Page 1 of 1

Tazaric Scratchpad

Posted: 30 Dec 2014 16:02
by Micamo
Yeah I should be writing for my Mithara blog but I feel like writing this instead. I am the night!

Tazaric (formerly known as Project Steel) is my attempt at a (mostly) isolating conlang; What inflection exists is mostly done through non-concatenative morphology rather than affixes. It's the major lingua franca of the Ngazir desert region, though most who speak it as a second language actually speak a semi-pidginized version called Kodul Tazar, which (among other things) dumps all the non-concatenative morphology entirely. But here I'll only be discussing Maal Diak Wiin, the conservative prestige dialect spoken by the upper classes of Tazar, the language's home city-state.

Boring phonology crap

/p t̪ t c k/
/b d̪ d ɟ g/
/h/
/m n̪ n ɲ ŋ/
/ɾ l w j/
/ǀk ǁk ǃk ǂk/
/ǀg ǁg ǃg ǂg/
/ǀⁿ ǁⁿ ǃⁿ ǂⁿ/

/a e ɜ i ɪ o ɔ u ʊ/
/a̤ e̤ ɜ̤ i̤ ɪ̤ o̤ ɔ̤ ṳ ʊ̤/

All vowels come in two length variants (short and long), and two tones (high and low). Rising and falling contour tones exist on bimoraic syllables, but not on monomoraic ones. The maximum Tazaric syllable is C(i,u)V(:)(p,t̪,t,c,k,m,n̪,n,ɲ,ŋ); Most roots are monosyllabic, and the few that aren't are all monomorphemizations of previously bimorphemic words.

There's coronal consonant harmony between /t̪ d̪ n̪ ǀk ǀg ǀⁿ/ and /t d n !k !g ǃⁿ/, and vowel harmony between /e i o u/ and /ɛ ɪ ɔ ʊ/, where /a/ is neutral.

Nominal Morphology

Here's an important point: All number/case morphology in Tazaric is suppletive. There are common repeating patterns, but no truly regular processes. Loanwords are always number and case-neutral.

Number

About 50% of monosyllabic Tazaric nouns form plurals by shortening/lengthening the stem vowel, and inverting the tone.

bɜ̀r "good.SG", bɜ́ɜ́r "good.PL"
ŋá̤̤á̤r "ant", ŋà̤r "ants"

(Tazaric adjectives in attributive position inflect for number and case, agreeing with their head noun.)

Most of the rest keep the tone constant while changing the length, or keep length constant while inverting the tone. Other, minor patterns also exist. No forms exist where both the singular and the plural have a high tone; Either they both have a low tone (or a contour) or the plural involves a tone change.

Construct State

The construct state is used when a noun doesn't stand alone; Whenever it's accompanied by an adjective, possessor, or when it's the head of a relative clause. Unlike number and case, the construct state form is completely predictable: When the noun ends in a plosive, the construct state form changes the plosive to a corresponding nasal. Otherwise, the construct form is identical.

gùááŋ wé lò̤mè lṳ̀ò̤p áhɔ̀ɔ̀t̪
gùááŋ wé lò̤m=è lṳ̀ò̤p á=hɔ̀ɔ̀t̪
NST\swim 1sg LOC.CONST\pond=3f.3s.POSS SG.GEN\pond DECL.SG=dry
The pond I swim in dried up.

Case

Tazaric has 4 cases: Neutral, Topical, Genitive, and Locative. Neutral is the unmarked form, and is used for non-topical core clausal arguments. The topical case is used for sentential topics, as well as any nouns in the clause whose possessor is coreferent with the topic (the topical case overwriting their usual case):

wá ácí ǂgòò wɜ̤́ɜ̤̀ŋ ɲíúk
wá á=cí ǂgòò wɜ̤́ɜ̤̀ŋ ɲíúk
1sg.TOP DECL.SG=PST give TOP\weng medicine
I gave my Weng medicine. (wá ácí ǂgòò wɜ́ɜ́ŋ ɲíúk "I gave the weng (not mine) medicine")

The genitive is used for possessors. The possessum is marked with both the construct state, and a special clitic which agrees with the gender of the possessor and the number of the possessum:

mʊ́ cí ǂgòò wɜ́ɜ́ŋá ǁkòt ɲíúk?
mʊ́ cí ǂgòò wɜ́ɜ́ŋ=á ǁkòt ɲíúk
2sg.TOP PST give CONST\weng=3m.3s.POSS GEN\boy medicine
Did you give the boy's Weng medicine?

The locative marks, well, locations; It can be used alone or in conjunction with a spatial preposition.

wá ácí wááp ɲíúk cʊ̀ɔ̀
wá á=cí wááp ɲíúk cʊ̀ɔ̀
1sg.TOP DECL.SG=PST put medicine LOC\bowl
I put the medicine in the bowl.

Re: Tazaric Scratchpad

Posted: 02 Jan 2015 18:18
by Baffin
/p t̪ t c k/
/b d̪ d ɟ g/
/h/
/m n̪ n ɲ ŋ/
/ɾ l w j/
/ǀk ǁk ǃk ǂk/
/ǀg ǁg ǃg ǂg/
/ǀⁿ ǁⁿ ǃⁿ ǂⁿ/
I enjoy this. To what extent do fricatives occur allophonically? Also, how standard of a click inventory would you call that? I know next to nothing about them.
Tazaric has 4 cases: Neutral, Topical, Genitive, and Locative. Neutral is the unmarked form, and is used for non-topical core clausal arguments. The topical case is used for sentential topics, as well as any nouns in the clause whose possessor is coreferent with the topic (the topical case overwriting their usual case):

wá ácí ǂgòò wɜ̤́ɜ̤̀ŋ ɲíúk
wá á=cí ǂgòò wɜ̤́ɜ̤̀ŋ ɲíúk
1sg.TOP DECL.SG=PST give TOP\weng medicine
I gave my Weng medicine. (wá ácí ǂgòò wɜ́ɜ́ŋ ɲíúk "I gave the weng (not mine) medicine")
Could you explain this to me in more detail? I've always stuggled understanding topic/comment structures (i.e. never put the time in to learn them.)

I am very intrigued by what you will come up with for the suppletive morphology. Are there actually natlangs where a majority is considered non-concatenative?

Re: Tazaric Scratchpad

Posted: 02 Jan 2015 22:03
by Micamo
Baffin wrote:I enjoy this. To what extent do fricatives occur allophonically?
Minimally; The names "Tazaric" and "Tazar" were conceived back when the language had /s z/ as phonemes but I like them too much to dump them even when the lang lost them; These were always intended to be anglicisms anyway.
Also, how standard of a click inventory would you call that? I know next to nothing about them.
Saner than usual, actually. I based its clicks off of Khoekhoe and Sandawe:

Sandawe inventory:

/ǀ ǁ ǃ/
/ǀʰ ǁʰ ǃʰ/
/ᶢǀ ᶢǁ ᶢǃ/
/ᵑǀ ᵑǁ ᵑǃ/
/ǀʼ ǁʼ ǃʼ/

Khoekhoe inventory:

/ǀ ǁ ǃ ǂ/
/ǀʰ ǁʰ ǃʰ ǂʰ/
/ᵑǀ ᵑǁ ᵑǃ ᵑǂ/
/ᵑ̊ǀʰ ᵑ̊ǁʰ ᵑ̊ǃʰ ᵑ̊ǂʰ/
/ᵑ̊ǀˀ ᵑ̊ǁˀ ᵑ̊ǃˀ ᵑ̊ǂˀ/

I took the 4 POAs from Khoekhoe and the 3 MOAs from Sandawe (removing glottalized and aspirated variants since the language does not distinguish those on normal plosives).
Could you explain this to me in more detail? I've always stuggled understanding topic/comment structures (i.e. never put the time in to learn them.)
Well, syntactically, the topic can be any* noun in the sentence (but usually the subject) raised to the highest position in the sentence, above the fronted verb. Said noun is striped of its normal case and takes the topical case instead. In subordinate clauses (such as relative clauses) the topic position must be empty. The (affirmative) verb is then inflected as either being in Subject topic form (neutral) and Non-Subject Topic (abbreviated NTS); Because their topic position is empty a verb must be. The topical case is also used to mark any noun in the comment that's possessed by the sentential topic. Note that there's a difference between the possessor being co-referent with a topic, and the possessor actually being the topic:

ǁkìt ácí lʊ́ʊ́k wɜ̤́ɜ̤̀ŋ
ǁkìt á=cí lʊ́ʊ́k wɜ̤́ɜ̤̀ŋ
TOP\boy DECL.SG=PST kill TOP\weng
The boy killed his own weng.

ǁkìt ácé wó lɔ́ɔ́k wɜ̤́ɜ̤̀ŋ
ǁkìt á=cé wó lɔ́ɔ́k wɜ̤́ɜ̤̀ŋ
TOP\boy DECL.SG=NST\PST 1sg NST\kill TOP\weng
As for the boy, I killed his weng.

Pragmatically, the topic slot is used for the most important referent in the sentence that has already been introduced into the discourse; They're what the sentence is "about" in a sense. Note that the topic is optional if the sentence only involves new referents;

ɪ́cɪ̀ɜ̀ŋ ɟáán ná kóò
ɪ́=cɪ̀ɜ̀ŋ ɟáán ná kóò
NT\DECL=NST\like John NST\eat PL\rabbit
John likes eating rabbits. (an exotic delicacy where the language is spoken)

EDIT: I should mention that "any noun in the sentence" refers to monoclausal sentences: You can't extract NPs out of a subordinate clause in the sentence, unless the subordinate clause is an argument or a modifier to an argument of the verb(s) and the entire clause is being brought upward for the ride, leaving the intended topic in-situ:

*d̪à̤n̪ ácé wó ŋɪ̤̀r lṳ́mé̤ ǁnù cí yàrʊ́
d̪à̤n̪ á=cé wó ŋɪ̤̀r lṳ́mé̤ ǁnù cí yàr=ʊ́
TOP\sheriff DECL.SG=NST\PST 1sg NST\shoot dog because PST kill wife=1sg.POSS
As for the sheriff, I shot the dog because he (the sheriff) killed my wife.
I am very intrigued by what you will come up with for the suppletive morphology. Are there actually natlangs where a majority is considered non-concatenative?
Well there are many languages in east africa that use non-concatenative morphology to mark negation, case, number, aspect, etc., but usually only one or two of these per language. Tazaric is an attempt at a mash-up of those different langs' morphologies to combine all of these things at once, and do much more with non-concatenation than any natlang probably does. Plus whatever ideas for a primarily analytic lang I happen to find cool or interesting.

Re: Tazaric Scratchpad

Posted: 04 Jan 2015 17:57
by Baffin
What are the rules for marking case?

Any verb morphology you're willing to speak on?

Re: Tazaric Scratchpad

Posted: 04 Jan 2015 19:06
by Micamo
Baffin wrote:What are the rules for marking case?
Well, all case morphology in Tazaric is suppletive; There are common patterns but no rules per se. As for what those patterns are... I'm not 100% decided yet. But syncretic case forms are quite common; Many nouns (even in Maal Diak Wiin) lack a distinctive genitive or locative form. About 70% of nominals with a distinctive locative case form this case by overwriting the stem tone with a low tone:

gʊ́ʊ̀m "fire.NEU" / gʊ̀ʊ̀m "in the fire"

The most consistent type of genitive marking is vowel lowering:

ǁkìt "boy.NEU" / ǁkòt "of the boy"
Any verb morphology you're willing to speak on?
Verb morphology is more regular than nominal morphology, but not by much. The important inflections and derivations are:

Non-Subject Topic
Topic Plurality
Causative
Benefactive
Instrumental
Verbal noun

A full conjugation of a verb kóm "eat" is given below:

--- | SG | NST.SG | PL | NST.PL | VN
Basic: kóm | kám | kòm | kàm | kómó
CAUS: kóóm | káám | kòòm | kààm | kóómóó
BEN: kó̤m | ká̤m | kò̤m | kà̤m | kó̤mó̤
INST: kóòm | káàm | kóòm | káàm | kóòmóò

Note that the benefactive and instrumental forms are not true applicatives; Their complement is still in a structurally oblique form:

wá ákóm kómó ɲɪ̤́ɜ̤̀ŋ
wá á=kóm kómó ɲɪ̤́ɜ̤̀ŋ
1sg.TOP DECL.SG=eat eat.VN weng_mucus
I'm eating nyieng. (A protein-rich mucus secreted by a weng to feed its young; Think cow's milk except with a very different glandular mechanism.)

wá ákóòm kóòmóò ɲɪ̤́ɜ̤̀ŋ kɜ́ !nɔ́ɔ́m
wá á=kóòm kóòmóò ɲɪ̤́ɜ̤̀ŋ kɜ́ !nɔ́ɔ́m
1sg.TOP DECL.SG=INST\eat INST\eat.VN weng_mucus PREP spoon
I'm eating nyieng with a spoon.

wá ákó̤m kó̤mó̤ ɲɪ̤́ɜ̤̀ŋ kɜ́ ǁkìt
wá á=kó̤m kó̤mó̤ ɲɪ̤́ɜ̤̀ŋ kɜ́ ǁkìt
1sg.TOP BEN\eat BEN\eat.VN weng_mucus PREP boy
I'm eating nyieng for the boy.

Re: Tazaric Scratchpad

Posted: 04 Jan 2015 20:11
by Creyeditor
I really like your click inventory and the breathy contrast in vowels. I also like the vowel harmony, but is it possible, that you made a mistake in the list above?
Micamo wrote:
/a e ɜ i ɪ o ɔ u ʊ/
/a̤ e̤ ɜ̤ i̤ ɪ̤ o̤ ɔ̤ ṳ ʊ̤/
[...]
There's coronal consonant harmony between /t̪ d̪ n̪ ǀk ǀg ǀⁿ/ and /t d n !k !g ǃⁿ/, and vowel harmony between /e i o u/ and /ɛ ɪ ɔ ʊ/, where /a/ is neutral.
Of course I also like the consonant harmony, although I am not sure what to think about the clicks being included.
And a question: Is there interaction between tone and voicing of vowels?

About the morphosyntax:
I always wanted to make a isolating/ablauting language. So short words and so short sentences [:)]
Although the suppletive seems to mess with a lot of established conlangs, I like that a lot.

Also the construct state, whenenver a noun is modified, is very refreshing.

Re: Tazaric Scratchpad

Posted: 04 Jan 2015 20:28
by Micamo
Creyeditor wrote:I really like your click inventory and the breathy contrast in vowels. I also like the vowel harmony, but is it possible, that you made a mistake in the list above?
Micamo wrote:
/a e ɜ i ɪ o ɔ u ʊ/
/a̤ e̤ ɜ̤ i̤ ɪ̤ o̤ ɔ̤ ṳ ʊ̤/
[...]
There's coronal consonant harmony between /t̪ d̪ n̪ ǀk ǀg ǀⁿ/ and /t d n !k !g ǃⁿ/, and vowel harmony between /e i o u/ and /ɛ ɪ ɔ ʊ/, where /a/ is neutral.
Yeah, my bad: /ɜ/ is the correct vowel here.
Of course I also like the consonant harmony, although I am not sure what to think about the clicks being included.
An earlier iteration of the lang excluded clicks from the harmony (and also had ejective clicks) but it felt wrong.
And a question: Is there interaction between tone and voicing of vowels?
Breathy-voiced vowels with a high tone are very rare in the unmarked form of a word (and never occur in uninflected words like adverbs and adpositions), but are merely uncommon as the result of morphological alternations; Breathy stems tend to resist tone shift more often than modal stems.

Re: Tazaric Scratchpad

Posted: 20 Mar 2015 20:13
by Micamo
As part of the "Get off my lazy ass and post more shit about my constuff" initiative:

Relative Clauses:

Tazaric (in Maal Diak Wiin form) implements relative clauses with a hybrid strategy: First, an embedded clause is created that contains the head of the RC in construct form in situ (even if the noun wouldn't be in construct form in a main clause). Then, the embedded clause is optionally possessed by a copy of the head of the RC in genitive form. For the purposes of possessive constructions an RC is feminine gender.

As with all embedded clauses a relative clause is obligatorially topicless, and the declarative particle cannot appear.

Example:

!nìì wó ǂgɪ̀ɜ̀ɜ́m ɲɔ̤̀mɔ́ ɲɪ̤̀ɔ̤́
!nìì wó ǂgɪ̀ɜ̀ɜ́m ɲɔ̤̀=mɔ́ ɲɪ̤̀ɔ̤́
NST\DESID 1s.NEU NST\kiss CONST\girl=3sF.POSS GEN\girl
the girl I want to kiss

It is also possible for the two nouns to be different; If so, the noun in-situ is more generic and the noun in genitive form is more specific:

cé búáàŋ lɔ́ɔ́k wɜ́ɜ́ŋɔ́ gɪ́ʊ́ʊ́m
cé búáàŋ lɔ́ɔ́k wɜ́ɜ́ŋ=ɔ́ gɪ́ʊ́ʊ́m
NST\PST CONST\dangerous_thing NST\kill NEU\weng=3sF.POSS GEN\fire
the fire that killed the weng