Another triconsonantal project

[Man in Space]

So in my setting there's this big Sprachbund where you end up with several families of triconsonantal languages. Here's the latest, because I am a glutton for punishment.

Phonology

Phonemic inventory

/m n nˤ ɲ ɲˤ ŋ/
/p b mb t d nd tˤ dˤ ndˤ tʲ dʲ ɲdʲ tʲˤ dʲˤ ɲdʲˤ k g ŋg q ʔ/
/ts dz ndz tsˤ dzˤ ndzˤ tʃ dʒ ɲdʒ tʃˤ dʒˤ ɲdʒˤ/
/ɸ s sˤ ɬ ɬˤ ʃ ʃˤ x χ~ʁ h/
/l lˤ j w/
/ʜ ʢ/

/i a ə u/

The consonant inventory is fairly large. This is mostly due to the fact that the pre-protolanguage root structure could include a second, reduced vowel that liked to lenite consonants. Additionally, pharyngealization developed from a distinction on the vowels in the pre-protolanguage.

Root structure

Being a (mostly) triconsonantal language, you obviously have roots of the form √C₁C₂C₃. Restrictions on radicals are as follows:

• C₁ can be any consonant. Prenasalized consonants are only allowed in this position.
• C₂ can be any non-prenasalized consonant.
• C₃ cannot be prenasalized, an affricate, or /h/.
• With two exceptions, C₁ and C₂ cannot both be fricatives. The exceptions are:
  • If one of these radicals is one of /s sˤ ʃ ʃˤ/, or
  • If C₁ = C₂ (this is usually due to assimilation with one of the radicals originally having been *ɾ.

Syllable structure

In most situations the maximal syllable is CVC. There are two major exceptions:

• Codas of consonant + sibilant are permissible, and
• Initial onsets are permitted to violate the sonority hierarchy (e.g. nh-).

Verbs

Verbal morphology

Your basic verbal forms are:

I. Base form
1SG C₁VC₂uC₃
1PL C₁VC₂əC₃
2 C₁VC₂C₃a
3 C₁C₂VC₃

II. Intensive
1SG C₁VC₁C₂uC₃
1PLC₁VC₁C₂əC₃
2 C₁VC₂C₃a
3 C₁əC₁C₂VC₃

III. Reflexive/autobenefactive
1SG C₁VC₂C₂uC₃
1PLC₁VC₂C₂əC₃
2 C₁VC₂C₂aC₃
3 C₁VC₂C₂VC₃

IV. Chaos
1SG VC₁C₂uC₃
1PLVC₁C₂əC₃
2 VC₁C₂aC₃
3 C₁iC₂VC₃

V. Causative
1SG VC₁C₁VC₂uC₃
1PLVC₁C₁VC₂əC₃
2 VC₁C₁VC₂C₃a
3 VC₁C₁VC₂C₃

VI. Motion purpose/intent/ask/seek/attempt
(For this form, if the initial radical is a prenasalized consonant, it "decomposes"; otherwise the initial N becomes an echo of the first consonant)
1SG NəC₁C₁VC₂uC₃
1PL NəC₁C₁VC₂əC₃
2 NəC₁C₁VC₂C₃a
3 NəC₁C₁VC₂VC₃

VII. Change of state
1SG dˤVC₁C₂uC₃
1PL dˤVC₁C₂əC₃
2 dˤVC₁C₂aC₃
3 dˤuC₁C₂VC₃

Theme vowels

This language, which is as yet unnamed, features two "theme" vowels that signify the affirmation or negation of the verb (this is what is denoted V above): i denotes the affirmative and a denotes the negative.

Nominalizations

So this is still something I'm working out. However, I do know that, thanks to reduplicative processes in the pre-protolanguage, you end up with stuff like copies of C₁ or C₃ ending up somewhere completely different in the word. I have a few nominalizations, in the singular at least, nailed down for Form I; for illustrative purposes, I'll here use the roots √qtl 'stand, stay, establish, set up' and √txŋ 'read'.

• Your basic Form I nominalizer is C₁VC₃C₂VC₃ (qiltil 's.th. standing there', qaltal 'absentee'; tiŋxiŋ 'reader', taŋxaŋ 'person not of the nobility' < 'illiterate person').
• The Form I locative nominalizer is C₁C₂VC₃C₃ə (qtillə 'location, place where s.th. is').
• The Form I tool nominalizer is C₁VC₂əC₃C₃iC₃ (tixəŋŋiŋ 'written passage, writing').

[bbbosborne]

i don't have much experience with triconsonantal root-based (con)langs but this looks like it is/will be pretty thorough. i'm curious as to what your syntax will look like or be capable of.
also, i think it's a bit interesting you have a marked affirmative and a marked negative. from a quick search i couldn't find any natlang examples -- maybe basis for a small bit of conculture development?

The consonant inventory is fairly large. This is mostly due to the fact that the pre-protolanguage root structure could include a second, reduced vowel that liked to lenite consonants.

other than the /ə/ you already have? (i'm guessing /ɪ/...)

Additionally, pharyngealization developed from a distinction on the vowels in the pre-protolanguage.

can you elaborate on this?

[Ketumak]

Looks good to me, too. Is gemination allowed? And how about adding phonemic vowel length? That might be a nice option for the verbal paradigm, somewhere.

[bbbosborne]

yes, i'm always curious how a triconsonantal root-based language deals with gemination. are there any rules to the root's consonants?

[Man in Space]

i don't have much experience with triconsonantal root-based (con)langs but this looks like it is/will be pretty thorough. i'm curious as to what your syntax will look like or be capable of.

Thank you!

also, i think it's a bit interesting you have a marked affirmative and a marked negative. from a quick search i couldn't find any natlang examples -- maybe basis for a small bit of conculture development?

It was an idea that popped into my head and I thought, I have a way to influence the first vowel, why not have sandhi mess with the negation marker and then erode it so that you just have the vowel left? In at least some of the daughters there'd be the development of a negative marker from some other construction (sort of like how French went from ne. . .pas to, in spoken French at least, just pas).

The consonant inventory is fairly large. This is mostly due to the fact that the pre-protolanguage root structure could include a second, reduced vowel that liked to lenite consonants.

other than the /ə/ you already have? (i'm guessing /ɪ/...)

Additionally, pharyngealization developed from a distinction on the vowels in the pre-protolanguage.

can you elaborate on this?

That /ə/ is not reduced, it's a full vowel.

In the pre-proto stage of the language, root structure was basically

C₁V₁(C₂)(V₂)(C₃)

And your vowels were **i **iˤ **ə **əˤ **a **aˤ **u **aˤ for the full vowels and **ɨ **ɨˤ **ɐ **ɐˤ for the reduced ones. V₁ was always a full vowel; V₂, if present, was always reduced. Eventually, vowel pharyngealization was reanalyzed as a consonantal effect and it became a consonantal feature, hence all the pharyngealized consonants.

Looks good to me, too. Is gemination allowed? And how about adding phonemic vowel length?

yes, i'm always curious how a triconsonantal root-based language deals with gemination. are there any rules to the root's consonants?

Gemination is allowed and is a feature of multiple word forms, including the following verbal forms:

√qtl

III qittil 'he stands, he remains'
V iqqitil 'he stands (s.th.) up, he sets (s.th.) upright'
VI qəqqitil 'he supports, he props up'

And the following Form I nominalizers:

√qtl qtillə 'location, place where s.th. is'
√txŋ tixəŋŋiŋ 'written passage, writing'

are there any rules to the root's consonants?

• C₁ can be any consonant. Prenasalized consonants are only allowed in this position.
• C₂ can be any non-prenasalized consonant.
• C₃ cannot be prenasalized, an affricate, or /h/. There is a marked tendency for C₃ to not be pharyngealized, though pharyngealized C₃ is not completely absent.
• With two exceptions, C₁ and C₂ cannot both be fricatives. The exceptions are:
  • If one of these radicals is one of /s sˤ ʃ ʃˤ/, or
  • If C₁ = C₂ (this is usually due to assimilation with one of the radicals originally having been *ɾ).

And how about adding phonemic vowel length? That might be a nice option for the verbal paradigm, somewhere.

Vowel length is, I think, a bit too much for this stage of the language. I do plan on deriving it into some of the daughters, though; I have a few ways I think I can do this.

[bbbosborne]

are there any rules to the root's consonants?

• C₁ can be any consonant. Prenasalized consonants are only allowed in this position.
• C₂ can be any non-prenasalized consonant.
• C₃ cannot be prenasalized, an affricate, or /h/. There is a marked tendency for C₃ to not be pharyngealized, though pharyngealized C₃ is not completely absent.
• With two exceptions, C₁ and C₂ cannot both be fricatives. The exceptions are:
  • If one of these radicals is one of /s sˤ ʃ ʃˤ/, or
  • If C₁ = C₂ (this is usually due to assimilation with one of the radicals originally having been *ɾ).

oh, pffft. i forgot that was already there. brain fart

[Man in Space]

Nominalizers

Verbal Noun

I C₁VC₂aC₃C₃u
II C₁VC₁C₂aC₃C₃u
III C₁VC₂C₃u
IV C₁C₂VC₃C₃u
V VC₁C₁aC₂C₃u
VI NVC₁C₂aC₃C₃u
VII DVC₁C₂aC₃C₃u

Agent

I C₁VC₃C₂VC₃
II C₁əC₁VC₂VC₃
III C₁C₃VC₂C₂VC₃
IV C₁VC₂C₂iC₃
V VC₁C₁VC₃C₂
VI NəC₁C₁VC₂VC₃
VII DC₁VC₂C₂VC₃ (D is underlyingly /dˤ/, but this changes depending on its context)

Patient

I BC₁VC₂aC₃ (B is underlyingly /b/, but this changes depending on its context)
II BC₁VC₁C₂aC₃
III BC₁VC₂C₂aC₃
IV BC₁VC₂aC₃
V BVC₁C₂aC₃
VI BNVC₁C₂aC₃
VII bdˤVC₁C₂aC₃

Locative

I C₁C₂VC₃C₃ə
II C₁əC₁C₂VC₃C₃ə
III C₁VC₂C₂əC₃
IV C₁C₂VC₃C₃ə
V VC₁C₁VC₂C₃ə
VI NəC₁C₂VC₃C₃ə
VII DuC₁C₂VC₃C₃ə

Tool

I C₁VC₂əC₃C₃i
II C₁VC₂C₂əC₃i
III C₁VC₂C₃iC₃
IV C₁iC₂VC₃C₃i
V vC₁C₁VC₂C₃iC₃
VI NəC₁VC₂C₂iC₃
VII DC₁VC₂əC₃C₃i

[Bessunire]

This looks really interesting. I love the idea of a triconsonental sprachbund.

IV. Chaos
1SG VC₁C₂uC₃
1PLVC₁C₂əC₃
2 VC₁C₂aC₃
3 C₁iC₂VC₃

I am curious to know what this form does, though.

[Man in Space]

This looks really interesting. I love the idea of a triconsonental sprachbund.

Thank you! Yeah, I was kind of thinking, "what if triconsonantal morphology were an areal feature?". I've got a few tricons or putative families thereof—Raholg, Wǫkratąk, something else from a while ago that I can't remember very well…figured I'd dump them all in the same general vicinity.

IV. Chaos
1SG VC₁C₂uC₃
1PLVC₁C₂əC₃
2 VC₁C₂aC₃
3 C₁iC₂VC₃

I am curious to know what this form does, though.

It basically confers some chaotic aspect to the verb. Changing the Conlang Wiki example slightly, think something along the lines of "break" > "fall apart", or "run" > "run amok", or "blow (of wind)" > "blow with gale-force winds", or "think" > "be severely intoxicated".

More nominalizers

Characteristic substance

I atʲC₁uC₂VC₃
II atʲtʲəC₁C₂VC₃
III atʲC₁uC₂C₂VC₃
IV atʲC₁iC₂VC₃
V tʲVC₁C₁uC₂C₃ə
VI aNNuC₁C₂VC₃
VII adˤdˤuC₁C₂VC₃

Resultative or end product

I C₁VC₂C₃uC₃
II C₁əC₁VC₂C₃uC₃
III C₁VC₂C₂VC₃uC₃
IV C₁VsC₂uC₃
V VC₁C₁VC₂C₃əC₃
VI NəC₁C₁VC₂C₃uC₃
VII dˤuC₁VC₂C₃uC₃

Concrete or inanimate

This one's a fun one. It was back-derived from the word atʲu 'thing', which was reanalyzed as ʔatʲuw with the root √ʔtʲw, yielding the form 1a2u3. And because a is the mark of the negative, it's a nominalization of something that isn't whatever √ʔtʲw is. Basically, √ʔtʲw has connotations of "not existing".

I C₁VC₂uC₃
II C₁VC₁C₂uC₃
III C₁VC₂C₂uC₃
IV C₁VsC₂uC₃
V VC₁C₁uC₂C₃ə
VI NVC₁C₁aC₂uC₃
VII dˤVC₁C₂uC₃

Plural nouns

Hoo boy…basically, what happened was that certain words got bleached semantically and came to serve as number-markers, and then you ended up with a triple threat of sandhi, metathesis, and analogy.

Verbal Noun

I tsˤC₁VC₂aC₃C₃u
II tsˤəC₁:VC₂aC₃
III tsˤC₁VC₂C₃u
IV tsˤəC₁C₂VC₃C₃u
V tsˤC₁VC₂C₃u
VI tsˤəNNVC₁C₂aC₃
VII tsˤəlˤːVC₁C₂aC₃

Agent

I mC₁uC₂VC₃
II məNC₁uC₂VC₃
III mC₁uC₂C₂VC₃
IV mC₁VC₂C₂uC₃
V muC₁C₁VC₃aC₂
VI muC₁C₁VC₂VC₃
VII mdˤuC₁C₂VC₃

Patient

I mBVC₁C₂uC₃
II mBVC₁C₂uC₃
III mBVC₁C₂uC₃
IV mBVC₁C₂uC₃
V mbVC₁C₂aC₃
VI mBVC₁C₂uC₃
VII mdˤVC₁C₂uC₃

Locative

I tsˤəC₁C₂VC₃C₃ə
II tsˤəC₁:əC₁C₂VC₃
III tsˤəC₁:VC₂əC₃
IV tsˤəC₁C₂VC₃
V tsˤC₁VC₂C₃ə
VI tsˤəNNəC₁C₂VC₃
VII tsˤəlˤːuC₁C₂VC₃C₃ə

Tool

I tsˤəC₁ːVC₂iC₃
II tsˤəC₁ːVC₂əC₃
III tsˤəC₁ːVC₂C₃i
IV tsˤəC₁ːiC₂VC₃
V tsˤC₁VC₂C₃i
VI tsˤəlˤːVC₁C₂iC₃
VII tsˤəlˤːVC₁C₂əC₃

Characteristic substance

I tsˤC₁aC₂VC₃
II tsˤaC₁C₂VC₃
III tsˤC₁aC₂C₂VC₃
IV atsˤC₁iC₂VC₃
V tsˤC₁VC₂C₃ə
VI tsˤəlˤuC₁C₂VC₃
VII tsˤtˤuC₁C₂VC₃

Resultative or end product

I tsˤəC₁:VC₂C₃u
II tsˤəC₁C₁VC₂C₃u
III tsˤəC₁:VC₂uC₃
IV tsˤəC₁:VsC₂uC₃
V tsˤC₁VC₂suC₃
VI tsˤəlˤC₁VC₂C₃u
VII tsˤtˤuC₁VC₂C₃uC₃

Concrete or inanimate

I tsˤəC₁:VC₂uC₃
II tsˤəC₁:VC₁C₂uC₃
III tsˤəC₁:VC₂uC₃
IV tsˤəC₁:VsC₂uC₃
V tsˤC₁VC₂uC₃
VI tsˤəlˤːVC₁aC₂uC₃
VII tsˤtˤVC₁C₂uC₃

[Man in Space]

A sound rule

Sequences /ij uw/ surface as i or u word-finally or before a consonant. This excludes geminate /jː wː/.

VII dˤuɲdʲgiw 'turn white, blanch, become pale, whiten' < √ɲdʲ-g-w 'white'
VII.RESULTATIVE dˤuɲdʲigwu 'stone ground for brickmaking' > √ɲdʲ-g-w 'white'

I.A.PL mndzˤuli 'bricklayer' < √ndzˤ-l-i 'hold together, fasten' (the implication being that a bricklayer does this with bricks and mortar)

[Man in Space]

Romanization

Because working with strictly IPA gets boring sometimes.

/m n nˤ ɲ ɲˤ ŋ/ m n ṇ nʸ ṇʸ ñ
/p b mb t d nd tˤ dˤ ndˤ tʲ dʲ ɲdʲ tʲˤ dʲˤ ɲdʲˤ k g ŋg q ʔ/ p b mb t d nd ṭ ḍ ṇḍ tʸ dʸ nʸdʸ ṭʸ ḍʸ ṇʸḍʸ k g ñg q ʔ
/ts dz ndz tsˤ dzˤ ndzˤ tʃ dʒ ɲdʒ tʃˤ dʒˤ ɲdʒˤ/ c z nz c̣ ẓ ṇẓ č ž nž č̣ ẓ̌ ṇẓ̌
/ɸ s sˤ ɬ ɬˤ ʃ ʃˤ x χ~ʁ h/ f s ṣ ł ł̣ š ṣ̌ ḫ r h
/l lˤ j w/ l ḷ y w
/ʜ ʢ/ ḥ ṛ

/i a ə u/ i a ə u

Miscellaneous stuff

Pč̣ič̣ḍʸaš > pč̣ič̣ḍʸaš
√č̣ḍʸš 'heat, heat up, cook, burn, brand'

B + NC > mC

B-ṇḍaṇḍḫas > mḍaṇḍḫas

√n-q-tʸ 'look at' > III 'study'
mniqqatʸ 'knowledge, tradition'

Numbers

So in the earliest stage of the language it was quite like Australian languages in which there were only a handful of numbers:

1 əḍḍadʸ
2 mbaḫt
3 łiḫañ
4 əḍḫañ

From these, roots and forms were derived:

əCCaC, CaCC, CiCaC 'numeral'
√ḍḍdʸ 'one, single, lone, alone, only, first, prime'
√mbḫt 'two, pair, double, copy, follow, mimic'
√łḫñ 'three, triple, triad, triangle'
√ḍḫñ 'four, corner, square, rectangle, quadrilateral'

The numbers for 'five' and 'six' were dawul 'hand' and əṇḍəks 'fist', respectively. There were no basic higher numbers; innovation in these areas is a hallmark of the different languages in the family.

[Arzena]

I dig this immensely.