Quiet Words

Just to post something, here's a simple conlang I'm working on: Apr29

I describe it as an artlang but it's probably better classified as an engelang, since it exists just to try out some morphosyntax ideas, such as the absence of nouns.

Verb Agreement

Verbs are either univalent or bivalent and verb forms are either finite or attributive.

Univalent attributive forms simply append -u to the stem and univalent finite forms append a person-number suffix. These are the same as the noun possessor suffixes:
K: -k, KP: -ka
L: -s, LP: -sa
M: -p, MP: -pa
3rd person: 0
Impersonal: -n

An /a/ may be inserted as with nouns.

Bivalent verbs have an A-argument and a P-argument. Bivalent attributive forms append a person-number suffix, possibly an inversion suffix -i, and finally -u. The person-number suffix is the same as above with the addition of:
Reflexive: -r

The inversion suffix isn't used with the reflexive suffix; it makes the person-number suffix refer to the A-argument instead of the P-argument. Examples (assuming statements):

kyat tansu (direct) "the cat that sees you"
kyat tankyu (inverse) "the cat I see"
kyat tanru (reflexive) "the cat that sees itself"

Bivalent finite forms append a person-number suffix, possibly an inversion suffix -i, and another person-number suffix. The persons are hierarchical:

K > L > 3 > Unspecified > Reflexive

with the higher of the 2 persons used for the latter suffix and the lower for the former. If needed, an /a/ is inserted before the suffix complex, but never between suffixes. Examples (also assuming statements):

tanask. (direct) "I see you."
tansik. (indirect) "You see me."
tanrk. (reflexive) "I see myself."

Note that if exactly one argument is 3rd person, it will be the former if the other argument is local (i.e. K(P), L(P), or M(P)) but the latter if the other is Unspecified, due to the hierarchy:

tans kyat. (tan-0-0-s) "You see the cat."
tanan kyat. (tan-n-0-0) "The cat sees."

(both examples are direct, hence the additional -0).

The 9C-KL project is inspired by the old 9-phoneme challenge.

9C-KL Phonology

The writing system is based on an "archaic" 9-phoneme inventory rather than the "contemporary" pronunciation. The 9 phonemes are:
Vowels: /a/
Semivowels: /u/, /i/, and /r/
Consonants: /n/, /s/, /p/, /t/, /k/.

When followed by /a/ or /i/, /u/ is transliterated |w|.
When followed by /a/ or /u/, /i/ is transliterated |y|.
When followed by /p/, /n/ is transliterated |m|.

There can't be more than 2 of the same phoneme in succession. Of the consonant pairs, all but the geminates and those beginning with /n/ may occur initially; only the latter plus /sp/, /st/, /sk/ may occur finally (however, I'm considering allowing final /ps/, /ts/, and /ks/ as well).

The writing system recognizes no prosody.

I'm afraid I got tired of the switch-reference and clause-chaining conlang and started on an easy SVO one. It lacks personal agreement, but there are enclitic object pronouns which attach to verb forms. A peculiarity is that verbs have 2 basic stems: one that can take only one enclitic and one that can take 2 enclitics. Besides the primary and secondary objects, there are directionals, which are also enclitic. No verb can take both a directional
and an indirect (primary) object.

The reflexive, passive, and object-omission suffixes go where the 1st enclitic would have gone. There are also inverse and causative forms.

There are 4 verb classes according to how the stems are forms. For 2 of these classes, the 1-enclitic stem is more basic and for another class, the 2-enclitic stem is more basic. The remaining class uses ablaut.

The directionals indicate whether motion is toward, away from, by way of, or to a reference point, that being either the location of the speaker or some other definite location. The last pair are also used for static locations. The directionals can themselves be used as verbs.

Verbs take tense prefixes; for dynamic verbs, these are followed by aspect prefixes.

In G2 (mentioned in the preceding post), relative clauses are internally headed. This means that the head noun of the phrase appears within the relative clause with the cases appropriate for that clause; a resumptive pronoun immediately following the relative clause has the case appropriate for the matrix.

Originally I had the head noun always appear first in the clause, since the possessor or the whole are usually relativized rather than the possessum or the part (genitives and partitives precede what they modify). Example

house* in man RP died John sold.

However, this made sentences like this:

dog cat mouse bread ate RP caught RP chased RP I saw.

So now I have the phrases containing the head noun appear right before the verb of the relative clause:

bread mouse ate RP cat caught RP dog chased RP I saw.

I'm not sure if it solves all the problems though.

* I don't have articles yet and am not sure where they would go.

* I worked a little bit on E2 today; otherwise I've only been working on the font for it.

* What I have been working on is GEL "Generic European Language", which is supposed to have typically european features without being related to any existing languages.

* The GEL morphosyntax has been outlined, but I'm a little stuck filling in the details. For example, I need a number of declensional stems and around 10 endings. This last is because even though there are only 6 cases, they result from combining with additional cases, such as genitive-partitive, dative-locative, accusative-ablative, and instrumental-ergative.

* There are also historical changes in the TAM system, mainly involving grammaticalization.

The almost current version of project E1/Toilan has been uploaded; the writing system (an abugida) is at:

http://qiihoskeh.conlang.org/cl/o/Toilan/E1Write.htm

The language features a small inventory: 7 consonant phonemes (p t k s m n l) and 4 vowel phonemes (i e a o). Phonological syllables are (C)V(C) with limited codas, while the morphology uses ((C)C)V plus word-final /n/ possible.

Only the imperative mood has person-number marking and there's no formal agreement. There are 4 cases: agentive, patientive, possessive, and partitive. Grammatical voice is marked only on attributive forms; otherwise only argument order and omission are needed.

I've done some more work on TIAL, Ampkohlaš, and (indirectly) the TriCons4 writing system. The Ampkohlaš formal writing system is briefly described at

http://qiihoskeh.conlang.org/cl/o/Pkohla/PKWrite.htm

The work left to do on TIAL is mostly tinkering, clean-up, and vocabulary, plus writing "Learn TIAL" (not counting translating that and the reference grammar into some other natlang).

Ampkohlaš (and the others) need a lot more work, especially documenting the syntax.

My newest conlang sketch is called Ampkohlaš [amp'.ko?.laS]. It currently has vowel harmony (more or less), hierarchical agreement, up to 3 arguments marked on the verb, evidentials, and temporal distance. The name actually means

"I told you that earlier today."
am-p-k-o-hla-š
today-1S-2S-Dir-tell.Aor-3Inan
with 1st person participant evidentiality implicit.

Everything could change, though, and suddenly, the o- doesn't look right ... I see, it should be

am-p-ko-0-hla-š
today-1S-2S-Dir-tell.Aor-3Inan

Parts of the Q-OVS (Ŋovasi) grammar are now online at:

http://qiihoskeh.conlang.org/cl/o/OVS/O-Intro.htm