Agree, position has nothing to do with it, just their higher-level

relationship.  They modify their heads just like other adjectives.
Genitives in construct (as opposed to prepositional genitives) also
function as adjectives, as do relative clauses.  The whole group
modifies nouns, and in Lusitanian they all happen to follow their heads,
as you'd expect with VS word order.  The difference of course is that
nouns are still nouns when not in construct, and there are other clauses
besides relative, but determiners have no function outside of modifying
heads.  Therefore they are adjectives in Lusitanian.

Re how to best fuse endings to stems, since Lusitanian determiners
aren't inflections there's no question of phonological changes.  The
noun either ends in a consonant and can accept a suffixed determiner or
it can't.  In fact all these determiners were independent words earlier
in the language, it's only when nouns began to lose stress to their
possessive determiners that determiners became suffixes in writing.  The
articles, always stressless, just followed suit once the orthographic
change was made with the possessives.  Beyond this, most Lusitanian
determiners have non-determinative functions and in these cases are
still independent words.

As far as derivation goes, it was my question.  I'd like to see how
other conlangs derive, for example, the noun "richness" from adj "rich"
(assuming that's the direction) - do you add the equivalent of "-ness"
to an adjective?  If you don't, how do you determine it, eg "my richness", or
distinguish class from quality, like "the rich", or from definite
reference, "a rich one"?


PS a noun is also closed-stem when its final consonant is organic /w/,
which cannot act as onset before a suffixed syllable.  An example was
the name /ã'drɛw/, where /w/ is organic.   But non-organic /w/,ie /w/
resulting from mutation, reverts back to the unmutated consonant when it
no longer coda.  An example is the noun stem /sɔl-/ "sun", which ends in
the mutating consonant (l).  The undetermined word is /sɔw/ because coda
/l/ > /w/, but add a syllable and /l/ resurfaces, eg /sɔlə/ "the sun".
By contrast no syllable can be added to /ã'drɛw/ in Lusitanian
phonology, so it is closed-stem

On 4/23/2017 1:12 AM, Alex Fink wrote:
> On Sat, 22 Apr 2017 18:55:24 -0400, Doug Barton <[log in to unmask]> wrote:
>> My conlang Lusitanian analyzes determiners as a class of adjectives.
>> Unmarked word order is noun-adj, but unlike other adjectives, Lusitanian
>> determiners don't just follow nouns but are suffixed to them
> So what facts motivate analysing determiners as adjectives, then?  Thatboth of them follow the noun isn't enough to justify that analysis, I don't think, especially when one's a suffix and the other's a freestanding word.
> But on your original question:
>> However many derived nouns,
>> nominalized adjectives for instance, cannot take suffixed determiners
>> because Lusitanian adjectives are morphologically invariable, no
>> particular word shape unlike nouns and verbs which usually end in a
>> consonant.  Instead, the most common way to determine an adjective is
> [via a particle to host the determiner.]
>> Any similar system you know of?
> I don't think derivedness is central to the matter: this is a question about what happens when you have a class of words which for morphophophonological reasons can't take usual inflections.  I can't think of a natlang example where this morphology marks categories like definiteness or possessor, but on the recent thread about case on proper names examples werenoted where names that weren't of inflectable shape just didn't inflect;they were invariant for case.  Your solution, hosting the determiners somewhere else, seems fine too, sort of like what light verbs do for their complements.
> If the problem were only phonological, that your language doesn't like vowels in hiatus, it's probably more common for a language to solve that on the level of phonology.  Delete one of the two vowels, or fuse them (e.g. into a diphthong), or insert some hiatus-breaking consonant.  I guessit's not only phonological for you, given that you say [emphasis added]
>> These same particles are also used to determine and/or pluralize the few
>> "closed-stem" nouns in the language, typically names, which end in a
>> vowel or *otherwise do not allow suffixes*.
> What are these other cases?
> Alex