Random phonology/phonemic inventory thread

[DV82LECM]

Would written geminated /ʔ/ be <">? It would be intuitive but it strikes my eye oddly.

Also, fan of them both.

Probably yes.
Some people use <q> for /ʔ/ but I don't like it either. Do you have some other options?

I thought the apostrophe was kinda cool. I don’t think i have ever seen anyone do it, that is what I meant. It would be different.

[Omzinesý]

My romlang idea.

p t t͡s k (ks) <p t c/ç c/qu (x)>
f θ s x <f ŧ s ꝁ>
v ð z ɣ <v d z g>
m n ɲ <m n gn>
l r j <l r j>

Having <ꝁ> without <k> is a bit strange. Maybe <c̄>.

[Creyeditor]

[...]

/p t ts k/
/f s x/
/m n/
/v~w l/
[...]
ps px tf tx tsf tsx kf ks
(fs) fx sf sx (xf) (xs)
ml (nv)
vl (lv)

I like the /pn kn/ cluster. I should maybe add them (and also /fn/) and allophonically devoice nasals in there. And maybe [f] is really an allophone of /v/ (or [v] of /f/) at least in clusters they are in complementary distribution. That leaves me with /pl/ and /kl/ clusters, which I want to exclude because they are boring but they are also kind of parallel to /pn/ and /kn/ and I really like the lateral in /ml/. Maybe /l/ is just special in that it cannot devoice or something.

Today, I thought about another way to make this work. I could introduce a voicing (or maybe aspiration contrast in obstruents.

/p b t d ts dz k g/
/f v s z x/
/m n/
/w l/

ps px tf tx tsf tsx kf ks
(fs) fx sf sx (xf) (xs)
bn (dm) (dzm) gn (gm)
(bl) dw dzw (gl) (gw)
(vz) zv
(vn) (zm)
ml (nw)
vl zw
wl (lw)

This gives me nice clusters /bn/, /gn/, /dw/, /dzw/ /zv/, /zw/. OTOH, it also creates /bl/, /gl/, which I would like to avoid. So maybe it really just shifts the problem.

[VaptuantaDoi]

A Stubby-Holder sister language, moving slightly away from the Australian aesthetic.

/p t̻ t k/ p th t k
/v s̻ ɣ/ v sh g
/m n̻ n ŋ/ m nh n ng
/d̻n̻ dn/ dnh dn
/l̻ l/ lh l
/d̻l̻ dl/ dlh dl
/ɾ ɹ/ rr r
/w ð̞ ̻ j/ w yh y

/e o a/ e o a

th sh nh dnh lh dlh yh → [c ç ɲ ɟɲ ʎ ɟʎ j] tj sj nj dnj lj dlj y / _ e
t n dn l dl rr → [ɖ ɳ ɖɳ ɭ ɖɭ ɽ] rt rn rdn rl rdl rd / _ o

Maybe also

g ng → Ø / # _

Final vowels elided with antepenultimate stress

Frequent morphophonological alternation between p k th tj t rt and v g sh sj rr rd. Maybe even full-fledged consonant mutation:

Code: Select all

 Plain Hard  Nasal
   v     p     mp
   g     k    ngk
   sh    th   nhth
   sj    tj   njtj
   rr    t     nt
   rd    rt   rnrt
   m     m     m
   ng    ng    ng
   nh   dnh    nh
   nj   dnj    nj
   n     dn    n
   rn   rdn    rn
   lh   dlh    nh
   lj   dlj    nj
   l     dl    n
   rl   rdl    rn

Some meaningless sample words:

mpathwaga
athamayorlonhk
rdnumpaljtjenop
rdosyevelnkaparram
Yhodnhamanrolkapngke
Yodnyamanhthardl
Shayompernrtongadly

[thethief3]

Littoral Zeluzhia Revamp
*m *mʲ *n *nʲ *ŋʲ *ŋʷ
*b *bʲ *d *dʲ *gʲ *gʷ
*p *pʲ *t *tʲ *kʲ *kʷ *qʲ *qʷ
*s *sʲ *ɬ *ɬʲ *ʁʲ *ʁʷ
*l *lʲ
*ɾ
*j *w

*i *i: *u *u:
*e *e: *ø *ø: *o *o:
*a *a:

[VaptuantaDoi]

Some more Stubby sisterlang ideas expanded from the Romanisation game:

/p t tɬ tʂ tʃ k kʷ/ p t tl rt ty k kw
/m n ɳ ɲ ŋ ŋʷ/ m n rn ny ng ngw
/w r l j ɻ/ w rr l y r

/e o a/ e o a
/aː/ aa


Syllable structure is probably CV(m n ɳ ɲ ŋ ŋʷ w l r j ɻ) with some CC clusters prohibited. All odd-syllable words gain a meaningless -pa following all morphological processes, and then following this all consonant-final words gain a. Stress is placed on the initial on a word with an even number of syllables and the second if odd. Second-syllable stress causes lenition of initial stops p t tl rt ty k kw → w rr l r y Ø w and collapse of all vowels to a.

rtowtya → rtowtya
rtowtya-perrng → rtowtyaperrngpa
rtowtya-kwalar → rawtyákwalara

---

/p t k/ p t k
/m n ŋ/ m n ng
/w r ɻ j/ w rr r y

/i ɨ u a/ i ı u a

Basically just a mini-Australian inventory. Stops and /r/ are prohibited word-initially; root-initial p t k rr become w r w y when exposed to a left word edge. This means all words begin with one of m n ng w r y.

mayu
nurapka
ngarrarpirka
wara
rintarmu
yarmkıngka

Word-final codas are repaired with a following k or a single consonant, ka otherwise:

runap- → runapka
ranwuk- → ranwuka

But

runap-ari-wayk-anga → runawariwaykanga

Word-internal clusters are allowed up to three consonants, e.g. mayngpirrka, raymparrarkmurnaripka, rarrinampkiyka. There's automatic nasal insertion into reduplicated forms too; this would produce more clusters:

runap- → runawntunapka
warr- → warrngkarrka

Maybe these two live next to each other, and dialects of #2 have borrowed the stress rule from #1:

rúnawntùnapka ~ ranáwntonàpka
ríntarmu ~ rantármo

[VaptuantaDoi]

Based on something Gleb gave me:

/p t k q/
/b d g/
/ɓ ɗ/
/f s/

/i u ɛ ɔ a/
/ɛ̃ ɔ̃ ã/

Now for a little twist:

/ɓ ɗ q/ → [m n ŋ] | _ /ɛ̃ ɔ̃ ã/

And also

/ɓ ɗ/ → [w l] | /i u ɛ ɔ a/ _

I will never do anything with this but hey.

[Solarius]

More of a random idea: morphophonology, but whatever. This is something I've been working on for my latest sketch.

Inventory is pretty simple:

/p t ts tɬ k q/<p t ts tl k q>
/s ɬ h/<s hl h>
/m n/<m n>
/l ɹ w/<l r w>

/i ḭ e ḛ a a̰ o o̰ u ṵ/

Phonotactics are in the most general terms C(V)V(V)(ɹ), but the rules are more complicated than that:
-The only triphthongs which are allowed begin with /i ḭ u ṵ/.
-Diphthongs and Triphthongs have to be homorganic in creaky voice.
-Final /ɹ/ can only coexist with monophthongs and the opening diphthongs /ie ia io ue ua uo/ (including creaky forms); thus /Caɹ/ and /Ciaɹ/ are allowed, but not */Caiɹ/ or */Cieaɹ/. It's also only allowed word-finally (and is generally pretty rare).
-Geminates are allowed intervocallically.

Stress is right-edged; there's no consistent rule as to whether it falls on the penultimate or final syllable.

Allophony takes these neat CV syllable shapes and messes them up a bit.
-Several phonemes fortite when geminated: /ɹː sː/ --> [r st]
-When the stress is on the pneultimate syllable, it causes the deletion of the posttonic vowel, thus creating a phonetic but not phonemic coda consonant: /ki.'ta̰.na̰/ --> [ki.'tan]. These can re-emerge when followed by affixes.
-When /w/ is forced into coda position by the preceding rule, it becomes {b}: /'t͡sḭ.we/ --> [t͡sḭb]
-Coda /ɹ/ becomes rhotacization on the preceding vowel: /t͡ɬea.'so.ɹa/ --> [t͡ɬea.'so˞˞]

These can all re-emerge when followed by affixes.

Morphophonology is kind of a wild time.

There are two kinds of suffixes (and clitics), stress-attracting and stress-neutral. Stress-attracting suffixes pull the stress onto themselves, unless they’re followed by another stress-attracting suffix or more than one subsequent syllable (wherein the stress defaults to the penultimate syllable, though this is rare). As a result, these suffixes usually retain their shape and aren't deleted, since they carry stress or aren't in posttonic position.

/t͡ɬoi/ --> /t͡ɬoi.'la/
[t͡ɬoi] --> [t͡ɬoi.'la]
burn --> burn-TR

Stress-neutral suffixes and clitics do not pull down stress in the same way. As a result, their final vowels are reduced frequently if preceded by a stressed syllable.

/t͡ɬoi/ --> /'t͡ɬoi.nːa/
[t͡ɬoi] --> ['t͡ɬoi.nː]
burn --> burn-INCH

Since stress has to fall on either the penultimate or final syllable, in a situation where there's a pileup of multiple stress-neutral suffixes it defaults to the penultimate.

/t͡ɬoi/ --> /'t͡ɬoi.nːaka/
[t͡ɬoi] --> ['t͡ɬoi.nːak]
burn --> burn-INCH=SR

Aside from the various stress-related rules, there's also consonant harmony processes in suffixes. When the preceding syllable has /q/, /k/ in a suffix becomes /q/ (and vice-versa).

[Flavia]

/i u ɤ/
/m n̻ n̻ʲ n̻ʷ/
/p b ɓ t tʼ d ɗ k kʼ g kʲ kʼʲ gʲ kʷ kʼʷ gʷ/
/s̻ x xʲ xʷ h/
/r j w/

(C)([r j w])V(C[+pulmonic])

i u ɤ > e o ɑ / _C[-pulmonic]
i u ɤ e o ɑ > y u o ø o ɒ / Cʷ_
i u ɤ e o ɑ > i ʉ ɘ e ɵ a / Cʲ_
o ɤ ɘ > u̯o ɨ̯ɤ i̯ɘ / _[r j w]

Based on gleb seed 14928697642.

[Visions1]

Phonology of that culture thing I just posted about in C&C.
Brackets mean I'm unsure if I want it. I'm not doing anything with this right now anyways.

(/m̥ n̥/)
/m n/
/p t k/
(/b d g/)
/s h/
/r/
(/l/)
/w j/

/i u/
/e ə/ alternatively, replace /ə/ with /y/
/a/

Consonants may be doubled. Syllables (morae, call them whatever) can only take one vowel, but you can string as many vowels as you like in a row. So /hyppeastrumayyeiu/ is valid. (Planned on making it agglutinating or highly highly polysynthetic or something.)

[Omzinesý]

pʰ tʰ kʰ
p t k
b d g
t͡s t͡ɬ t͡ʂ ʈ͡ɬ
d͡z d͡ɮ d͡ʐ ɖ͡ɮ
s ɬ ɕ ʂ ɬ
z ʑ ʐ
mʰ nʰ
m n ŋ
l r ɭ ɽ
s ʃ x
z ʒ
ʋ j

i ɨ u
e ə o
ä

Most words are monosyllabic.
The syllable structure is: C(G)V(C), where G is a glide (ʋ j).

When the coda is an obstruent or lacks, the syllable has one of two tones (rising or lowering).
If the coda is voiceless, the syllable has one of three tones (neutral, rising, lowering). The vowel with a neutral tone is shorter and the coda consonant longer. (This is somewhat based on Lithuanian tones.)

a:˩˥r
a:˥˩r

a:˩˥t
a:˥˩t
at:

[VaptuantaDoi]

Something a little bit silly – the phoneme inventory of Delaware singing:


/t tʃ k/
/n j w h/

/a e i o/ + length

/tʃ/ only occurs before /i/
/w/ does not occur before /o/.


I love how it forms a complete little inventory all of its own.

[Visions1]

I was wondering what a vocables language would look like. Nice.
I'd like to see an Amerind one (maybe that's too vast?)

Oh hey, I'm Greek now.

[Nortaneous]

/p k r~n/
/a e o i u/

/ioroa arkaoor paraikueiap epporkukaik̩ eikapirparko rikakkaoapir/
[tsoɾo̯a aŋˑaoːn paɾai̯su̯ei̯ap epːoŋguxai̯t ei̯saɸimˑaŋgo nisakˑao̯aɸin]

[Omzinesý]

An unnamed lang


Vowels (The inventory happens to be like that of Pashto.)
i u <i u>
e ə o <ai e au>
æ ɒ <a o>

Diphthongs
ei~æi, eu~æu, oi~ɒi, ou~ɒu <ei, eu, oi, ou>
(The first component can thus be closed or mid.)

The stress always lies on the second syllable.

- ə only appears in unstressed syllables.
- Diphthongs usually lie on the stressed / second syllable. They can though appear on the morphemic border.

Words that are longer than two syllables and end in a vowel always end in ə.


The consonant inventory has African feel.

p t t͡ʃ k
b d d͡ʒ g
ɓ ɗ ɖ͡ʒ
f θ s ʃ x
v ð z ʒ ɣ
m n ɲ
l r ʎ
j w

Phonotactics is mostly (C)V(C). Very rarely coda clusters nasal + obstruent can appear.
Nasals and liquids could be syllabic and even stressed.

Verbs are prefixing, and verb roots usually CCVC.

d͡ʒ-æ-xled
SG1-PRES-sing
'I'm singing.'

Nouns are suffixing and noun roots typically CVCC.

All nouns have a portmanteau suffix that codes
1) (around ten) Bantu-style noun class,
2) case (direct, possessed (like edafa), and ablative (appears with some prepositions and numerals)), and
3) number (singular plural).

This makes around 60 nominal suffixes (though some plural classes merger).

xileb is an abstract root for singing. 'A song' would be xilb-or.

[thethief3]

*p *k *ʔ
*h
*w *j
*m *n

*i *u
*e *o
*a

C(w, j)VC syllable structure were any consonant except *w can't be a coda
It's a descendant of Littoral Zeluzhia btw.

[thethief3]

and its descendant

*mb (ᵐb) *nd (ⁿd) *ng (ᵑɡ) *ngw (ᵑɡʷ)
*p (pʰ) *t (tʰ) *k (kʰ) *kw (kʷʰ)
*’p (p˭) *’t (t˭) *’k (k˭) *’kw (kʷ˭)
*ndl (ⁿd͡ɮ) *njl (ᶮɟʎ̝) *njj (ᶮɟ͡ʝ)
*tł (t͡ɬʰ) *cł (c͜ʎ̥˔ʰ) *cc (c͡çʰ)
*’tł (t͡ɬ˭) *’cł (c͜ʎ̥˔˭) *’cc (c͡ç˭)
*s (sʲ) *h (h)
*v (v) *l (l)
*w (w) *y (j)
*m (m) *n (n) *ń (nʲ)

*i (i) *u (u)
*e (ɛ) *o (ɔ)
*a (a)

*ī (i:) *ū (u:)
*ê (e:) *ô (o:)
*ē (ɛ:) *ō (ɔ:)
*ā (a:)

[Porphyrogenitos]

Another one of my ideas for an essentially monosyllabic language. Taking a more naturalistic approach than some of my previous attempts. I wanted to have at least 1000 distinct syllables, but it was difficult to reach that without getting too kitchen-sinky.

This scheme has about 766 distinct syllables, barring any accidental phonological gaps.

/m mʲ mʷ n nʷ ɲ ŋ ŋʷ/ <m my mw n nw ny ng ngw>
/m̥ m̥ʲ m̥ʷ n̥ n̥ʷ ɲ̊ ŋ̊ ŋ̊ʷ/ <mh mhy mhw nh nhw nhy ngh nghw>
/p pʲ pʷ t tʷ c k kʷ/ <p py pw t tw c k kw>
/b bʲ bʷ d dʷ ɟ g gʷ/ <b by bw d dw j g gw>
/ᵐp ᵐpʲ ᵐpʷ ⁿt ⁿtʷ ᶮc ᵑk ᵑkʷ/ <mp mpy mpw nt ntw nc nk nkw>
/ᵐb ᵐbʲ ᵐbʷ ⁿd ⁿdʷ ᶮɟ ᵑɡ ᵑɡʷ/ <mb mby mbw nd ndw nj ngg nggw>
/f fʲ fʷ s sʷ ç x xʷ h/ <f fy fw s sw yh x xw h>
/v vʲ vʷ z zʷ j~ʝ ɣ w~ɣʷ/ <v vy vw z zw y gh w>
/r rʷ l lʷ/ <r rw l lw>
/r̥ r̥ʷ l̥ l̥ʷ/ <rh rhw l lhw>

/i e a o u/ <i e a o u>
/ai̯ au̯ eu̯ oi̯ iu̯ ui̯/ <ai au eu oi iu ui>

Plain and palatalized consonants do not contrast before /i iu̯/.
Non-velar plain and labialized consonants do not contrast before /u ui̯/.

After plain velar consonants, /u ui̯/ are realized as [ɨ ɨi̯].

Vowels are nasalized after nasal consonants (not including pre-nasalized consonants).

Nasalization spreads rightward across tightly-tied morphemes, e.g. in compounds and certain syntactic constructions. Nasalization spreads through /j w/ (and their voiceless variants) as well as /h/ and zero-onset initials (which are often realized with [ʔ]), but is blocked by other segments. In some dialects, nasalization may spread through voiced fricatives.

Morphemes and basic words are monosyllabic. Some lexical items (mainly auxiliary verbs) undergo stem changes to encode inflectional categories.

Historical notes:

Original palatalized velar and palatalized alveolar consonants merged as palatals.

One of the main sources for the nasal/prenasalized voiced stop distinction is a historical nasal vs. oral vowel contrast (which in turn probably originated from a historical nasal coda). Nasal consonants became voiced prenasalized stops before oral vowels, and then nasality disappeared after non-nasal consonants.

Voiceless prenasalized stops, which are less frequent, had other sources, mainly remnants of prefixation of nasal elements.

I may end up adding series of voiceless and voiced prenasalized fricatives, which would also have origins in nasal prefixation.

[Shemtov]

Possible third version of Classical Hanese:
/p pʰ b t tʰ d t͡s t͡sʰ d͡z ʈ͡ʂ ʈ͡ʂʰ d͡ʐ t͡ɕ t͡ɕʰ d͡ʑ k kʰ g/ <p p' b t t' d c c' z ch ch' zh q q' j k k' g>
/m n ɳ ɲ ŋ/ <m n nh ń ŋ>
/s ʂ ʐ ɕ h/ < s sh r x h>
/l j w/ <l î w>

/ i ɨ ʉ u/ <i y ư u>
/e ɵ o/ <ê ơ ô>
/ɛ ɐ ɔ/ <e ă o>
/a:/ <a>

/iʉ iu ie iɵ io iɛ iɐ iɔ iɑ/ <iư iu iê iơ iô ie iă io ia>
/ui uɨ ue uɵ uo uɛ uɐ uɔ uɑ/ <ui uy iê iơ iô ie iă io ia>
/ai aɨ au aʉ ɐi ɐu ɐɨ ɐʉ ɛu ɔi oɨ/ <ai ay au aư ăi ău ăy ăư eu oi oy>
/iai iaɨ iau iɐi iɐu iɐɨ iɛu iɔi iɔɨ/ <iai iay iau iaư iăi iău iăy iăư ieu ioi ioy>
/uai uaɨ uau uaʉ uɐi uɐu uɐɨ uɐʉ uɛu uɔi uɔɨ/ <uai uay uau uaư uăi uău uăy uăư ueu uoi uoy>

/˩ ˥ ˥˩/ (shown on a) <a á ah>

Phonotactics: (C)V(C)

Final consonants: /p t k m n ŋ/

The falling tone can only occur on open syllables.

The consonant ineventory is inspired by Middle Chinese, and thus the <C' > for /Cʰ/ is inspired by Wade-Giles.

[Porphyrogenitos]

Another one of my ideas for an essentially monosyllabic language. Taking a more naturalistic approach than some of my previous attempts. I wanted to have at least 1000 distinct syllables, but it was difficult to reach that without getting too kitchen-sinky.

Actually I think I'm going to discard the monosyllabic idea - I think this works much better as an inflectional language, and I'm going to drop the palatalized/labialized interpretation since that was basically trying to pound a square peg into a round hole. /j/ and /w/ will be recognized as the second component of onset clusters, and I will also let /l/ and /r/ be part of onset clusters.

/j w r l/ all assimilate in voicing to the preceding consonant. I'm on the fence about the nasal spreading.