Ast əmu (Ast-Amu 2.0)

[Omzinesý]

My first version of Ast-Amu did not develop to what I wanted it t be. So this is a second attempt. In this version I'm going to take a more diacronic approach.

The goals are basically the same as in the first attempt.

Features
Phonology
- French-style stressing of phrases instead of words
- Allows many consonant clusters
- a small-scale vowel harmony between ɑ and æ

Edit: It maybe appears only at the underlying level.

Morphosyntax
- Some Semitic-style transfixes (not as much as in Semitic)

Edit: Transfixes will be replaced by three-gade ablaut.

- Cases Nominative/Direct, Partitive, and (maybe) genitive cases.
- VSO (relatively strict)
- Verbs agree with subject and object

Edit: There is a problem with the object alternation (partitive vs. nominative). If it's coded in the verb it is just aspect marking in the verb. So ,i is possible that object markers will be just pronominal clitics that are not used beside object NPs.

- Split alignment (pronouns vs. nouns)

Edit: Probably just nominative-accusative.

- Most verbs consist of a simple verb and a coverb* that are phonologically a compound. (I don't know if they can be discontinuous in some constructions.)
- Some simple verbs can also incorporate nouns ('eat'+'coffee' => 'to drink coffee'), which is morphologally like simple verb + coverb, but only some simple verbs do that.

Edit: Not sure about this.

*Coverbs are a word class that always appears with a simple verb. They don't inflect like nouns. They express the specific meaning of the verb, while the simple verb has some very broad semantics.

[Omzinesý]

Old Ast-amu phonology (This may change very much, when the project goes on.)

The phonology of Old Ast-amu is simple. It's easier to start deriving Modern Ast əmu from it.

Vowels
There are five vowels. They have a front-back harmony. A word cannot contain vowels from both front and back groups. *u can appear with both.

Front: *æ *i
Back: *ɑ *ɨ
Neutral: *u

Consonants
*p *t *k
*f *s
*β *ð *ɣ
*m *n
*l *r

Phonotactics
Old Ast-amu phonotactics is very simple. Syllables are (C)V(l/r/n).

Length of consonants or vowels is not phonemic.

Stress
Primary stress was phonemic and could appear on any syllable.
Secondary stresses appeared on the odd syllables before the primary stress but not on the syllable preceding it. After the primary stress, secondary stresses appeared on the thirst fifth etc. syllable from the main stress. (This can change and maybe secondary stresses will be more based on strength or something. It is also possible that I don't even need the concept of secondary stress and vowel reduction is fully determined by allowed and forbidden consonant clusters.)
Stress realized as a tone because at least the partitive marker had its own tone/stress.

Some genrated words (C)V(N)CV(N)(CV(N))/(C)VCV(CV(N)) :
[+ATR]:
urtǽn
fúti
ælrǽrir
ænpænǽn
ðúnæ
nirínæn
lirfí
βúði
ðælír
kullǽl


[-ATR]:
ɑ́ɣʊmɪ
nɪrsʊ́nβʊ
ʊðɪtʊ́n
ɪ́lnɑ
ʊrɑ́ðʊ
mʊfɑ́
pɪlrʊ́
ɑ́rβʊrnʊn
fʊrʊ́
ðɑsʊtɪ́n
ʊ́rkɑ

[Omzinesý]

.

[Omzinesý]

I think I will make the sound changes a bit simpler and not have ATR contrasts.
There simplest way is:

kaká
-> kəká (reduction of unstressed vowels)
-> kəka (loss of word stress)
-> kək (apocope)

[Omzinesý]

Nouns

Old Ast-amu had two cases that are relevant for Modern Ast-amu.

Nominative/Direct was unmarked.
Partitive was marked with i + reduplication of the last consonant. The i has a central variant in back-vocali words. The i preserved its own tone.

NOM nirúkæt
PART nirúkætít

Later, the vowel before the i was lost. Modern Ast-Amu partitive marker is thus an infix. It is the only situation where a word has two full-grade vowels.

NOM nruxt
PART nruxit

hælt 'money'
hælit 'some money'

Nominal morphology is relatively regular.

[Omzinesý]

Modern Ast-əmu phonology (revised)

Ast-əmu has five vowels:

i u
ɪ ə
æ ɑ

Ast-əmu morphology distinguishes between full (i u æ ɑ) and reduced (ɪ ə) vowels. A word can have only one full vowel. All others are reduced ones. It is however possible that a word has only reduced vowels.

Ast-əmu consonants are:

p t k q
f s r̥ ɬ x~χ
v z r l ɣ~ʁ
m n

On the distribution of velars and uvulars

1) If there is /æ/ or /ɪ/ in the environment, only velars are allowed.
2) If there is /ɑ/ or /ə/ in the environment, only uvulars are allowed.
3) If both 1) and 2) appear in the environment, the one following the consonant is prominent.
4) In other environments, either one is allowed.

a) The plosives /k/ and /q/ are distinct phonemes in the context 3).
b) Fricatives ɣ and ʁ as well as x and χ have free variation in the context 3). By most speakers, [ɣ] is however preferred over [ʁ] but [χ] is preferred over [x].

Other allophony
- The closed vowels /i/ and /u/ become mid-closed when adjacent to uvulars.

[Omzinesý]

Now, we get to the verb.

The morphological pattern is:
-2 Subject agreement
-1 mood
0 stem
1 object agreement

Subject agreement and mood prefixes are relatively straightforward.

There are two finite verb stems: the present stem and the past stem. For most verbs, they are VCC, where the vowel has the full grade. The tense difference is distinguished by changes of the two consonants.

uxt 'wants'
uks 'wanted'

Some rules can be derived for groups of verbs but there are quite few verbs in the beginning, so all verbs can be handled irregular.

The infinitive is typically CVC where the vowel has a full grade.

kit 'to want'