Re: head-initial structure

From: "Garrett Jones" <alkaline@...>
> My aim is to make Minyeva completely head-initial. This applies to both
the
> syntax and the morphology. so, the word order is VSO-NAR, compounds are
> head-initial, and it is exclusively prefixing. I have a few questions:
I'm going to answer this one first.

> 4. Has anyone else tried exclusively head-first morphology in their
> language? (head-first compounds & exclusive prefixing).

Yes, in fact it sounds like you and I are on parallel trajectories,
with the exception that, because my language will be relatively isolating,
I am not sure that I am going to have prefixes yet.  As a general design
principle, I decided a while ago that a viable "logical" language would
be greatly enhanced by starting with a phonology, morphology, and
syntax that is as simple and rigorous as possible--more specifically, it
should have as few production rules as possible.  I think a VSO-NAR
syntax of the sort you and I are contemplating goes along way towards
achieving that simplicity.  The only reservation, however, that I have
about VSO topography is that it embeds the subject in the middle
of the clause--and since the subject is usually the most topical
element of the clause, I would truly prefer it to be occur
sentence-initially
(and I suspect that it is for this very reason that a solid majority of
natlangs are either SVO or SOV).  But although I may contemplate
a SVO surface structure in the future, right now my slogan is,
"start simple, you can always complexify later." Thus, VSO.

> 1. How good is the correlation between head-initial compounding and
> head-initial syntax in natlangs (and in which ones do they correlate)?
>
> 2. How good is the correlation between prefixing and head-initial
> compounding in natlangs (and in which ones do they correlate)?
I have wondered this myself and wish I could answer it definitively.
I can say that all the SOV languages I have studied--Latin, German,
and Japanese--employ head-final compounds so it is reasonable
to suspect that VSO languages would be head-initial here.  This
is only speculation.  Actually I can't think of a single language that
regularly forms head-initial compounds.  English has a few like
"scofflaw" and "killjoy", ancient Greek has some like "philosophy"
and there is the occasional oddball like Italian "portfolio", but in
general head-initial compounds seem rare (keyword here: seem).
Hopefully we can get a more expert perspective on this.

> 3. Where should i put determiners, demonstratives, quantifiers, and
> numerals? at the moment i have them all after the adjective, but i'm
> considering putting them all before the noun. Or, i could put
> determiners/demonstratives before and quantifiers/numerals after.
> coffee store owners:
> nidokafigrefka
>
> ni = plural
> do = person associated with
> kafi = store
> grefka = coffee
>
> it makes sense to me that if the plural is prefixing, then quantifiers
would
> also occur before a word. (doesn't necessarily happen in natlangs though,
> like english: all car-s)
It makes sense to me as well. Because the plural marker is really nothing
more than a quantifier with the meaning "more than one", it should, I think,
regularly appear wherever quantifiers are supposed to appear--and they
do not need to be redundant either (no need to say the "s" in "three cars").

As for word order, the only thing I myself have decided is to put the
article
in front.  I think several other positions are justifiable, (with of course
the exception it should *not* appear after the relative clause.)

Maybe it's time to dig up a book on Language Universals, and wherever
logic does not have a preference, go with that...

Regards
---   Mike

Replies