Zompist Bboard Again

Ares Land wrote: ↑Mon Apr 19, 2021 7:25 am

More: show

Leaving tone aside for the moment, that would be klɵ so '(I see) they are sleeping' then?

Very good! You got it.

(Further exercise: now figure out where the tones come from. It’s not too difficult, if you remember that dɵ has low tone and =ė is a toneless clitic.)

I admit you'll have to explain that! I'm not sure I really got the parts about tones (I'm not familiar with autosegmental analysis; I'm going to brush up on it a bit)

Ares Land wrote: ↑Tue Apr 20, 2021 4:00 am I admit you'll have to explain that! I'm not sure I really got the parts about tones (I'm not familiar with autosegmental analysis; I'm going to brush up on it a bit)

It’s surprisingly simple. The underlying form of this sentence is, as I presume you figured out already, //t⟨dɵ̀⟩é=e e̤b//. Morphophonological processes change this to /klɵ̌=e e̤b/. Then apply Contour Relinking (nothing else applies) to give /klɵ̀=é e̤b/ Klɵ é ehb.



_______________
I’m not at the stage where I can write another post yet, but here’s some of my thoughts in the meantime:

• Classifiers: Obligatorily used with numerals and demonstratives. Noun classifiers can be used with bare nouns and anaphorically as a discourse strategy (to compensate for the lack of third person pronouns). I’m very undecided as to whether classifiers should be one or two classes, open or closed, or should use repeaters.
• Verbal complex: I want this to allow very complex auxiliary constructions à la English, though ideally with very non-English-like semantics. It probably won’t mark mood, there being obligatory evidentials anyway, but I’d like to do some sort of interesting fused tense-aspect paradigm. (I’m thinking of making the unmarked TA something like stative/habitual — or is that too much like English? Does it matter?) It should have directionals somewhere in it as well.
• Syntax: At some point I’ll need to figure out coordination, subordination etc. Perhaps I can do something interesting with alignment or classifiers here? (Apinajé has ergative marking only in subordinate clauses; I was reading about some language, possibly more than one, which uses classifiers to form relative clauses. Jakaltek?) Given the Papuan influence on Hlʉ̂, it would be interesting to experiment with some sort of clause chaining or switch reference.

Ah, thanks! got it now.

Hlʉ̂ ASCII orthography

I tried to set up a dictionary for Hlʉ̂, only for SIL Toolbox to choke on the non-ASCII filename. I thus designate the following an Official Hlʉ̂ Orthography in situations when non-ASCII characters are unavailable (or inconvenient):

[m n] ⟨m n⟩
[p t~k ʔ] ⟨p t~k q⟩
[b d~ɡ~l~l̥ s~h] ⟨b d~g~l~l s~h⟩

/i ʉ u/ ⟨i ue u⟩
/e ɵ o/ ⟨e oe o⟩
/a/ ⟨a⟩

/a á ǎ â a̤ a̰/ ⟨a aj av ax ah aq⟩

(Sample: Mahd hluej donoj hup bax nov hemj e noe. Yes, I hate it too.)

Classification in Hlʉ̂ (and some discourse)

Hlʉ̂ utilises two separate systems of noun classification. Noun classifiers are optionally used to modify, well, nouns; they play an important part in Hlʉ̂ discourse. Numeral classifiers obligatorily accompany numerals when modifying nouns.

Syntactically, both constructions are pretty straightforward. Noun classifiers, when present, appear immediately after the noun. They form a closed set of around 15–20 words, many of which (though not all) are transparently derived from generic nouns. Examples include:

hlʉ́ ‘CL.human’ ← hlʉ̂ ‘human’
blep ‘CL.domestic.animal’ ← ?
siʔ ‘CL.fire’ ← siʔ ‘fire’
hup ‘CL.stone’ ← hup ‘stone’
mat ‘CL.plant’ ← mǎt ‘bush’

Once I know a bit more about Hlʉ̂ culture, I plan to do a separate post on the semantics of Hlʉ̂ classifiers sometime, at which point I will give a full listing of noun classifiers.

Syntactically, noun classifiers appear immediately after the noun and any demonstrative present. Both noun and classifier are optional, though classifiers are obligatory with demonstratives:

Mahd hlʉ́ hém a ehb. ʔlih blep hlʉ́ da ni e ehb. ʔlih plɵʔe a nɵ.
//ma̤d hlʉ́ hém=a e̤b | ʔli̤ blèp hlʉ́ dà nì e e̤b | ʔli̤ plɵ̀ʔe=a nɵ̀//
man CL.human see=1/2 VIS. dog CL.domestic.animal CL.human to COP=3 VIS. dog like=1/2 DECL.

lit. I see a human_CL man_N. The human_CL has an animal_CL dog_N. I like dogs_N.

In contrast to noun classifiers, numeral classifiers in Hlʉ̂ constitute a large and semi-open class. While noun classifiers mostly classify nouns in terms of general categories, numeral classifiers place a larger emphasis on shape and size, though the two categories do overlap somewhat:

hlʉ̂ ‘CL.human’ ← hlʉ̂ ‘human’
mɵʔ ‘CL.thin.flat’ ← mɵʔ ‘leaf’
sen ‘CL.long.thin’ ← ?
hup ‘CL.small.round’ ← hup ‘stone’
nǒ ‘CL.bump’ ← ?

When a noun is modified by a numeral, an appropriate numeral classifier — if there is one — must be placed immediately after the numeral. However, not all nouns have a numeral classifier. When such nouns are modified by a numeral, they use a so-called ‘repeater’ construction, in which a copy of the head noun is placed after the numeral: naʔ dî naʔ lit. ‘day one day’.

Thus, summing up, the noun phrase thus has the following maximal structure:

[possessor/relative LNK] [noun [demonstrative classifier]] adjective [numeral classifier]

It’s worth noting that the two classifier slots aren’t exclusive; e.g. donó hup dí nǒ ‘hill CL.stone one CL.bump’ is perfectly valid.

Perhaps more interesting than the syntax of classifiers is their semantics and pragmatics. Noun classifiers in particular play an important role in Hlʉ̂ discourse. Generally speaking, the presence of a noun classifier indicates specificity. As a first attempt at a definition, a noun is specific when it refers to one particular entity (or group of entities); a non-specific noun is then generic. Note that this is not the same as definiteness! Consider:

Nɵd hlʉ́ noʔ mla a nɵ ‘I talk to a woman’
Nɵd hlʉ̂ ni ‘A woman is a person’

In both sentences, the underlined noun is indefinite; however it occurs with a noun classifier in the first sentence but without one in the second. In the first sentence, the speaker refers to a specific woman, even if the woman is not explicitly identified yet. However, in the second sentence, the speaker is talking about women in general. We might say that a definite noun is one for which the listener can identify the referent, whereas a specific noun is one which for which the speaker can identify the referent. (Full disclosure: I know I’m not the one who came up with that definition, but I’m not sure where I saw it.)

The other parameter of variation, of course, is the presence or absence of the head noun. Generally, the head noun is included when introducing a new referent. However, a noun classifier by itself has anaphoric function: it refers to an old referent rather than introducing a new one. (Of course, in most cases the most idiomatic way to do this is simply by eliding the NP entirely.) This effectively ends up compensating for the lack of a third person pronoun, since any such pronoun can be replaced by a classifier, possibly with a demonstrative.

Thus we can look at the previous example again:

Mahd hlʉ́ hém a ehb. ʔlih blep hlʉ́ da ni e ehb. ʔlih plɵʔe a nɵ.
//ma̤d hlʉ́ hém=a e̤b | ʔli̤ blèp hlʉ́ dà nì e e̤b | ʔli̤ plɵ̀ʔe=a nɵ̀//
man CL.human see=1/2 VIS. dog CL.domestic.animal CL.human to COP=3 VIS. dog like=1/2 DECL.

lit. I see a human_CL man_N. The human_CL has an animal_CL dog_N. I like dogs_N.

The referent man is introduced with a full noun phrase, as it is a new referent. A classifier is included as this is a specific NP: I’m looking at some particular person. In the next clause, the same referent is now ‘old news’, so is referred to with a classifier only, acting in an anaphoric function. But we also introduce the person’s dog, which is specific and a new referent, and thus gets both a classifier and a noun. Finally, I make a statement about dogs generally; the non-specific NP is referred to as ʔlih without a classifier. If I were to now say something like Mahd hlʉ́ sí a nɵ, this could only mean ‘A man [not the same one as before] goes’; it could not mean *‘The [same] man goes’, since I include the head noun thus implying this is a new referent.

(An interesting observation: note that the Hlʉ̂ verb plɵʔe cross-references a singular object. However, in English, the object dogs is plural. What’s happening here is that English has no way to refer to a non-specific noun: both articles a and the imply specificity, or at least are not incompatible with it. I like a dog says that I like a specific dog, and I just haven’t specified which. Accordingly, it would appear that the only way to express a non-specific object in English is to pluralise it. But this means I have no way to pluralise a generic object, except by periphrastic means: I like many dogs at a time, say, or I like groups of dogs. By contrast, the distinction between singular specific ʔlih blep plɵʔe a nɵ, singular generic ʔlih plɵʔe a nɵ, and plural generic ʔlih plɵlɵʔe a nɵ is straightforward in Hlʉ̂.)

By the way, exactly the same applies to numeral classifiers. The presence of a numeral classifier — and hence a numeral — implies specificity. (You could admittedly have non-specific numerals such as ‘I like bunches of two bananas’, but as in English, Hlʉ̂ expresses these with periphrasis.) Similarly, a numeral+classifier can be used anaphorically without a head noun:

Mud bâ hup dʉ é di? / Dí hup ní e nɵ.
//mùd bâ hùp n⟨lɵ̀⟩í=e dì ||| dí hùp ní=e nɵ/
banana two CL.small.round ⟨PL⟩is=3 Q? / one CL.small.round is=3 DECL

Do you have two bananas? / I have one.

EDIT: Corrected word order in last example

If I can use a term WeepingElf uses…ROCK’N’ROLL! I first really started reading this at the doctor’s the other day and I’m very impressed by how much cool stuff phonologically is going on and by how much character this language has despite its brief descriptions here.

Man in Space wrote: ↑Sun Jun 13, 2021 7:43 pm If I can use a term WeepingElf uses…ROCK’N’ROLL! I first really started reading this at the doctor’s the other day and I’m very impressed by how much cool stuff phonologically is going on and by how much character this language has despite its brief descriptions here.

Wow, thanks for the praise! Glad you like it. I did deliberately try to give this one more ‘character’ than usual; it helps that I had a good idea of where I wanted to go from the beginning.

And that reminds me: I really need to work more on Hlʉ̂. I should probably figure out the TAM system soon, though that’ll require a bit of research on auxiliary verb systems first. Perhaps when I get more time.

I would like to join in the congrats, and I definitely would like to see more.

Qwynegold wrote: ↑Sat May 15, 2021 10:33 am I have made a conlang design challenge …

OK, so I might have gotten just a little bit carried away with this…

For reference, here were my choices:

bradrn wrote: ↑Sat May 15, 2021 10:15 pm

• A2 Fixed stress location of a type that you seldom or never use in conlangs. When affixes are added to a word, the stress moves so that it will stay on the specificly numbered syllable.
• A4 Consonant mutation
• B6 Uses prefixes way more often than suffixes
• C3 A retroflex series, and not just on fricatives/affricates

Phonology

Consonants and vowels

The surface consonant inventory of Wēchizaŋkəŋ /ˈweːɰiˌzaŋkəŋ/ appears fairly straightforward:

/m n ɲ ŋ/ ⟨m n ɲ ŋ⟩
/p t ʈ c k/ ⟨p t tr c k⟩
/ᵐp ⁿt ᶯʈ ᶮc ᵑk/ ⟨mp nt ntr nc nk⟩
/s ʂ ɕ/ ⟨s sr sh⟩
/β z ɹ ɰ/ ⟨ꞵ z th ch⟩
/r j w/ ⟨r y w⟩

However there are a number of oddities here:

• The retroflex series has a very restricted distribution, appearing only in the onset. Additionally, all stop+/r/ clusters occur except for /tr/. Thus /ʈ ᶯʈ ʂ/ appear underlyingly to be |tr ntr sr|.
• Similarly, the palatal series /ɲ c ᶮc ɕ/ seems to be in complementary distribution with the alveolars /n t ⁿt s/ and velars /ŋ k ᵑk/: the former occur only after underlying /i j/, while the latter two appear elsewhere. Thus the palatals are underlyingly either alveolars or velars. (Of course, sometimes the underlying consonant is impossible to determine, in which case I transcribe them as simply palatals.)
• The evidence for considering the prenasalised stops to be unit phonemes is very weak, and extends mostly to the facts that (a) they include /ɳ/ which is elsewhere non-phonemic, and (b) they pattern as units with respect to consonant mutation and syllable structure. Neither of these is particularly convincing, especially in light of the fact that these are usually unvoiced [mp nt ɳʈ ɲc ŋk]; thus these too are analysed as underlying clusters.
• The voiced approximants/fricatives /β z ɹ ɰ/ most commonly arise from intervocalic lenition. But these phonemes also occur in other environments and thus must be underlying in some cases.

Accordingly, the underlying i.e. morphophonemic consonant inventory must be:

|m n ŋ|
|p t s k|
|β ɹ z ɰ|
|r j w|

In contrast to the consonants, vowels are much simpler, and have fewer distributional oddities:

/a e i o u ə ɨ/ ⟨a e i o u ə ɨ⟩
/aː eː iː oː uː/ ⟨ā ē ī ō ū⟩

Phonotactics and stress

Maximal underlying syllable structure is ((C₁)C₂(r))V((C₃)C₄). All consonants but /ŋ/ are allowed in the onset; the coda can be a nasal, a stop or a prenasalised stop. In onset clusters, C₂ can be a stop or /s ɰ/; if it is a stop, C₁ may be a homorganic nasal. Syllabification is predictable via the Maximal Onset Principle. Roots tend to be underlyingly disyllabic, though are often monosyllabic at the surface.

Surface primary stress is consistently on one of the first two syllables. In most case, the stressed syllable can be predicted by the following rules, although they occasionally fail:

1. If one of the first two syllables is heavy (has a long vowel), stress that one.
2. If both are heavy, stress the initial syllable.
3. If both are light, and only one has rime /a/ or /i/, stress that one.

Stress is however fully predictable given the morphophonemic representation.

Phonological rules

I divide phonological rules into segmental and post-prosodic rules. The segmental rules apply first, followed by foot assignment and then the post-prosodic rules.

Segmental rules:

1. Geminate Deletion: C₁ C₁ → C₁
2. Retroflexion: tr sr → ʈ ʂ
3. Nasal Assimilation: nasals followed by a stop assimilate in PoA
4. Consonant–Cluster Deletion: C₁ % C₂ C₃ → C₂ C₃ (where % is a syllable boundary)
5. Palatalisation: n ŋ t k s ʂ → ɲ ɲ c c ɕ ɕ / {i,j} _ (applying to all members of a consonant cluster)
6. Intervocalic Lenition: p t s k → β ɹ z ɰ / V_V

Following the segmental rules is syllabification and foot assignment. Feet are iambic and assigned left-to-right; primary stress is given to the leftmost foot, and degenerate feet are not tolerated. Heavy syllables — i.e. those with a long vowel — cannot occur in the weak part of the foot, and are assigned a separate foot if this would occur. Thus e.g. |əɰəwəzəekrəaːɰampəicəβuniːɨŋ| is footed as (əˈɰə)(wəˌzə)(eˌkrə)(aː)(ɰaˌmpə)(iˌcə)(βuˌniː)ɨŋ.

Post-prosodic rules can modify this prosodic structure, and are as follows:

1. Metathesis: If a foot has structure (CV.V(C)), the first syllable undergoes metathesis becoming (VC.V(C))
2. Unstressed Vowel Deletion: If a foot begins with an unstressed vowel, that vowel is deleted, with consequent resyllabification of any following consonant. If the previous foot ends in a vowel, that vowel may mutate: ə+V ɨ+V a+e a+i i+a a+o a+u e+u u+a u+e u+i i+u → V V e e e o o o o o ɨ ɨ, otherwise there is no mutation.
3. Cluster Resolution: If the previous rule formed an illegal cluster, a copy of the following vowel is inserted.
4. Stress Clash Resolution: If two stressed syllables are adjacent, the second is destressed (becoming a degenerate foot in the process)
5. Stressed Vowel Lengthening: /a e i o u/ → /aː eː iː oː uː/ in stressed syllables. /ə ɨ/ do not lengthen, but mutate to /a i/.

Note that the action of Palatalisation leaves some underlying consonants unknowable, and Unstressed Vowel Deletion leaves some underlying vowels unknowable. In these cases the relevant phonemes will be written |ɲ|, |c|, |sh| or |V| as necessary.

Consonant mutation

Consonant mutation is a key part of morphophonology in this language. Mutation grades are lexically specified for particular affixes, and act to alter the phoneme immediately following the affix. There are three mutation grades, namely:
Unlisted consonants do not mutate; ‘0’ indicates deletion, or lack of initial consonant. In morphophonemic representation I write these grades as diacritic features at the right edge of the morpheme: |ⁿ| for nasal mutation, |ʷ| for weak mutation and |ˡ| for lenition. e.g. third person nonplural patientive is |əʷ-|, and patientive plural is |ꞵiⁿ-|.

Lenition is unproblematic: it simply corresponds to triggering of rule Intervocalic Lenition without a preceding vowel. (Indeed this is surely its historical origin in the only place where it is found, the first person singular possessive clitic.) However the ordering of the other mutations with respect to other rules is more interesting. On the one hand, the weak mutation treats palatals as a separate series to alveolars and velars, ignoring the triggering vowels if they are deleted: |wə-ikɨŋ| /ˈwiːcɨŋ/ ‘I run’, |rəʷ-ikɨŋ| /ˈriːɕɨŋ/ ‘It runs’ (not */ˈriːkɨŋ/ or */ˈriːɰɨŋ/). On the other hand, the nasal mutation can cause vowels to surface when otherwise they would be deleted: |rɨ-e-əmeŋ| /ˈreːmeŋ/ ‘You (dual) die’, |rɨ-ꞵiⁿ-əmeŋ| /rɨˈβiːnəˌmeŋ/ ‘You (plural) die’ (not */rɨˈβiːnmeŋ/). And mutating affixes cannot trigger lenition: |kə-e-in-təɰa| /ˈkeːɲcəɰa/ ‘I finish seeing you’, |kə-e-ɹa-təɰa| /ˈkeːɹaˌɹaɰaː/ ‘I start seeing you’, but |kə-e-səʷ-təɰa| /ˈkeːzəˌsaɰa/ ‘I keep on seeing you’ (not */ˈkeːzəˌɹaɰa/). I thus propose that the different mutations are applied in different places in the phonology: nasal mutation is applied just before syllabification, whereas weak mutation is applied at some point after Unstressed Vowel Deletion, while both mutations are considered non-vocalic unrealised segments for the purpose of segmental rule Intervocalic Lenition.

Basic syntax and morphology

Word classes

As with most languages, there is a basic distinction in the open word classes between nominals and verbals. The vast majority of words are one or the other; very few can act as both.

Nominals are words which can head an NP and act as arguments in a clause. This comprises three word classes: personal pronouns, (common) nouns and proper nouns. Common and proper nouns both comprise open classes; proper nouns are distinguished mostly by the fact that they are unpossessable. Pronouns make up a closed class of three members: first person exclusive |wā|, first person inclusive |wət| /wat/, and second person singular |tā|. Generally these are used only for emphasis, verbal cross-referencing prefixes being sufficient in most cases. Interestingly they distinguish clusivity, which is unmarked on the verb; on the other hand, the verbal prefixes distinguish three numbers absent in the pronouns.

Verbals are words which can take verbal morphology, most typically cross-referencing prefixes. This comprises four word classes: intransitive S=A, intransitive S=O, transitive and ditransitive verbs. These are distinguished primarily by how many and which cross-referencing prefixes are required by the underived root. There are few ditransitive verbs, most prominently |yomp| ‘give’; these require three cross-referencing prefixes. Transitive verbs include |punī| ‘see’, |othem| ‘eat’ and |awət| ‘wash’; these require two prefixes. In some languages the copula is a separate word class, but here the copula |tham| is simply a regular transitive verb. Intransitive verbs are split depending on which series of prefixes they take: intransitive S=A verbs like |waŋ| ‘run’ and |ikɨŋ| ‘dance’ take the agentive series, whereas intransitive S=O verbs like |əmeŋ| ‘die’ and |kaꞵa| ‘fall’ take the patientive series. Thus this language can be categorised as having split-S alignment. It appears that S=O is the less marked class; there are far fewer S=A intransitives than S=O, and the former may even be a closed class.

Aside from nominals and verbals there are a few other minor classes, all of which are closed. The six demonstratives distinguish three degrees of distance:

	Nominal	Adverbial
Proximal	|nā|	|nīm|
Medial	|wē|	|wēm|
Distal	|mpē|	|mpī|

True adjectives make up a small closed class of only six members: |itē| ‘small’, |kapa| ‘big’, |yəthu| ‘good’, |wəŋkom| ‘bad’, |anɨm| ‘new/young’ and |chaŋkri| ‘old’. When used attributively they behave as S=O intransitive verbs; however they differ in that they can modify a noun directly, rather than requiring a relative clause. Finally, postpositions make up another small closed class, consisting of |nəm| ‘dative/ablative’, |trəŋ| ‘instrumental’, |want| ‘comitative’, |chənu| ‘locative’, |mpun| ‘inside’, |yeŋ| ‘outside’. These words are less commonly used than the English equivalents, applicative constructions being preferred when possible.

Cross-cutting the above word classes are the interrogative words: |ꞵəyə| ‘who’, |ꞵiɲ| ‘what/which’, |ꞵənu| ‘where’, |ꞵoru| ‘when’, |ꞵethe| ‘how’, |ꞵachi| ‘how many/much’. The first four are syntactically nouns; the other two act as verbs. There is no separate word for ‘why’, which is expressed as a combination of |ꞵiɲ| plus a benefactive.

Basic syntax

The typical sentence is verb-final and SOV, though basic word order varies extensively depending on semantic and pragmatic considerations, and arguments are consistently elided when recoverable. NPs however have the following fairly rigid order:

numeral — possessor/deictic — noun — adjective — relative clause

(Note that relative clauses are actually internally headed, which in most cases is formally identical to being head-initial.)

Most non-declarative clause types such as polar questions and imperatives are simply marked on the verb without requiring a specific construction. (Content questions are the exception here; see below for details.) Copular and existential clauses similarly have no special properties and simply use the copular verb |tham|, though existentials are unusual in taking a dummy subject, and thus being ambiguous with a copular interpretation:

Parē yēzəkrəthāmɨŋ?
/paˈreː ˈjeːzəkrəˌɹaːmɨŋ/
|pare əʷ-rəʷ-e-sə-krə-tham-ɨŋ|
dog 3s.O-3s.A-NP-sit-IRR-is-Q

Is there a dog sitting? / Is it a sitting dog?

Comparatives also have no special syntax, and consist of a biclausal construction contrasting the two participants. Usually the subject clause takes an intensifier:

Narē ērashē, Rāmpu ēchaꞵā.
/naˈreː ˈeːraˌɕeː | ˈraːmpu ˈeːɰaˌβaː/
|Narē əʷ-e-riaʷ-itē, Rāmpu əʷ-e-kapa|
Nare 3s.O-NP-INT-small Rāmpu 3s.O-NP-big

Narē is smaller than Rāmpu (lit. Narē very big, Rāmpu small)

However, predicative possession uses a special copular construction, in which the possessor is cross-referenced as an object. The possessum is usually incorporated into the verb, though more bulky possessums need to be given as subjects. If no explicit subject is given, as with existentials the construction takes a dummy subject:

Rēꞵarethāmɨŋ.
/ˈreːβareˌɹaːmɨŋ/
|ə-rəʷ-e-pare-tham-ɨŋ|
1s.O-3s.A-NP-dog-is-Q

I have a dog. (lit. ‘there dog-is to me’)

Nominal morphology

Nominal morphology is minimal. Nouns inflect only for possessor, via the following clitics:

	Singular	Non-singular
1	ˡ=	tɨ=
2	kə=	rɨ=
3	ʷ=	ʷ=

These are transparently derived from the verbal object cross-referencing prefixes. Note that a third person possessor is signalled solely by weak mutation of a word-initial consonant, and first person singular is similarly signalled by lenition: parē ‘dog’ vs ꞵarē ‘my dog’ vs arē ‘their dog’. (Their analysis as clitics is therefore arguable.) The optional possessor is placed before the modified word: melē arē ‘the boy’s/boys’ dog’.

Verbal morphology: the word

In contrast to the minimal nominal morphology, verbal morphology is extremely complex and polysynthetic:

Tētharēchriɲəŋazishanaŋkrazēwathiŋ?
/ˈteːɹaˌreːɰriɲəŋˌaziˌɕanaŋkraˌzeːwaˌɹiŋ/
|ti-e-tha-riaʷ-krə-iɲ-ŋəzi-sə-nə-am-ŋkrase-wat-ɨŋ|
2ns.O-NP-AND-INT-IRR-TERM-PERF-ABL-MID-BEN-food-make-Q

Were you two able to have really gone and finished making each other some food?

The verbal word is formed from a verb stem (itself a composite entity) using a slot-based template:

-10: object
-9: indirect object
-8: agent
-7: direction/posture
-6: object plurality
-5: adverbial
-4: irrealis
-3: aspect 1
-2: negation
-1: aspect 2
0: stem
+1: illocution

Person marking

The finite verb begins with personal prefixes. There are two sets of prefixes:

Person/number	Agent (slot -8)	Object (slots -9,-10)
1 singular	wə-	ə-
2 singular	tə-	kə-
3 singular	rəʷ-	əʷ-
1 non-singular	wi-	tɨ-
2 non-singular	ti-	rɨ-
3 non-singular	iʷ-	ɨʷ-

As this is a split-S language, intransitive verbs take a single prefix from a lexically determined series, whereas transitive verbs take an object prefix followed by an agent prefix. In the case of ditransitive verbs, the indirect object is specified by a second object prefix placed immediately after the direct object prefix. The object series appears to be less marked, insofar as loans and derived intransitive verbs have a strong tendency to take object prefixes.

If an object prefix is used, an obligatory number marker is placed in slot -6: ꞵiⁿ- when the object is plural, and e- when the object is non-plural. Note that objects thus distinguish a separate dual number, whereas agents do not. When both a direct and indirect object are specified, ꞵiⁿ- is used when either object is plural, creating some ambiguity.

Direction/posture

A direction or posture prefix is placed in slot -7. There are four such prefixes: tha- ‘andative/going’, wəŋ- ‘venitive/coming’, sə- ‘sitting/lying’, əŋə- ‘standing’. The locatives and posturals relate to the agent argument for transitive verbs, and the only argument for intransitive verbs.

Adverbial

Slot -5 contains one of a number of miscellaneous prefixes with a generally adverbial function:

• riaʷ- ‘intensifier’ — a highly productive general-purpose intensifier; can be used in any sentence involving large quantities or strong emotions
• əⁿ- ‘diminutive’ — used with actions carried out only ‘a little bit’ or ‘incompletely’
• yat- ‘today’, ic- ‘at night’ — specify time period of action
• ꞵaza- ‘quickly’ — modifies speed of action
• yu- ‘but’ — contrasts this verb to the previous clause
• ꞵo- ‘also’ — gives additional information
• naŋ- ‘just; first time’ — specifies that an action has soon been or will just be started, or is being done for the first time
• chaꞵə- ‘still; again’ — specifies that an action is continuing from before the reference point, or is being repeated

Mood, polarity, illocution

Negative polarity is expressed by placing the prefix mpə- in slot -2:

Wīmantacha. / Wīŋkrəmpamantacha.
/ˈwiːmaˌntaɰa/ /ˈwiːŋkrəˌmpamaˌntaɰa/
|ɨʷ-ꞵiⁿ-man-təcha| / |ɨʷ-ꞵiⁿ-krə-mpə-man-təcha|
3ns.O-PL-PFV-fall / 3ns.O-PL-IRR-NEG-PFV-fall

They fell / They did not fall

Slot +1 contains the three illocutionary suffixes:

• -ɨŋ ‘question’ (polar, alternative or content)
• -tən ‘imperative’
• -əmp ‘epistemic uncertainty’

All four affixes mentioned above obligatorily co-occur with the slot -4 irrealis prefix krə-. This prefix can also occur in other situations: specifically, it marks future and counterfactual events.

Ēkrəzənūthənā…
/ˈeːkrəzəˌnuːɹəˌnaː…/
|əʷ-ə-e-krə-sə-nu-təna …|
3s.O-1s.O-NP-IRR-ABL-LOC-go …

If I could reach it, then…

Aspect

The category of aspect is marked over two slots. What I will arbitrarily label ‘aspect 1’ is marked in slot -3 using the following prefixes:

• ∅- ‘imperfective’ — unmarked aspect, denotes anything not covered by the prefixes below (stative, ongoing activities, etc.)
• maⁿ- ‘perfective’ — emphasises a holistic view of the action; usually past or future
• āka- ‘iterative’ — denotes a repeated action
• tha- ‘inchoative’ — emphasises the start of an action or state
• iɲ- ‘terminative’ — emphasises the end of an action or state

‘Aspect 2’ is marked in slot -1 using the following prefixes:

• ŋəzi- ‘perfect’ — denotes an action performed before the time of reference, emphasising present relevance
• weʷ- ‘habitual’ — denotes an action or state regularly or consistently performed or held; excludes generics, which are unmarked
• naⁿ- ‘distributive’ — denotes an action performed over a large area of space, usually repeatedly

Verbal morphology: the stem

Nested within the affixes outlined above is the verb stem. This unit consists of the verbal root itself, any incorporated nouns and various ‘inner affixes’. The verb stem may be as simple as just one verbal root, or as complex as the underlined portion of the following example:

Tētharēchriɲəŋazishanaŋkrazēwathiŋ?
|ti-e-tha-riaʷ-krə-iɲ-ŋəzi-sə-nə-am-ŋkrase-wat-ɨŋ|
2ns.O-NP-AND-INT-IRR-TERM-PERF-ABL-MID-BEN-food-make-Q

The ‘inner affixes’ of the verb stem differ from the ‘outer affixes’ outlined above in being scope-ordered: rather than being arranged in slot-and-template morphology, each affix is applied to a smaller verb stem, modifying it to make a larger stem. In this way they are much more explicitly ‘derivational’ than the outer affixes. They also have much more derivational semantics: for instance all valency-changing prefixes are part of the verb stem.

Mediopassive

The prefix nə-, which I will term the ‘mediopassive’, is the sole valency-decreasing affix in Wēchizaŋkəŋ. In prototypical usage it acts as a reflexive or reciprocal, setting the agent equal to the object:

Ēnawat.
/ˈeːnaˌwat/
|ə-e-nə-awət|
1s.O-NP-MED-wash

I am washing myself.

Tēnəꞵūnī.
/ˈteːnəˌβuːniː/
|tɨ-e-nə-punī|
1ns.O-NP-MED-see

We two see each other.

In some cases it may delete the agent entirely, acting as a passive:

Ēraꞵaŋazinakrampī!
/ˈeːraβaŋˌaziˌnakraˌmpiː/
|ə-e-riaʷ-maⁿ-ŋəzi-nə-krampī|
1s.O-NP-INT-PFV-PF-MED-hit

I’ve been hit!

A handful of ‘deponent verbs’ (using Kemmer’s terminology (1993)) occur only in the mediopassive, e.g. |nə-taneŋ| ‘vanish’, |nə-ŋkāwa| ‘wake’:

Ēmanathanēŋ.
/ˈeːmaˌnaɹaˌneːŋ/
|əʷ-e-maⁿ-[nə-taneŋ]|
3s.O-NP-PFV-[MED-vanish]

It vanished.

*Wēmantanēŋ.
*|əʷ-wə-e-maⁿ-taneŋ|
*3s.O-1s.A-NP-PFV-vanish

(Intended: I vanished it)

Wēmanicanəthāneŋ.
/ˈweːmaniˌcanəˌɹaːneŋ/
|əʷ-wə-e-maⁿ-icə-[nə-taneŋ]|
3s.O-1s.ANP-PFV-CAUS-[MED-vanish]

I made it vanish.

Causatives and applicatives

Wēchizaŋkəŋ has five valency-increasing prefixes: one causative and four applicatives.

The causative icə- introduces a semantic role of Causer, in the syntactic role of agent. The former agent is demoted to indirect object, displacing the former indirect object if any. An example of the causative was supplied above: applying the causative to ēmanathanēŋ ‘it vanished’ gives wēmanicanəthāneŋ ‘I made it vanish’.

The four applicatives each introduce or promote a former indirect object to direct object position; the former object is demoted to indirect object. The most frequent applicative is the ‘benefactive’ am-. (Really a more general dative applicative, but ‘benefactive’ seems to be the traditional name.) As the name suggests, prototypically this introduces a beneficiary:

Chaꞵūŋ ayekrāmawathən
/ɰaˈβuːŋ ˈajeˌkraːmaˌwaɹəŋ/
|chaꞵuŋ əʷ-əʷ-rəʷ-e-krə-am-awət-tən|
fruit 1s.O-3s.O-3s.A-NP-IRR-BEN-wash-IMP

Wash the fruit for me.

It also is commonly used to introduce a stimulus, or less commonly a manner (usually given as an action nominal):

Nā kaꞵāchəŋ wīreāŋkrem.
/naː kaˈβaːɰəŋ ˈwiːreˌaːŋkrəm/
|nā kapa-kəŋ əʷ-wi-e-riaʷ-am-krem|
this big-OBJ 3s.O-1ns.A-NP-INT-BEN-worry

We are all very worried about this big one.

Chaꞵūŋ ēzaramic ayekrāmawathən
/ɰaˈβuːŋ ˈeːzaˌramic ˈajeˌkraːmaˌwaɹəŋ/
|chaꞵuŋ əʷ-e-sarəm-ic əʷ-əʷ-rəʷ-e-krə-am-awət-tən|
fruit 3s.O-NP-rub-ACT 3s.O-3s.O-3s.A-NP-IRR-BEN-wash-IMP

Wash the fruit by rubbing it.

Occasionally it can even be used for a maleficiary:

Ꞵarē aꞵanamameŋ.
/βaˈreː ˈaβanaˌmameŋ/
|ˡ=parē ə-əʷ-e-maⁿ-am-əmeŋ|
|1s=dog 1s.O-3s.O-NP-PFV-BEN-die|

My dog died on me.

The other applicatives are rarer, and have a correspondingly smaller range of meaning:

• trə- ‘instrumental’ — introduces an instrument
• yo- ‘comitative’ — introduces an accompaniment to the agent
• nu- ‘locative’ — introduces a location, source, destination etc., as well as times

Other affixes

In addition to the valency-changing prefixes, a handful of other miscellaneous prefixes can appear in the verb stem.

The abilitative sə- denotes that the agent is able to do the specified action:

Thēchepkɨ sēkrəzəꞵūnīŋ?
/ˈɹeːɰepkɨ ˈseːkrəzəˌβuːniːŋ/
|˻əʷ:tha:kiep:kɨ˼ əʷ-tə-e-krə-sə-punī-ɨŋ|
˻3s.O:INCH:fenestrated:AGT˼ 3s.O-2s.A-NP-IRR-ABL-see-Q

Can you see the delicious monster?

The direct evidential taꞵə- denotes that the action was directly seen or heard by the speaker:

Nīm īyesāꞵəwāt.
/niːm ˈiːjeˌsaːβəˌwaːt/
|nīm əʷ-iʷ-e-riaʷ-taꞵə-wat|
here 3s.O-3ns.A-NP-INT-DIR-do

I saw them doing it right here.

The reduplicative continuative C(C)(C)ə- denotes a long and continuing action; or, more formally, it de-emphasises the event boundaries:

Rishi wiŋazinūchəchātrə…
/riˈɕi wiŋˈaziˌnuːɰəˌɰaːʈə/
|rishɨ wi-ŋəzi-nu-kə~katrə|
sand 1ns.A-PF-LOC-CONT~jump

We had been jumping and jumping on the sand [when]…

Noun incorporation

Incorporation of O and S_O argument nouns into the verb stem is a highly productive process. The incorporated noun is placed immediately before the smaller verb stem to which it is applied: usually this is immediately before the root, but placement elsewhere is also quite common.

In Mithun’s typology (link, link), Wēchizaŋkəŋ displays noun incorporation of Types I, II and III:

• Type I noun incorporation is the creation of a compound to narrow the scope of a verb:
  
  Wəchāꞵuntēche.
  /wəˈɰaːβuˌnteːɰe/
  |wə-chaꞵuŋ-teche|
  1s.A-fruit-cut
  
  I was fruit-picking.
• Type II noun incorporation is the incorporation of an object — often a body part — with simultaneous promotion of an oblique argument to direct object:
  
  Kəwamaŋkōꞵeŋkrāmpī.
  /kəˈwamaˌŋkoːβeˌŋkraːmpiː/
  |kə-wə-maⁿ-koꞵen-krampī|
  2s.O-1s.A-PFV-head-hit
  
  I hit you in the head.
• Type III noun incorporation is the incorporation of an object for discourse backgrounding:
  
  Waŋiɲrīshiꞵūnī
  /ˈwaŋiɲˌriːɕiˌβuːniː/
  |iʷ-wəŋ-iɲ-rishi-punī|
  3ns.A-VEN-INCH-sand-see
  
  They began to see the sand. (having already mentioned the sand)

Note that Types I and III are valency-decreasing, whereas Type II does not alter valency.

All examples of incorporation given above involve straightforward compounding of the noun and the verb. A less familiar pattern is what I might term ‘postpositional incorporation’. In this construction, the incorporated noun is followed by an applicative: the applicative adds an extra argument to the verb, only for its referent to be immediately incorporated:

Nantrāmənūthənāit waꞵewēyem.
/naˈɳʈaːməˌnuːɹəˌnaː.it ˈwaβeˌweːjem
|naⁿ-tramə-nu-təna-it əʷ-wə-pe-weʷ-iɲem|
DIST-forest-LOC-go-ACT 3s.O-1s.A-NP-HAB-like

I enjoy walking in the forest.

Another common collocation is the co-occurrence of the mediopassive — giving a reflexive interpretation — with an incorporated body part:

Wēꞵanāwət.
/ˈweːβaˌnaːwət/
|əʷ-weꞵa-nə-awət|
3s.O-hair-MED-wash

I’m washing my hair.

Verbal morphology: other forms

Participles

Each verb has two participle forms: an agent participle in -kɨ and a patient participle in -kəŋ. The agent participle may be formed from transitive and agentive intransitive verbs, by removing the agent cross-referencing prefix and adding -kɨ to the end. Similarly the patient participle may be formed from transitive and patientive intransitive verbs, by removing the patient prefix and adding -kəŋ.

Participles are most commonly used to form relative clauses. A participle by itself gives a ‘headless relative’:

ēꞵunīkɨ
/ˈeːβuˌniːkɨ/
|ə-e-punī-kɨ|
3s.O-NP-see-AGT

‘Who sees me’

mēŋkəŋ
/ˈmeːŋkəŋ/
|əmeŋ-kəŋ|
die-PAT

‘The dead one’

More complex relative clauses are internally headed (which in most cases places the head before the participle):

ŋkrazē təmānothēŋkəŋ
/ŋkraˈzeː təˈmaːnoˌɹeːŋkəŋ/
|ŋkrase tə-maⁿ-othem-kəŋ|
food 2s.A-PF-eat-PAT

The food you ate

Headless relatives in particular see extremely productive use in Wēchizaŋkəŋ discourse for nominalising verbs, and many have been lexicalised (I gloss these with colons and square brackets following Seri convention):

chathaŋkəŋ
/ɰaˈɹaŋkəŋ/
|chat:ən:kəŋ|
˻one:ORD:OBJ˼

‘the first one’ = leader

thēchepkɨ
/ˈɹeːɰepkɨ/
|˻əʷ:tha:kiep:kɨ˼|
˻3s.O:INCH:fenestrated:AGT˼

‘it has holes’ = delicious monster (M. deliciosa)

The language name Wēchizaŋkəŋ is itself a headless relative: |iʷ-weʷ-kizəm-kəŋ| ‘what we speak’.

Participles are also used to form content questions. For nominal interrogative words, the questioned component is participlised and equated to the question word:

Ꞵīɲ kəꞵarē rəꞵāmpunīchəŋ yachrəthāmɨŋ?
/βiːŋ kəβaˈreː rəˈβaːmpuˌniːɰəŋ ˈjaɰrəˌɹaːmɨŋ/
|ꞵiɲ kə=pare rəʷ-maⁿ-punī-kəŋ əʷ-rəʷ-krə-tham-ɨŋ|
what 2s.POSS=dog 3s.A-PF-see-OBJ 3s.O-3s.A-IRR-is-Q

What did your dog see? (lit. ‘what is it that your dog saw’)

Verbal question words do not require the copula, but their complement is still participalised:

Kəꞵarē tirī rəꞵāmpunīchəŋ chraꞵachīŋ?
/kəβaˈreː tiˈriː rəˈβaːmpuˌniːɰəŋ ˈɰraβaˌɰiːŋ/
|kə=pare tiri rəʷ-maⁿ-punī-kəŋ əʷ-krə-ꞵachi-ɨŋ|
2s.POSS=dog bird 3s.A-PF-see-OBJ 3s.O-IRR-how.many-Q

How many birds did your dog see? (lit. ‘how many are the birds your dog saw’)

Note of course that nouns cannot be participlised: thus a question like ‘how many birds are there’ is simply Tirī chraꞵachīŋ?.

As there are is no indirect object participle, only subjects and direct objects can be relativised on or questioned. Extracting indirect objects thus requires an applicative:

Sēmanamūtriŋkaŋ chraꞵethēŋ?
/ˈseːmanaˌmuːʈiˌŋkaŋ ˈɰraβeˌɹeːŋ/
|əʷ-tə-e-maⁿ-am-utriŋ-kəŋ əʷ-krə-ꞵethe-ɨŋ|
3s.O-2s.A-NP-PF-BEN-draw-OBJ 3s.O-IRR-how-Q (lit. ‘how is how you drew it’)

How did you draw it?

Action nominals and complement clauses

Action nominals are formed similarly to agent participles, by omitting the agent prefix and adding the action nominal suffix -ic. Action nominals are highly restricted in their inflectional possibilities: they may maximally take object prefixes, directional/postural prefixes, and negation with obligatorily co-occurring irrealis prefix. If an agent is present, it is cross-referenced using the nominal possessive proclitics:

Kəēwathīc wēthīshu.
/kəˈeːwaˌɹiːc ˈweːɹiːɕu/
|kə=əʷ-e-wat-ic əʷ-wə-e-tīshu|
2s.POSS=3s.O-NP-do-ACT 3s.O-1s.A-NP-ask

I’m asking you to do it.

As shown in the example above, action nominals are regularly used as a complementation strategy. Others include participals and relative clauses, but most commonly such structures simply use juxtaposition:

Ēkrəzənūthənā, ēkrəꞵamū.
/ˈeːkrəzəˌnuːɹəˌnaː | ˈeːkrəβaˌmuː/
|əʷ-ə-e-krə-sə-nu-təna əʷ-ə-e-krə-pamu|
3s.O-1s.O-NP-IRR-ABL-LOC-go, 3s.O-1s.A-NP-IRR-take

If I could reach it, then I would take it.

There is exactly one type of ‘genuine’ complement clause, formed by placing the complementiser tə(y)= after the clause. (Note that this is a ‘wrong-way’ clitic, i.e. it attaches to the following word rather than the preceeding complement clause. The clitic-final semivowel is included only if the following word starts with a vowel.) This construction is rather restricted in its range, occurring primarily with mental process predicates:

Wēthawat təyēntishō.
/ˈweːɹawat təˈjeːntiˌɕoː/
|əʷ-wə-e-tha-wat təy=əʷ-e-ntisho|
3s.O-1s.A-NP-INCH-do COMP=3s.O-1s.A-think

I think I am finished making it.

bradrn wrote: ↑Mon Jul 26, 2021 10:35 pm OK, so I might have gotten just a little bit carried away with this…

I think I like it when you get carried away with ideas, this is great! I was hoping to learn more about the morphophonological shenanigans going on with this particular conlang.

Oddly, my phone browser (chrome on android) doesn't like the bold ꞵ, but normal ꞵ is fine. Neither could it display the mutation table in the code box.

bradrn wrote: ↑Mon Jul 26, 2021 10:35 pm Wēthawat təyēntishō.
/ˈweːɹawat təˈjeːntiˌɕoː/
|əʷ-wə-e-tha-wat təy=əʷ-e-ntisho|
3s.O-1s.A-NP-INCH-do COMP=3s.O-1s.A-think

I think I am finished making it.

Is a conlang ever finished?? Either way, I look forward to seeing more of it! I'd especially like to hear some audio recordings of it

I like it. The design choices you mention seem pretty limited compared to how things turned out. Did you find that those choices led to others that just sort of felt right, or was the rest sort of ad hoc and just whatever you fancied in the moment?

This specifically jumped out at me:

bradrn wrote: ↑Mon Jul 26, 2021 10:35 pm Direction/posture

A direction or posture prefix is placed in slot -7. There are four such prefixes: tha- ‘andative/going’, wəŋ- ‘venitive/coming’, sə- ‘sitting/lying’, əŋə- ‘standing’. The locatives and posturals relate to the agent argument for transitive verbs, and the only argument for intransitive verbs.

I haven't seen posture mixed in w/ deixis, and I think it's a cool feature. You call it "direction", and with the postures it does remind me of "directionals" from Qiang & other langs, which I used in my Dwarven. I could see some interesting directions for descendant langs. These could additional directionals that are more spatial, like Qiang. More interestingly, I could maybe see these developing into volition marking or perhaps voice. Having a wide range like that in different descendant lines is interesting to me.

Any plans for this lang, or is it a sort of "conlang isolate", never to be used in any wider work?

quinterbeck wrote: ↑Tue Jul 27, 2021 1:18 pm

bradrn wrote: ↑Mon Jul 26, 2021 10:35 pm OK, so I might have gotten just a little bit carried away with this…

I think I like it when you get carried away with ideas, this is great! I was hoping to learn more about the morphophonological shenanigans going on with this particular conlang.

There is no more to learn: though I may have more to say about some other sections, I tried to be as comprehensive as possible in describing the morphophonology.

Oddly, my phone browser (chrome on android) doesn't like the bold ꞵ, but normal ꞵ is fine. Neither could it display the mutation table in the code box.

Could you provide a screenshot please? As for ⟨ꞵ⟩, I’ve noticed fonts on other platforms tend to struggle a bit with U+A7B5 LATIN SMALL LETTER BETA.

Either way, I look forward to seeing more of it! I'd especially like to hear some audio recordings of it

Alas, my tongue gets all twisted when I try to say these sentences, so no audio recordings for now.

Vardelm wrote: ↑Tue Jul 27, 2021 7:13 pm I like it. The design choices you mention seem pretty limited compared to how things turned out. Did you find that those choices led to others that just sort of felt right, or was the rest sort of ad hoc and just whatever you fancied in the moment?

There were some areal constraints as well — I needed to have a way to create headless relatives in one word, for instance. I also definitely had some ideas floating around in my head for what I wanted to do. And of course, this being me, the whole thing needed to be as naturalistic as possible, which did add quite a few additional constraints — there’s only so many ways to make a polysynthetic language. Beyond that, there were essentially no limits.

bradrn wrote: ↑Mon Jul 26, 2021 10:35 pm Direction/posture

A direction or posture prefix is placed in slot -7. There are four such prefixes: tha- ‘andative/going’, wəŋ- ‘venitive/coming’, sə- ‘sitting/lying’, əŋə- ‘standing’. The locatives and posturals relate to the agent argument for transitive verbs, and the only argument for intransitive verbs.

I haven't seen posture mixed in w/ deixis, and I think it's a cool feature.

I got it from Caddo.

More interestingly, I could maybe see these developing into volition marking or perhaps voice.

Sounds interesting, could you elaborate?

Any plans for this lang, or is it a sort of "conlang isolate", never to be used in any wider work?

Wait and see

bradrn wrote: ↑Tue Jul 27, 2021 8:42 pm I got it from Caddo.

Groovy. I think I've looked at Caddo a tiny bit in the past, but not enough to pick this feature up.

bradrn wrote: ↑Tue Jul 27, 2021 8:42 pm Sounds interesting, could you elaborate?

I picture "standing" as "being on your feet; able to act/react", while "sitting/lying" is more passive. So, "standing" could maybe evolve into an active (transitive? causative?) voice or volitional mood. "Sitting/lying" would be for a passive voice or non-volitional mood. You could have 3 different descendant lines: one develops more directionals, another the posture/direction morphs in voice, and the final one posture/direction changes to volition.

bradrn wrote: ↑Tue Jul 27, 2021 8:42 pm Wait and see

Oh, you tease!!!

Vardelm wrote: ↑Tue Jul 27, 2021 9:23 pm

bradrn wrote: ↑Tue Jul 27, 2021 8:42 pm Sounds interesting, could you elaborate?

I picture "standing" as "being on your feet; able to act/react", while "sitting/lying" is more passive. So, "standing" could maybe evolve into an active (transitive? causative?) voice or volitional mood. "Sitting/lying" would be for a passive voice or non-volitional mood. You could have 3 different descendant lines: one develops more directionals, another the posture/direction morphs in voice, and the final one posture/direction changes to volition.

Interesting idea. Personally I tend to associate ‘standing’ with lack of movement: languages often represent predicative locationals with a verb ‘stand’. I’d be interested to know if this is a universal, since your interpretation sounds reasonable also.

But in my experience, idiomatic meanings tend to attach more often to directionals than posturals: ‘I finished up’ (perfective), ‘You should sing along’ (imperfective I think), ‘He went and burned it’ (adversive? malefactive?). Or perhaps that’s just because directionals are far more common than posturals? I’ll have to read more about this.

bradrn wrote: ↑Tue Jul 27, 2021 8:42 pm

quinterbeck wrote: ↑Tue Jul 27, 2021 1:18 pm I think I like it when you get carried away with ideas, this is great! I was hoping to learn more about the morphophonological shenanigans going on with this particular conlang.

There is no more to learn: though I may have more to say about some other sections, I tried to be as comprehensive as possible in describing the morphophonology

My bad, I meant to imply my hopes had been fulfilled!

bradrn wrote: ↑Tue Jul 27, 2021 8:42 pm

Oddly, my phone browser (chrome on android) doesn't like the bold ꞵ, but normal ꞵ is fine. Neither could it display the mutation table in the code box.

Could you provide a screenshot please? As for ⟨ꞵ⟩, I’ve noticed fonts on other platforms tend to struggle a bit with U+A7B5 LATIN SMALL LETTER BETA.

Here's the mutation table


Here's an ok ꞵ and a problematic one


Here's lots of problematic ones


The ꞵ used in the IPA transcription lines of the glosses is always ok

quinterbeck wrote: ↑Wed Jul 28, 2021 11:40 am Here's the mutation table

Oh dear, that’s worse font coverage than I thought. Here’s what it looks like for me:

Here's an ok ꞵ and a problematic one
[img]

Here's lots of problematic ones
[img]

The ꞵ used in the IPA transcription lines of the glosses is always ok
[img]

Ah, I see the problem. The character in the IPA transcriptions is U+03B2 GREEK SMALL LETTER BETA, but the romanisation uses U+A7B5 LATIN SMALL LETTER BETA. The former has good font coverage, the latter doesn’t.

Oh, this is that conlang with weird romanization choices for some consonants that you've been posting in the Conlang fluency thread!

bradrn wrote: ↑Mon Jul 26, 2021 10:35 pmSurface primary stress is consistently on one of the first two syllables. In most case, the stressed syllable can be predicted by the following rules, although they occasionally fail:

This is not fixed stress location though...

bradrn wrote: ↑Mon Jul 26, 2021 10:35 pm

Code: Select all

underlying:  p  t  ʈ  tɕ  k β r j m n ɲ ŋ w 0
     nasal: ᵐp ⁿt ᶯʈ ᶮtɕ ᵑk m n ɲ m n ɲ ŋ m n
      weak:  0  s  ʂ  ɕ   ɰ w j j β r j ɰ w 0
  lenition:  ꞵ  ɹ  ʈ  z   ɰ β r j m n ɲ ɰ w 0

Is tɕ a cluster? Because you haven't listed any affricates in the phoneme inventory.

I have to say this doesn't look like any other language, at least phonology-wise. All the rules are a little too complicated for me to keep track of, but that's okay. I have a conlang with so complicated allophony that even I can't comprehend it all myself.

Note to self: Basic syntax and morphology is the next part to read through.

Qwynegold wrote: ↑Sat Jul 31, 2021 3:17 pm Oh, this is that conlang with weird romanization choices for some consonants that you've been posting in the Conlang fluency thread!

Yes, indeed it is!

The romanisation, by the way, is taken from the Baining language Mali. The idea is that ⟨t k⟩ lenite to ⟨th ch⟩. (And for consistency I should also have done /β/ ⟨ph⟩, but I like ⟨ꞵ⟩. Mali uses ⟨v⟩ for the same sound.)

bradrn wrote: ↑Mon Jul 26, 2021 10:35 pmSurface primary stress is consistently on one of the first two syllables. In most case, the stressed syllable can be predicted by the following rules, although they occasionally fail:

This is not fixed stress location though...

It’s fixed given the underlying form.

bradrn wrote: ↑Mon Jul 26, 2021 10:35 pm

Code: Select all

underlying:  p  t  ʈ  tɕ  k β r j m n ɲ ŋ w 0
     nasal: ᵐp ⁿt ᶯʈ ᶮtɕ ᵑk m n ɲ m n ɲ ŋ m n
      weak:  0  s  ʂ  ɕ   ɰ w j j β r j ɰ w 0
  lenition:  ꞵ  ɹ  ʈ  z   ɰ β r j m n ɲ ɰ w 0

Is tɕ a cluster? Because you haven't listed any affricates in the phoneme inventory.

Whoops, that’s a typo; it should be /c/. (It was /tɕ/ in an early revision of the phonology.) I shall fix that momentarily.

Now I've finished reading it. How long did it take you to make this? I never get that much done on grammar.

bradrn wrote: ↑Sat Jul 31, 2021 7:41 pm

bradrn wrote: ↑Mon Jul 26, 2021 10:35 pmSurface primary stress is consistently on one of the first two syllables. In most case, the stressed syllable can be predicted by the following rules, although they occasionally fail:

This is not fixed stress location though...

It’s fixed given the underlying form.

Is this how it's usually done? I mean in natphonology.