Re: Construct case and genitive pronouns

On 8/27/05, tomhchappell <tomhchappell@...> wrote:
> 
> 
> In the phrase "Carsten his Birthday"
> "Carsten" is the Dependent, "Birthday" is the Head;
> "his" is a Floating Marker, attached neither to the Head nor to the
> Dependent. If the order is typically (Dependent, Marker, Head) this
> might diachronically evolve into a Genitive Dependent Marker; if the
> order is typically (Head, Marker, Dependent) it might diachronically
> evolve into a Construct-State Head Marker.

So far on this list we've mostly talked about morphologically marked 
construct state vs. a zero-marked absolutive state, and how it might come 
about. (Pretty much as above.) There's two other interesting sources for 
statey-ness that we might also mention. 

The first is the underspecification of some feature on the head, with an 
assumption that the head will "inherit" the feature from its dependent. I 
don't know how idafa originally came about, but this a reasonable guess. 
That's underspecification of definiteness; we played around with the idea of 
underspecification of number earlier. What else might be possible? That'd be 
interesting to consider, but I'm way too tired at the moment. State as 
underspecification of *case*? That'd be pretty strange. I'll have to think 
about that later.

The other is the emergence of a morphologically marked absolutive state as 
an indication that a usually inalienably-owned possession is being talked 
about without regard to any general posessor. For example, in some languages 
it's rare to talk about body parts or familial relations without marking 
them as mine or yours or his or hers. (It's rare to talk about an arm that 
doesn't belong to anyone, and about a mother that isn't anyone's mother.) In 
Smith's Manual of Spoken Tzeltal (available on the internet; search for it), 
he talks about the difficulty in eliciting the unpossessed word for "hand". 
The usual response for a hand that isn't mine or yours or his... "our hand".

So we might have the usual, unmarked form of a noun being the construct 
state -- the possessed one -- while there's a separate, less common version 
for the very marked occasion of talking about hands and mothers "in 
general", such as "Mothers are for comfort; fathers are for play."

This pattern would be a bit strange, however, for *all* nouns. It's weird to 
talk about a hand that doesn't or didn't belong to anyone, but to whom 
belong clouds, centuries, and colors? I don't believe ANADEW, but I wouldn't 
find it strange to find a language with two different state-marking 
paradigms depending on the ownership-tendencies of the noun in question. One 
in which nouns of the first type (body parts, relatives, personality traits, 
etc.) are in construct state unless specifically marked with an absolutive 
affix, and in which nouns of the second type (colors, wild animals, 
earthquakes, etc.) are in absolutive state unless specifically marked with a 
construct affix. (And, if we're being naturalistic, one in which everything 
in between -- houses, horses, hatchets, etc. -- are unpredictably assigned 
to one or the other!)

Hope this doesn't overwhelm you, Carsten, but I find it to be interesting 
food for thought. It's a birthday present of random information!

-- Pat

P.S. While you're on a head-marking kick don't forget about inflected 
prepositions!

Reply