but I thought I should reply to Christophe's questions
and comments.

Christophe wrote, in reply to Sally:

>>In the meantime... do I front or not front the interrogative in an
>>OSV/SOV language?
>>
>        Why not just leave the interrogative in place, that's to say where the
>answer should be? I don't have any statistics but it seems that fronting
>the interrogative is mainly an Indo-European feature. Japanese (SOV) does
>very well without it.

Fronting the interrogative is not necessarily a characteristic feature
of Indo-European.  In fact, many of the Indo-Aryan languages are wh-in-situ
(the linguistics term for leaving the interrogative in place, as in
Japanese).

Wh-fronting (i.e. moving the interrogative to the front of the clause) is
actually extremely common among the languages of the world.  Most of the
Austronesian languages have wh-fronting, as do numerous Bantu and Afro-
Asiatic languages, as well as Mayan and Salishan languages.  Whether a
language is wh-fronting or wh-in-situ seems to have more to do with
basic word order than with genetic affiliation.  The pattern seems to
be as follows:

-- If a language is verb-initial (VSO, VOS), then there is an extremely
strong tendency for it to have wh-fronting, at least optionally.  Off
the top of my head, I can't think of *any* verb-initial languages which
don't have wh-fronting, and I've looked at quite a few of them.

-- If a language is strictly verb-final (SOV, OSV), then there is a
somewhat less strong tendency for it to be strictly wh-in-situ.  I've
read about a dialect of Quechua (SOV) which has wh-fronting, but otherwise
all of the strictly verb-final languages I know of (Japanese, Korean,
Turkish, Hindi, etc.) have wh-in-situ.

-- If a language is verb-medial (SVO, OVS), then it can go either way.
SVO languages like English, Spanish, and Finnish have wh-fronting, while
SVO languages like Mandarin and (I think) Yoruba have wh-in-situ.  Then
there are languages like colloquial French, which have both options.

>        BTW, in an interrogative sentence, what is an interrogative? Focus
>(strange, it doesn't give any new information, it asks for one)? Topic
>(also strange, it's not the thing we're talking about)? Because if you can
>determine that, the place of the interrogative will be easier to choose.

Semantically, interrogatives don't really pattern with either Topic or
Focus, but *syntactically* they clearly pattern with Focus.  There are
lots of languages which have a particular position in the clause that
is associated with Focus (usually immediately in front of the verb), and
in such cases the interrogative phrase invariably occupies that position.
This is what we find in Hungarian, for example.  I don't know enough
Hungarian to make up examples with real words, but here's what the
word order looks like:

        John gave Bill-Dat the book
        "John gave Bill the book"
                (neutral order, with "John" = Topic)

        John the BOOK gave Bill-Dat
        "John gave Bill the BOOK"
                ("John" = Topic, "the book" = Focus)

        John WHAT gave Bill-Dat?
        "What did John give Bill?"
                ("John" = Topic)

Here we see that non-focussed, non-topicalised elements occur after the
verb, while focussed elements occur immediately before the verb (after
the topic).  This is also the position where interrogatives go.

You find much the same pattern in Hindi - modulo the fact that Hindi is
verb-final whereas Hungarian is verb-medial.  In Hindi, focussed elements
tend to occur immediately before the verb, while interrogatives occur
either in this position or in their original position:

        John book read-Past
        "John read the book"

        book JOHN read-Past
        "It's JOHN who read the book"

        book WHO read-Past
        WHO book read-Past
        "Who read the book?"

What all this means for wh-questions in Teonaht I'm not sure.  Since
Teonaht appears to be pretty strictly verb-final, the typologically
realistic thing to do would be to have interrogatives occur either in
their original position, or else immediately before the verb complex
(perhaps immediately before tense marker).

Matt.

------------------------------------
Matt Pearson
[log in to unmask]
UCLA Linguistics Department
405 Hilgard Avenue
Los Angeles, CA 90095-1543
------------------------------------