Hi all.

I've been having some strange ideas for my conlang Obrenje. In
general, it is head-first, e.g. from <tsan> "mind" and <laj> "fellow"
we can build

<laj tsan> "a kinsman of the mind, a person who thinks likewise"
<tsan laj> "a kindred spirit".

Quantifiers (including some pronouns, like possessives) come before the
head noun, though an Obrenaj would probably tell you that the quantifier
is the head noun in those cases.

<torve> "man/men"
<con torve> "two men"
<xim torve> "my man/men" (lit.: "mine of men", since <xim> is the head
noun according to Obrenje grammar...)

Anyway, my problem arises when it comes to compound words. All
established compound words I've made so far are head-final! Compare:

<sampe> "heart"
<fros> "fire"
<fros sampe> "fire of the heart"
<sampros> "passion"

Similarly, the name of the Obrenje nation is <obrenaja>, although the
translation of "Obrenje nation" would be <naja obrenje>.

Naturally, I could scrap all my old compounds and re-make them to fit
the head-first rule. I wouldn't want to do that, though, since I like
my established compound words!

So I made up the following excuse: Old Obrenje used to have cases and
numeri, allowing for pretty arbitrary adjective sequences in a noun
phrase. The compounds like <sampros> were created at this stage.
Later, Obrenje lost case inflections and adopted a strict head-first
syntax, but didn't change the established compound words. Thus, a
modern speaker of Obrenje will use a regular head-first noun phrase
where no established compound words are available (e.g. <tsan laj>),
but may also create new compound words in direct analogy to already
established ones, (e.g. <kjobaja> "head of hair" from <kjobe> "single
hair", or <ingwaja> "choir" from <ingo> "voice").

Does that make sense from a linguistic point of view? I know it's a
bit forced, but modern Romance languages also mostly have head-first
syntax, although Latin used to be much more liberal.

Opinions, anyone?


-- Christian Thalmann