Håmwo Väh, tongue of the people (Was: A scratchpad)

[Travis B.]

This is a scratchpad for a language I have not named yet (Edit: now named Håmwo Väh, lit. tongue of the people). I am starting off with the phonology, because that is where I am coming up with ideas at the moment, even though I have some idea of what kind of morphosyntax I want (agglutinative verb morphology with polypersonal agreement, SOV word order, split-ergative alignment, combination of a complex case system and relational nouns, verb-like adjectives, etc.)

So to start off with phonology, here is the vowel system:

/a aː ə ɨ/

Yes, that's it. As you would expect, this implies an elaborate consonant system, which goes as follows:

/mʷ m mʲ nʷ n ɲ ŋʷ ŋ/
/bʷ b bʲ dʷ d dzʷ dz dʑ ɟ ɡʷ ɡ/
/pʷʰ pʰ pʲʰ tʷʰ tʰ tsʷʰ tsʰ tɕʰ cʰ kʷʰ kʰ qʷʰ qʰ/
/pʷʼ pʼ pʲʼ tʷʼ tʼ tsʷʼ tsʼ tɕʼ cʼ kʷʼ kʼ qʷʼ qʼ/
/ɸʷ ɸ fʲ sʷ s ɕ ç xʷ x χʷ χ/
/βʷ β vʲ zʷ z ʑ ʝ ɣʷ ɣ ʁʷ ʁ/
/rʷ r rʲ ɫʷ l ʎ/
/j w/

Blue consonants are fronting/palatalizing consonants. Red consonants are backing/rounding consonants.

Their effect on the vowels above is pretty simple. When a fronting/palatalizing consonant is next to a vowel, it gives it a fronted realization. When a backing/rounding consonant is next to a vowel, it gives it a backed and rounded realization, unless it is also next to a fronting/palatalizing consonant, it then has a fronted and rounded realization (aside from /a/, where then the two neutralize each other, resulting in [a]). Uvular consonants have the added effect of backing adjacent vowels, even if they are adjacent to a fronting/palatalizing consonant, without any effect on rounding (they may still be rounded if adjacent to a backing/rounding consonant). The exception is /aː/, which is not affected by adjacent consonants. This gives the allophones [i y ɨ ɯ u] for /ɨ/, [e ø ə ɤ o] for /ə/, [æ a ɑ ɒ] for /a/, and just [aː] for /aː/.

The underlying phonotactics are simple, (C)V(C). However, consonant clusters are frequently reduced after they have their effect on adjacent vowels. For instance, nasals are lost before egressive plosives, deaspirating them and voicing them.

Similarly, non-sibilant voiced fricatives along with /ɸʷ ɸ fʲ/ are lost both before and after any kind of nasal or plosive. Also, ejectives are elided adjacent to egressive plosives, with the effect of lengthening the egressive plosive.

Supersegmentally there is a phonemic pitch accent on one of the syllables of a root; if the root contains at least one instance of /aː/, phonemic pitch accent will be on one of these. Pitch accent consists of a downstep after the accented syllable; if the accented syllable is the last syllable of a word, this will apply to the next word. Otherwise there is an upward drift of pitch within an utterance.

[Travis B.]

The Verb Complex

The verb complex (except in imperative mood) consists of a root (no surprise there), either an ergative/accusative agreement marker, an absolutive/nominative agreement marker, or both (I figured I would make this fluid-S combined with split-ergative after all), aspect/tense markers (the key distinction is between perfective and imperfective (including stative and habitual) aspect; secondarily perfective aspect normally refers to past events while imperfective aspect normally refers to present events - to express imperfective past events, perfective future events, or iterative aspect extra markers are added), modal markers, negation markers, directional markers (think like separable prefixes in German), and interrogative markers.

The split ergativity comes into play when a 1st person or 2nd person core argument is present in a transitive clause. In this case, A is indicated with absolutive/nominative case-marking and agreement and O is indicated with ergative/accusative case-marking and agreement. (If it were not for the case marking, this would not be apparent.)

The overall structure of the verb complex is:

root-voice-lesser agreement-primary aspect-tense-secondary aspect-greater agreement-modal-direction-negation-evidential-mood-interrogative


The voice markers are as follows:

Reflexive	-ça
Reciprocal	-nʷan
Antipassive	-ɣə
Passive	-sʷa
Causative	-sa
Applicative	-(ɨ)βə

The lower personhood/animacy/topicality markers are as follows:

	Sg.	Pl.
1st (excl.)	-la	-mʲə
1st (incl.)	n/a	-tʼɨ
2nd	-rɨ	-ja
3rd (A)	-ma	-tʰə
3rd (B)	-jɨ	-ɕə
3rd (C)	-xa	-wɨ
3rd (D)	-ra	-rʲə
3rd (E)	-mɨ	-na

The primary aspect markers are as follows:

Imperfective (present)	-
Perfective (past)	-jə

The tense markers are as follows:

Past (imperfective)	-tʼəɣ
Future (perfective)	-jamʲ
Future (imperfective)	-wɨ

The secondary aspect markers are as follows:

Iterative	-xaːn
Inchoative	-ɡʷa
Cessative	-rʷɨ
Prospective	-vʲə
Retrospective	-kʰə

The higher personhood/animacy/topicality markers are as follows:

	Sg.	Pl.
1st (excl.)	-tʰa	-kʷʰə
1st (incl.)	n/a	-cʼə
2nd	-nɨ	-tsʰa
3rd (A)	-la	-mʲɨ
3rd (B)	-ɲə	-βʷɨ
3rd (C)	-βa	-ʁa
3rd (D)	-χaː	-mʲɨ
3rd (E)	-mʷɨ	-da

The inverse marker is as follows:

Inverse

-ɲa

The modal markers are as follows:

Necessity	-ɲaː
Obligation	-dzatʰ
Suggestion	-χəm
Possibility	-rʲa
Ability	-pʷʰə
Desirability	-ɟɨ
Desire	-ɣʷa

The directional markers are as follows:

Up	-χa
Down	-ɕɨn
In	-rʷɨχ
Out	-çəmʲ
Over	-mʲa
Under	-nɨʁ
Through	-saɣʷ
Around	-fʲam
Back	-ʁaː
Uphill	-βʷaχ
Downhill	-ɕaːrʷ

The negation markers are as follows:

Negative	-ɕa
Prohibitive (used in the imperative mood)	-kʰaː

The evidential markers are as follows (note they are only marked for indicatve mood):

Direct	-(ɨ)tʰ
Indirect	-(a)n

The mood markers are as follows:

Conditional	-(a)m
Subjunctive	-χɨ

The interrogative markers are as follows:

Yes/no question	-ɫʷa
Wh-question	-ʑɨ

[bradrn]

(agglutinative verb morphology with polypersonal agreement, SOV word order, split-ergative alignment, combination of a complex case system and relational nouns, verb-like adjectives, etc.)

…

/a aː ə ɨ/

…

Blue consonants are fronting/palatalizing consonants. Red consonants are backing/rounding consonants.

I like this conlang already!

an ergative/accusative agreement marker, an absolutive/nominative agreement marker, or both

I’d just call these series ‘agentive’ and ‘patientive’ markers respectively.

The split ergativity comes into play when a 1st person or 2nd person core argument is present in a transitive clause. In this case, A is indicated with absolutive/nominative case-marking and agreement and O is indicated with ergative/accusative case-marking and agreement. (If it were not for the case marking, this would not be apparent.)

This doesn’t sound like a split system. Could you go into more detail about the conditioning factors please?

[Travis B.]

There are two moods, the indicative and the interrogative. The interrogative differs from the indicative in that the ergative/accusative markers are used for the object and the absolutive/nominative, aspect, and tense markers are omitted, the prohibitive is used instead of the negative, and to mark a plural subject the marker -χa is placed immediately after the verb stem.

[Travis B.]

(agglutinative verb morphology with polypersonal agreement, SOV word order, split-ergative alignment, combination of a complex case system and relational nouns, verb-like adjectives, etc.)

…

/a aː ə ɨ/

…

Blue consonants are fronting/palatalizing consonants. Red consonants are backing/rounding consonants.

I like this conlang already!

an ergative/accusative agreement marker, an absolutive/nominative agreement marker, or both

I’d just call these series ‘agentive’ and ‘patientive’ markers respectively.

The split ergativity comes into play when a 1st person or 2nd person core argument is present in a transitive clause. In this case, A is indicated with absolutive/nominative case-marking and agreement and O is indicated with ergative/accusative case-marking and agreement. (If it were not for the case marking, this would not be apparent.)

This doesn’t sound like a split system. Could you go into more detail about the conditioning factors please?

What makes a split system is the nominal case-marking with regard to person. 1st and 2nd person A take absolutive/nominative case-marking and 1st and 2nd person O take ergative/accusative case-marking (and the corresponding agreement). Because this is also a fluid system that uses agreement and case-marking to mark volitionality, 1st and 2nd person volitional S take absolutive/nominative case-marking and 1st and 2nd person non-volitional S take ergative/accusative case-marking (and he corresponding agreement).

[bradrn]

What makes a split system is the nominal case-marking with regard to person. 1st and 2nd person A take absolutive/nominative case-marking and 1st and 2nd person O take ergative/accusative case-marking (and the corresponding agreement). Because this is also a fluid system that uses agreement and case-marking to mark volitionality, 1st and 2nd person volitional S take absolutive/nominative case-marking and 1st and 2nd person non-volitional S take ergative/accusative case-marking (and he corresponding agreement).

How is this a split system? It’s just… well, normal polypersonal agreement in a fluid-S system.

(Also, I don’t see how your names correspond to the actual usage of the cases. There seems to be nothing ‘absolutive’ or ‘ergative’ about any of these cases.)

[Travis B.]

This case-marking/agreement system may make no sense until one considers that the ergative/accusative case is the marked case and the absolutive/nominative case is the unmarked case. So the default pattern is ergative-absolutive, so A gets assigned to the marked case and O gets assigned to the unmarked case. However, in the situations where split-ergativity applies, i.e. when a speech participant is present as a core argument, this is reversed as a nominative-accusative pattern applies, so A gets assigned to the unmarked case and O gets assigned to the marked case, as it is unmarked for a speech participant to be A and marked for a speech participant to be O.

[Travis B.]

Also note that in the case-marking, which I have not unveiled yet, the unmarked case is not marked on the NP (i.e. a bare noun stem is used), whereas the marked case is marked on the NP, indicating that it is more marked overall.

[foxcatdog]

Any naturalistic precedents for the two way contrast between aspirated and ejective? I've heard of aspirated vs tenuis, tenuis vs ejective and aspirated vs voiced but never this.

Also looks cool i like languages with complex inflection.

[bradrn]

This case-marking/agreement system may make no sense until one considers that the ergative/accusative case is the marked case and the absolutive/nominative case is the unmarked case.

Ah, this system. To my knowledge it’s attested (for a given value of ‘attested’) in Proto-Pamir and nowhere else.

So the default pattern is ergative-absolutive, so A gets assigned to the marked case and O gets assigned to the unmarked case. However, in the situations where split-ergativity applies, i.e. when a speech participant is present as a core argument, this is reversed as a nominative-accusative pattern applies, so A gets assigned to the unmarked case and O gets assigned to the marked case, as it is unmarked for a speech participant to be A and marked for a speech participant to be O.

I understand, thanks for clarifying! I believe it was the first sentence I had been missing.

But in this case, I struggle to see how this could be combined with a fluid-S system. Generally, such split intransitive systems rely on a difference in agentivity between the cases. But here there’s no such difference: both cases may be both agentive or patientive as needed.

[Ahzoh]

Any naturalistic precedents for the two way contrast between aspirated and ejective? I've heard of aspirated vs tenuis, tenuis vs ejective and aspirated vs voiced but never this.

A fair number of languages have at the minimum a distinction between ejective and aspirated:
https://en.wikipedia.org/wiki/Tlingit_language
https://en.wikipedia.org/wiki/Lezgian_language
https://en.wikipedia.org/wiki/Abkhaz_language#Phonology
https://en.wikipedia.org/wiki/Chechen_l ... #Phonology
https://en.wikipedia.org/wiki/Ju%C7%80% ... n_language

[foxcatdog]

Any naturalistic precedents for the two way contrast between aspirated and ejective? I've heard of aspirated vs tenuis, tenuis vs ejective and aspirated vs voiced but never this.

A fair number of languages have at the minimum a distinction between ejective and aspirated:
https://en.wikipedia.org/wiki/Tlingit_language
https://en.wikipedia.org/wiki/Lezgian_language
https://en.wikipedia.org/wiki/Abkhaz_language#Phonology
https://en.wikipedia.org/wiki/Chechen_l ... #Phonology
https://en.wikipedia.org/wiki/Ju%C7%80% ... n_language

True but all these languages you listed except for abkhaz (which has voiced stops instead) have tenuis stops according to the wikipedia phonologies you posted (and some have even more).

Through there's nothing wrong with deviating from natlangs for conlanging purposes.

[Darren]

Any naturalistic precedents for the two way contrast between aspirated and ejective? I've heard of aspirated vs tenuis, tenuis vs ejective and aspirated vs voiced but never this.

A fair number of languages have at the minimum a distinction between ejective and aspirated:
https://en.wikipedia.org/wiki/Tlingit_language
https://en.wikipedia.org/wiki/Lezgian_language
https://en.wikipedia.org/wiki/Abkhaz_language#Phonology
https://en.wikipedia.org/wiki/Chechen_l ... #Phonology
https://en.wikipedia.org/wiki/Ju%C7%80% ... n_language

True but all these languages you listed except for abkhaz (which has voiced stops instead) have tenuis stops according to the wikipedia phonologies you posted (and some have even more).

Through there's nothing wrong with deviating from natlangs for conlanging purposes.

Nuxálk

[Ahzoh]

True but all these languages you listed except for abkhaz (which has voiced stops instead) have tenuis stops according to the wikipedia phonologies you posted (and some have even more).

You asked if there was a two-way distinction (contrasting both) between aspirated and ejective, not if such languages only make that distinction. But of course, then there's the aforementioned Nuxalk which also satisfies that part of your question.

[Travis B.]

This case-marking/agreement system may make no sense until one considers that the ergative/accusative case is the marked case and the absolutive/nominative case is the unmarked case.

Ah, this system. To my knowledge it’s attested (for a given value of ‘attested’) in Proto-Pamir and nowhere else.

So the default pattern is ergative-absolutive, so A gets assigned to the marked case and O gets assigned to the unmarked case. However, in the situations where split-ergativity applies, i.e. when a speech participant is present as a core argument, this is reversed as a nominative-accusative pattern applies, so A gets assigned to the unmarked case and O gets assigned to the marked case, as it is unmarked for a speech participant to be A and marked for a speech participant to be O.

I understand, thanks for clarifying! I believe it was the first sentence I had been missing.

But in this case, I struggle to see how this could be combined with a fluid-S system. Generally, such split intransitive systems rely on a difference in agentivity between the cases. But here there’s no such difference: both cases may be both agentive or patientive as needed.

How this works is that there is a threshold for core arguments for agentivity in transitive clauses where the ergative-absolutive system changes to a nominative-accusative system, and the threshold is where a core argument is an SAP. This is still compatible with a fluid-S system (in intransitive clauses of course), because there case/agreement is being used to mark agentivity of S arguments.

[Travis B.]

True but all these languages you listed except for abkhaz (which has voiced stops instead) have tenuis stops according to the wikipedia phonologies you posted (and some have even more).

You asked if there was a two-way distinction (contrasting both) between aspirated and ejective, not if such languages only make that distinction. But of course, then there's the aforementioned Nuxalk which also satisfies that part of your question.

I should note that in this language ejectives are weak rather than strong, so they are not very different from tenuis voiceless plosives, hence why I did not include a tenuis voiceless plosive series.

[Travis B.]

I have decided that the language should have a noun class system, albeit a small one, not one that goes to Bantu-esque levels of excess. There will be five noun classes, which I will mark as classes A, B, C, D, and E. One could call class A "masculine" and class B "feminine" but it does not pattern that strongly with natural gender. Note that core arguments' gender is marked on the verb, but SAP's are not marked for gender.

Note that I have updated the verb conjugation info above to reflect this change.

[Elancholia]

True but all these languages you listed except for abkhaz (which has voiced stops instead) have tenuis stops according to the wikipedia phonologies you posted (and some have even more).

Tzeltal, a Mayan language, contrasts aspirated and ejective stops and affricates. Though /p'/ is only realised as such word-finally, the alveolar, postalveolar, and velar stops/affricates don't appear to have a tenuis form.

https://en.wikipedia.org/wiki/Tzeltal_l ... Consonants

[Travis B.]

Case

The language has the following cases, which take the form of clitics following NP's (or in the case of the adverbial case, nominalized VP's):

agentive	-na
patientive	-jə
dative/benefactive	-rʲɨ
genitive	-ɣa, formerly -wə
adverbial	-χɨa

The following are old cases that have been removed:

instrumental/comitative	-xa
adessive	-tʷʼə
ablative	-sɨ
allative	-βa
inessive	-tʰə
elative	-mʲa
illative	-kʷʰɨ

Locative cases are commonly used in combination with directional markers on verbs, even when they may seem redundant.

[Travis B.]

More Noun Stuff

Nouns are marked for number.

Sg.	Pl.
-	-mə

Nouns are unmarked for gender, which is solely marked on verbs and verb-like adjectives.