Tshyak-family scratchpad

[Travis B.]

Note: I previously named this thread "Address", but I figured that was too narrow and I really wanted a general thread for Tshyak-family languages.

I am coming up with an address/register system for Old Zlang, and have put together the following system:

Normal pronouns are used when speaking to people one is familiar with, people who are lower in status than oneself, or people younger than oneself (unless one specifically wants to be respectful).

All address terms are associated with an inalienable possessive determiner, 1st person when speaking of someone else, and 2nd person when speaking of oneself.

I have only found one term for the 1st person which seems appropriate, which is jam "slave" when speaking to someone of markedly greater social rank.

On the other hand, I have come up with a variety of different terms to be used for the 2nd person when speaking in a respectful fashion:

sat moh "paternal uncle", lit. "father brother", for speaking respectfully to a man older than oneself
nay di "maternal aunt", lit. "mother sister", for speaking respectfully to a woman older than oneself
sat "brother", for speaking respectfully to a male similar in age to oneself
nay "sister", for speaking respectfully to a female similar in age to oneself
ŋes "son" or ŋes gak "nephew", lit. "sibling son", for speaking respectfully to a male younger in age than oneself
qa "daughter" or qa gak "niece", lit. "sibling daughter", for speaking respectfully to a female younger in age than oneself
bźat "lord", for speaking respectfully to a male of markedly greater social rank
ndak "lady", for speaking respectfully to a female of markedly greater social rank

I want to figure out some more 1st person forms to use, for speaking respectfully to people whom one would not use bźat or ndak with.

[Travis B.]

I am thinking of making a descendent of Old Zlang that is tonal, combining attributes of onsets, codas, and pharyngealization to generate tone.

The variables to consider are: voiced versus voiceless unaspirated versus voiceless aspirated onsets, pharyngealized versus non-pharyngealized vowels, coda /ɻ/, and glottal (both /ʔ/ and /h/) codas.

I plan on doing a three-way split of tone between unaspirated, voiceless unaspirated, and voiceless aspirated onsets, corresponding to basic low, middle, and high tones. Also, glottal codas will add rising tone to low and middle tones, resulting in low, low rising, middle, middle rising, and high tones. Furthermore, pharyngealization will lower middle or high tone that would otherwise be present. Coda /ɻ/ will result in falling tone on top of existing high or middle tones. This results in low, low rising, middle, middle rising, middle falling, high, and high falling tones.

Along with this there will be considerable simplification of possible syllables. For starters, glottal codas will be lost, leaving behind their influence on tone. Likewise, pharyngealization and coda /ɻ/ will be lost, leaving behind their influence on vowel quality and tone. Likewise, obstruent onsets will be simplified to a two-way distinction between voiceless and prenasalized voiced obstruents.

[Travis B.]

The name for this language will be Middle Kraw Zlang, or natively kaw /kɑw˧/, named after the Kraw valley in which it was spoken. It will be written with the same orthography as Old Zlang, except that fewer possible syllables will be represented.

I have rethought the tonal system for Middle Kraw Zlang. Rather, Old Zlang voiceless consonants will condition high tone and Old Zlang voiced consonants will condition low tone, unless the vowel was pharyngealized, where then pharyngealization is replaced with creaky voice (which counts as a tone), or there was a final /ʔ/ or /h/, which is lost and conditions mid tone.

Old Zlang non-prenasalized voiced obstruents merge to Old Zlang unaspirated voiced obstruents, and in turn Old Zlang prenasalized voiced obstruents become non-prenasalized voiced obstruents. Final Old Zlang /ɻ/ is lost and conditions vowels it follows (all front vowels) to become centering diphthongs.

[Travis B.]

Another change is that nasalized codas are lost, becoming vowel nasalization.

[Travis B.]

Okay, there will be the following tones/registers:

level: low-mid pitch, flat contour, modal voice, corresponding to initial voiced consonants
heavy: initial mid-high, falling, slightly breathy voice, corresponding to initial voiceless consonants
creaky: high, flat contour, creaky voice, corresponding to pharyngealized vowels
stopped: initial mid, rising, short vowel length, corresponding to final /ʔ/
creaky stopped: high, flat contour, creaky voice, short vowel length, corresponding to pharyngealized vowels combined with final /ʔ/
breathy: low, flat contour, breathy, corresponding to final /h/

[Travis B.]

Okay, I figure I needed to create another language (family) which is distinct from the Tshyak-family languages but which exists in the same world with them, to provide a substratum for languages like Middle Kraw Zlang. These languages are known by the derogatory exonym "slave" (Old Zlang jam, from Proto-Tshyak rəgyam) or, slightly less derogatorily, "tribe" (Old Zlang chaw, from Proto-Tshyak kyaw) languages since in Tshyak language-speaking society they formed the lowest rungs of society and were limited to the poorest land, the hills, and so on, but many of the later speakers of Tshyak-family languages are descended from speakers of these languages, since the Tshyak languages had spread onto the land where these languages had been previously spoken.

In many ways these languages are similar to the Tshyak-family languages, tending towards SVO word order, heavy use of serial verbs, and often having a sesquisyllabic word structure (their pre-proto-languages must have influenced each other a good deal in the past), but these are not genetically related to the Tshyak-family languages, and there are some notable differences, such as having a productive prefixing morphology (whereas Tshyak-family languages tend towards suffixing morphology or, for verb aspect, ablaut, with sesquisyllables not being analyzable in morphological terms). Also, these languages are basically tone-register languages, whereas only later Tshyak-family languages acquired tone and register, under their influence.

[Travis B.]

The basic structure of a word in Proto-Rəgyam is an optional presyllable followed by zero or more minor syllables followed by a final stressed syllable, similar to the structure of a word in Proto-Tshyak. In the end it takes the form of (C₁ə)[C₁{ɐ ɪ ɨ ʊ}]C₂({ɾ l})V(V)({p t k m n ŋ}). The consonants found where C₁ is specified before are /m n ɲ ŋ b d ɟ g p t c k ʔ s h ɾ l w j/. The consonants found where C₂ is specified before are /m n ɲ ŋ ɓ ɗ b d ɟ g p t c k ʔ pʰ tʰ cʰ kʰ s h ɾ l w j/. The vowels found in stressed syllables are /æ ɑ ɛ ɜ ɔ e ɘ o i ɨ u æː ɑː ɛː ɜː ɔː eː ɘː oː iː ɨː uː æi æɨ æu ɑi ɑɨ ɑu/. Stressed syllables have tone and register, with breathy low, low, mid, high, and creaky high tones/registers. Note that the vowel /ɨ/ found in minor syllables is lower than the vowel /ɨ/ found in stressed syllables or the endpoint [ɨ] of diphthongs in stressed syllables.

[Travis B.]

The autonym for Proto-rəgyam is *cɨɓo̰k, where *cɨ- is a prefix used for languages and *ɓo̰k means "human". It is politically correct to refer to these languages as cɨɓo̰k languages, but it is common even in academic circles to refer to the languages as rəgyam. (Compare that in most rəgyam languages the exonym for the Tshyak family of languages corresponds to *cɨhɛ̀ɛn, where *hɛ̀ɛn means "enemy".)

[Travis B.]

Proto-rəgyam relies heavily on serial verbs and attributive relative clauses to express what many other languages would express via adpositions and cases. It also has minimal inflectional morphology, with almost all of its morphology being derivational (mind you it does have a good amount of derivational morphology, and some things that may be considered inflectional are probably best considered derivational in the case of Proto-rəgyam such as its marking for voice), and with auxiliary verbs and adverbs being used to express what would expressed via verb conjugation in many other languages. This has the result that things that would be expressed via adpositions or case in other languages do not need separate verbs to link things together when expressed non-attributively, and things that would be expressed via adpositions or case in other languages can have properties characteristic of verbs such as tense, aspect, and mood.

Of course, in many of these respects Proto-rəgyam is quite similar to some Tshyak languages like Old Zlang. This is due to various rəgyam and Tshyak languages having had influence on one another from early on, to the point that they on the surface look quite similar. There are underlying differences, though - Tshyak languages fundamentally express tense and aspect together through ablaut of the verb stem, even though various Tshyak languages then innovated auxiliary verbs which result in more elaborate TAM systems, whereas Proto-rəgyam verbs all default to present imperfective, and auxiliary verbs are used to express other tenses and aspects. Likewise, Tshyak languages tend to express evidentials as part of the verb complex, such that they are best considered inflections (even if they etymologically are not so and even though they are written out as separate words), whereas in Proto-rəgyam evidentials are verbs used as parts of serial verb constructions, and can be used as standalone verbs.

Phonologically, though, Tshyak languages have significant differences from Proto-rəgyam. Tshyak languages have innovated complex systems of sibilants and in general complex consonant inventories, whereas Proto-rəgyam has only one sibilant, two fricatives, no prenasalized consonants, no uvular consonants, and a straightforward system of distinguishing stops and nasals at four POA (aside from implosives, distinguished at only two POA). Conversely, Proto-rəgyam has implosives, which the Tshyak languages lack. WIth regard to vowels, the main distinguishing factor between the Tshyak languages and Proto-rəgyam is that Tshyak languages as a whole lack phonemic vowel length and Proto-rəgyam does and that Tshyak languages do not have "true" diphthongs but rather have vowel-semivowel sequences whereas Proto-rəgyam does (as shown by that Tshyak languages cannot have arbitrary consonants after such pseudo-diphthongs whereas Proto-rəgyam allows the same consonants after them as in any other coda).

[Travis B.]

Both Proto-rəgyam and many of the Tshyak languages, but not Proto-Tshyak, are direct-inverse languages, where the core argument of a transitive main verb complex that is highest on the person/animacy/definiteness hierarchy is assumed to be the agent unless the inverse marker is present, where then the core argument of the main verb that is lower on the person/animacy/definiteness hierarchy is the agent. All the rəgyam and Tshyak languages are either ergative or fluid-S in alignment. Some Tshyak languages have case-marking, whereas other Tshyak languages lack case, as does Proto-rəgyam. Note that even many Tshyak languages that lack case mark which argument is the agent for transitive verbs by having the main verb complex agree with the subject or agent in number; conversely Proto-rəgyam does not mark number on the verb, or on nouns for that matter - it only marks number on pronouns and demonstratives and with actual cardinal numbers.

[Travis B.]

Proto-rəgyam expresses causatives of transitive verb by placing the causer in the agent position, the causee in the object position, and demoting the object to the argument of the coverb sæ̤u, which can be omitted. The causer can also be omitted, whereas to omit the causee requires marking the verb as antipassive.

Proto-rəgyam has two different kinds of resultative construction. The first is just the direct derivation of a state-change verb with the resultative prefix *ʔɪ- from another verb, very frequently a stative verb. The second is the combining of such a verb with an action verb, in a construction that typically takes the form of A action O *ʔɪ-state; these form an atypical (for Proto-rəgyam) sort of serial verb construction, where the object of the main verb is the argument of the state-change verb. Resultative constructions are naturally telic, and when *ʔɪ- is applied to an action verb it specifically indicates a telic action. However, this is optional, and telic actions need not be marked with *ʔɪ-.

[Travis B.]

Proto-rəgyam lacks grammatical number except on pronouns and demonstratives, but has a productive means of forming collective nouns from nouns, through reduplication. Specifically, reduplication operates such that a presyllable or a minor syllable (depending on whether there is another presyllable or minor syllable already) is added before the major syllable that takes the initial consonant of the major syllable, stripped of aspiration or implosiveness, and if the syllable is a minor syllable, the vowel it takes depends on the vowel of the major syllable, with front, central, and back close and close-mid vowels mapping to front, central, and back near-close vowels, and open and open-mid vowels and diphthongs mapping to a central near-open vowel (whereas if it is a presyllable the vowel is simply a schwa).

Examples of this include:

*ɾəɾǣɨ
"children" (note *ɾǣɨ "child")

*sɨɟɐɟǣm
"tribes, peoples, nations" (note *sɨɟǣm "tribe, people, nation")

*ʔɐɲɪɲé
"brothers" (note *ʔɐɲé "brother")

*jəsɨsɨ̄ɨm
"young girls" (note *jəsɨ̄ɨm "young girl")

*ʔɨsʊsṵum
"cities" (note *ʔɨsṵum "city")

[Travis B.]

Proto-rəgyam has evidentials which include an egophoric evidential, which are used in declarative, interrogative, and quoted clauses. The evidentials are:

*mæ̀ egophoric
*sí non-egophoric direct knowledge
*kɑ̤ deductive
*tʰɛ̀ reportative
*ɗṵ assumption
*tōo dubitative

The egophoric evidential is in declarative clauses used to indicate the personal involvement of the speaker; in interrogative clause it is used to ask about the personal involvement of the person spoken to; in quoted statements its used to indicate that the person being quoted indicated that they were personally involved in that spoken of.

Syntatically these evidentials are coverbs, and may occur anywhere in the clause between the main verb and any nominalized subclause; however, they normally follow the direct object. A number of these have other (related) meanings when used as main verbs unto themselves.

[Travis B.]

There are a variety of particles used to introduce subclauses; these include:

*hí attributive transitive relativizer
*lǣ complementizer
*nɑ̰ nominal relativizer
*hæ̀ time relativizer
*ɗi̤ place relativizer
*wɛ̤ truth relativizer
*bō reason relativizer
*jè manner relativizer
*ŋū method relativizer

The time, place, truth, reason, manner, and method relativizers are also interrogative particles for introducing interrogative clauses. However, the nominal relativizer is not used for introducing interrogative clauses; rather wh-movement combined with an interrogative determiner are used for nominal interrogative clauses.

[Travis B.]

Note: *wɛ̤ is used for introducing yes/no questions and when used for relative clauses is similar to the use of "whether" in English, as in "I don't know whether the Packers will make it to the Superbowl this coming year".

[Travis B.]

Proto-rəgyam has ergative-absolutive alignment, which is primarily indicated by how ambitransitivity operates in it (as it has no case). Ambitransitive verbs in Proto-rəgyam are normally ergative in nature; Proto-rəgyam has no passive constructions, and what would be expressed with them in other languages are expressed in it by simply omitting the agent. It should be noted that for serial verb constructions ergative verbs with an omitted agent behave as if there were an agent, but it simply is not stated. Conversely, to omit the object of a transitive verb the antipassive voice is needed.

There are some verbs which, strictly speaking, are intransitive, but whose object is expressed via a separate coverb. These are particularly verbs whose second arguments are normally other verbs which it forms a verb complex together with (e.g. *kɾɛ́ "seem"), and the separate coverb could be interpreted as being a dummy main verb which it together forms a verb complex. Take, for instance

lɨ́ kǣ kɾɛ́ kʰɜ́ sí
DIST.S CLASS seem large DIR
"It seems big."

lɨ́ kǣ kɾɛ́ ɾɑ̤ cʰǣp sí
DIST.S CLASS seem as tiger DIR
"It seems to be a/the tiger."

In the first sentence *kʰɜ́ "large" is used by itself, since it is an intransitive verb. In second sentence the coverb *ɾɑ̤ "as, like" is being used to introduce the argument *cʰǣp "tiger".

[Travis B.]

I have a much-updated page on Proto-rəgyam at https://travisb.neocities.org/conlangin ... index.html.

As Proto-rəgyam has started to solidify, that makes me think about what I should do with Proto-rəgyam daughter languages. Potential changes I could make include:

• I could make reduplication for expressing collectives become reduplication for expressing plurals when some other word implying plurality (e.g. a number, various determiners) are not present.
• I could reduce unstressed minor syllables into more complex initial clusters. Unstressed minor syllables beginning in glottal stops would turn into glottalization of the following consonants.
• I could eliminate (non-implosive) voiced consonants and merge them with tenuis consonants, turning them into an added tone or register distinction.
• I could eliminate some register distinctions, e.g. eliminating breathy voicing, turning it into a vowel distinction through diphthongization. This combined with the previous change, if voiced consonants triggered breathy voice, would allow such diphthongization to be present in syllables with tenuis initial consonants regardless of tone.
• I could add a distinction between volitional and non-volitional verbs in the intransitive, e.g. by deriving the reflexive causative form of the verb (e.g. ék cha hiuka'ala "I rose" > ék cha hal'a "I rose").
• I could have more development of compound words.
• I could front the palatal consonants ɟ c cʰ > dʒ tʃ tʃʰ > dz ts tsʰ.