Snewáhk'toò

A forum for all topics related to constructed languages
Post Reply
User avatar
Omzinesý
runic
runic
Posts: 3545
Joined: 27 Aug 2010 08:17
Location: nowhere [naʊhɪɚ]

Snewáhk'toò

Post by Omzinesý »

Snewáhk'toò is a project I have been working some days.

Snewâh is the name of the tribe that speaks it. Katóò means 'language'.

My intention is to concentrate more on phonology, morphology, and morphophonology. I think I'm better at syntax so lets do something else.


There are still morphosyntactic ideas
- incorporation and compounding
- a word class of statives
- ergative word order: SVO, VS, but VSObl
- definite suffix articles
Last edited by Omzinesý on 14 Sep 2022 01:58, edited 4 times in total.
My meta-thread: viewtopic.php?f=6&t=5760
User avatar
Omzinesý
runic
runic
Posts: 3545
Joined: 27 Aug 2010 08:17
Location: nowhere [naʊhɪɚ]

Re: Snewáhk'toò - Phonological Stuff

Post by Omzinesý »

Its consonant inventory is the one below. It is inspired by Ket (spoken in Siberia). It lacks both p and g. Old Snewáhk'toò actually had both but *p -> f and *g -> h.

t k q
b d
f s h
m n
l ɾ
j w

There isn't much allophony. I'll explain it after phonotactics.
/s/, /h/, and /j/ are the only consonants that can geminate. They are also the allowed coda consonants.

The inventory of short vowels is the one below. It feels cozy to me, simplified Scandinavian. It's basically Hungarian vowel inventory, but all short and long vowels appear in the same places.

y i u
ø e o
ä

Vowels can be short or long (written with two letters.)
There is also a reduced/overshort vowel. Its length varies between a full vowel and complete syncope, depending on lect. The distinction between it and the full vowel does not distinguish meanings, so I deem it non-phonemic.

Phonotactics

Snewáhk'toò phonotactics is relatively complex. This is an easy system to explain because there are no clusters like /nl/ that would brake up the nice groups.

The pattern below does not take into account the reduced vowel. I'll explain it later.

(O)(R)V(C)

O = obstruent = {t k q b d f s h}
R = resonant = {m n l r j w}
V = vowel = {a aa ai e ee ei o oo oi ø øø øi i ii u uu y yy}
C = coda = {s h}

/ɾ/ cannot appear in as the first sound of a word. That's like Turkic. Words can however begin with a cluster containing it, like /tɾ/. My artistic goal for that is that I don't want the lang have [r] and I think it would be the natural allophone word-initially.

Consonant clusters having a nasal as the second component cannot shere the same POA, so *bm, *dn, and *tn are not allowed.

The semivowels /w/ and /j/ appear in consonant clusters only on morpheme boundaries.

bjóòna 'I speak', b- 'SG1', jóòna 'to speak'

Nahuatl allows only stops in coda, and my Dleesoop is inspired by it. Many languages allow only resonants. Allowing only fricatives was an innovation of this lang. It didn't fully succeed because of/N/ and /j/ but they are not as frequent codas as /s/ and /h/.

/N/ and /j/ are actually in complementary distribution in coda. /N/ appears before stops and /s/ and /h/, /j/ before other consonants and in the absolute word-end.
Edit: I got rid of coda nasals.

Allophony

- The coda nasal /N/ assimilates with the following consonant in POA.
- /h/ is something like [x] as the first part of a consonant cluster. In coda, it is something like [ħ].
- Nasals can be devoiced when they are preceded by a voiceless obstruent in a consonant cluster.

Accent

I got the idea for the accent system from Japanese, but IDK if Japanese actually has anything similar. This system is an old idea of mine but I haven't used it in any of my longer projects.

There are two tones, high and low. Phonemically, there are no contour tones. Phonetically, a low tone has to change to a high tone through a rising tone and the other way around through a lowering tone, and a single high tone is actually a rising and then lowering tone.

Words can either lack an accent (i.e. have no high tone) or have one accent domain. What I call an accent domain consists of adjacent syllables with high tones. It can cover all the word (i.e. the word has no low tones) or some syllables. In practice, most words longer than one syllable have a tone domain and it rarely covers all the word.

The Autosegmental Theory would say that all words have LHL pattern but any of the tree tones can be joined to any number of syllables, as long as they are not joined to same syllables, and all the tones don't have to be joined.

I decided to write the first syllable of the accent domain with an acute <á> and the last one with a grave <à>. If the accent domain is only one syllable long, it is written with a circumflex <â>.

Vowel reduction

Usually the vowels reduced are the closed ones, /y/, /i/, and /u/. Other vowels can be reduced if the vowel of the preceding syllable is the same as it.

There are many constraints for the reduction.

1) Vowels of closed syllables cannot be reduced.
2) Vowels of word-initial syllables cannot be reduced.
3) Vowels of word-final syllables cannot be reduced.
4) Vowels immediately preceding the first syllable of an accent domain cannot be reduced.
5) Vowels of the first syllable of an accent domain cannot be reduced.
6) Vowels of the last syllable of an accent domain cannot be reduced.
7) Vowels of syllables adjacent to a reduced vowel cannot be reduced.

As you see, the important thing in the accent domain is actually its boundaries, not what happens inside it.

If there are several syllables where reduction could happen, one starts from the leftmost one. There can be several reduced vowels per word but not adjacent ones.

It depends on your lect and register how many vowels you reduce. The actual degree of reduction also varies. Sometimes the vowel is just very short but preserves its place. Sometimes it is a short schwa. Sometimes it is omitted all together.

See that the onset of the syllable does not affect the reduction. There are words like tadm'kóò that basically have syllabic nasals.

I decided to write reduced vowels with the hyphen <'>.
Last edited by Omzinesý on 15 Sep 2022 15:58, edited 3 times in total.
My meta-thread: viewtopic.php?f=6&t=5760
User avatar
Omzinesý
runic
runic
Posts: 3545
Joined: 27 Aug 2010 08:17
Location: nowhere [naʊhɪɚ]

Re: Snewáhk'toò - Some Vocabulary

Post by Omzinesý »

Some vocabulary

NOUNS
kâba 'man'
kâfa 'woma'
katóò 'language'

VERBS
jóòna 'to speak'

STATIVES
fnimi 'to love'
My meta-thread: viewtopic.php?f=6&t=5760
User avatar
Omzinesý
runic
runic
Posts: 3545
Joined: 27 Aug 2010 08:17
Location: nowhere [naʊhɪɚ]

Re: Snewáhk'toò - Morphophonological Stuff

Post by Omzinesý »

Old Snewáhk'toò had four vowels: *i, *u, *e, *a. Coda *w merged with them creating:
*aw -> o
*ew -> ø
*iw -> y

There are some suffixes starting with /u/ that assimilate with the morpheme-final vowel. The vowel does not lengthen.

I think the definite article could be just "u".
kaba 'a man' -> kabo 'the man'

This is borrowed from my old conlang with the same vowel inventory.


Consonant stems with consonant-initial suffixes

This change is somewhat similar to Japanese onbin.

Coda stops are deleted lengthening the vowel
murat + qa -> muraaqa
mip + te -> myyte

Nasals stay nasal but assimilate in POA with the consonant
hewen + qa -> heweɴqa
qarem + qa -> qarøɴqa

Compounding is very common in Snewáhk'toò. It is 'normal' left-branching compounding of N+N and N+V (incorporation).

The rightmost root has an initial mutation when compounded.

f-> b
t -> d
k/q -> h
b -> m
d -> n
s -> r

The three first ones apply to first sound of clusters as well.

jûda 'house', frâ 'build' -> júdabrà 'to house-build'

All syllables between the accent domains of the roots have an accent.

If a consonant stem ending in /t/ or /d/ is attached to a suffix beginning with /t/ or /d/, the plosives become /s/s.

murat + te -> murasse
Last edited by Omzinesý on 15 Sep 2022 15:49, edited 1 time in total.
My meta-thread: viewtopic.php?f=6&t=5760
User avatar
Omzinesý
runic
runic
Posts: 3545
Joined: 27 Aug 2010 08:17
Location: nowhere [naʊhɪɚ]

Re: Snewáhk'toò

Post by Omzinesý »

Word classes

Yeah, I'm still going to syntax.

Nouns
- heads of NPs
- can modify a noun it follows and then appears in Oblique/Genitive case
- cannot be a predicate, must be preceded by any stative as a copula

Statives
- cannot be heads of NPs, but must be headed by a "dummy noun": 'person that is A-ing'
- can directly modify an NP it follows
- can be a predicate whose subject follows it

Verbs
- cannot be heads of NPs, but must be headed by a "dummy noun": 'person that is V-ing'
- cannot directly modify an NP but must be preceded by a relativizer hee ~ he
- can be a predicate
My meta-thread: viewtopic.php?f=6&t=5760
User avatar
eldin raigmore
korean
korean
Posts: 6094
Joined: 14 Aug 2010 19:38
Location: SouthEast Michigan

Re: Snewáhk'toò

Post by eldin raigmore »

Verbs
- cannot be heads of NPs, but must be headed by a "dummy noun": 'person that is V-ing'
If the verb isn’t intransitive or unaccusative or unergative,
can the dummy noun mean “person/place/thing that is V-ed” as well as or instead of
“person/thing that is V-ing”?


What about ditransitive verbs? Do you have any? You don’t have to; and you don’t have to have made up your mind yet. But when you know I’m sure we’d like to know too!
Last edited by eldin raigmore on 14 Sep 2022 21:58, edited 1 time in total.
User avatar
Omzinesý
runic
runic
Posts: 3545
Joined: 27 Aug 2010 08:17
Location: nowhere [naʊhɪɚ]

Re: Snewáhk'toò

Post by Omzinesý »

eldin raigmore wrote: 14 Sep 2022 21:57
Verbs
- cannot be heads of NPs, but must be headed by a "dummy noun": 'person that is V-ing'
If the verb isn’t intransitive or unaccusative or unergative,
can the dummy noun mean “person/place/thing that is V-ed” as well as or instead of
“person/thing that is V-ing”?


What about ditransitive verbs? Do you have any? You don’t have to; and you don’t have to have made up your mind yet. But when you know I’m sure we’d like to know too!
A good question
How is the role of the antecedent of a relative clause interpreted. There could be a inflecting relative pronoun or in-situ pronoun or it could be interpreted "form the context". I dunno yet. Surely the dummy word can be 'place' or 'time' or alike, too.
I don't even know if there will be nominalizations. There will not be participles or converbs. Embedded clauses just lack some categories that matrix clauses have.
My meta-thread: viewtopic.php?f=6&t=5760
User avatar
eldin raigmore
korean
korean
Posts: 6094
Joined: 14 Aug 2010 19:38
Location: SouthEast Michigan

Re: Snewáhk'toò

Post by eldin raigmore »

In English there are active “-ing” participles for agents and passive “-ed” participles for patients.
In your conlang I think you’ve said you won’t do it that way.
I know you’ll need to do it somehow, and I don’t know why one choice would be better than another.
Whatever you choose, I’ll assume you had good reasons for that choice.
I’ll be interested to read what it is.
User avatar
Omzinesý
runic
runic
Posts: 3545
Joined: 27 Aug 2010 08:17
Location: nowhere [naʊhɪɚ]

Re: Snewáhk'toò - Morphology of Nouns

Post by Omzinesý »

Nouns have
- 2 cases
- 2 two numbers
- 2 degrees of definiteness

Definiteness
The indefinite form is the unmarked one.
The definite suffix is -u.

- It rounds the last vowel of the stem.

kafa 'a woman'
kafo 'the woman'

- There are few stems that end in a rounded vowel. Then the suffix is -wo.

tlehto 'a worker'
tlehtowo 'the worker'

- There are few nouns with only a consonant stem. They have just -u added to them, but the stem can look very different.

qaah 'a place (that is somehow build, village ect.)'
qaatu 'the place'

Cases
I call the unmarked case Nominative. Typologically, it should be called Direct. It codes subjects, direct objects, and complements of copulas.
I call the other case Oblique. It has several functions: genitive, adjunct, complement of verbs that don't govern a direct object, complement of statives.

Oblique is usually formed suprasegmentally changing the accent domain.
- The accent domain will generally move to right. The exact rules will come later.
- If the word has just one syllable, a partial reduplication will appear to make it longer for the accent domain to move.

Number
Usually, Singular is the unmarked number.
Plurals are formed by
i) reduplicating the first syllable.
ii) lengthening the last vowel.
iii) combination of the two.

kâfaa 'women'
qaqaah 'places'
My meta-thread: viewtopic.php?f=6&t=5760
User avatar
Omzinesý
runic
runic
Posts: 3545
Joined: 27 Aug 2010 08:17
Location: nowhere [naʊhɪɚ]

Re: Snewáhk'toò - Basic syntax

Post by Omzinesý »

Basic clausal syntax

Verb clauses (clauses whose head is a verb)

Snewáhk'toò has an ergative word order, that is, the transitive construction has the word order: S V O X, (1),

(1) Kafo jehnu boboly.
kafa-w jehnu bo~boli-w
woman-DEF [3]burn PL-leaf-DEF
'The woman burned the leaves.'

but the intransitive construction has the word order V S X, (2).

(2) Boboly jeh.
bo~boli-w jeh
PL-leaf-DEF [3]burn
'The leaves burned'

The word order can thus be understood via A (absolutive) and E (ergative) roles as E V A X.

Such a word order really appears in some native languages.


Stative clauses (clauses whose head is a stative)

Statives appear in the intransitive construction. They can have an argument that is not a direct object in the X slot. It appears in the oblique case, (3).

(3) Fnihi kâfo kabâ.
fnihi kafa-w kaba
love woman.NOM-DEF man[INDEF].OBL
'The woman loves a man.'


I'm not sure about NP-internal word orders. The basic typology of S, V, and O isn't really applicable here. I think right-branching should be more "natural".
My meta-thread: viewtopic.php?f=6&t=5760
Post Reply