Random phonology/phonemic inventory thread

[Omzinesý]

Something with Germanic style

i: u: <í ú>
ɪ ʊ <i u>
e: o: <é ó>
ɛ ɔ <e o>
æ: ä ɒ: <á a å>

p t tˠ k (ʔ) <p t tg k q>
b d dˠ g <b d dg g>
f θ θˠ x <f þ þg x>
m n nˠ <m n ng>
l lˠ <l lg>
rˠ <r>
s sˠ <s sg>
j <j>

Voiced stops and voiceless spirants merge, when non-final, and become voiced spirants.

[Shemtov]

A proto-lang:
/p b t d c ɟ k g k͡p g͡b ʔ/ <p b t d c j k g kp gb ʔ>
/m n ɲ ŋ/ <m n ny ng>
/ɸ s ç h/ <f s ś h>
/ⁿv ⁿð ⁿʝ ⁿɣ/ <v z ź ğ>
/ɾ/ <r>
/l/ <l>
/w j/ <w y>

/i u ɐ a/ <i u ă a>
/i: u: a:/ <í ú á>
/ai au ɐi ɐu/ <ai au ăi ău>
/iɐ uɐ/ <iă uă>

Phonotactics: CV(C)
Permitted finals: /p b t d k g m n ŋ s l/
/k͡p g͡b/ can only occur root-initial.

[thethief3]

Prenasal fricatives but not stops. What's the justification here?

[Shemtov]

No reason in particular, it just feels right. I'm also thinking of adding /m: n: ɲ: ŋ: l: r:/ <mm nn nny nng ll rr>

[thethief3]

I do doubt its realistic since prenasal fricatives are so rare compared to stops.

[Shemtov]

Redo:
/p b ⁿb t d ⁿd c ɟ ⁿɟ k ⁿg k͡p g͡b ʔ/ <p b mb t d nd c j nj k g ng kp gb ʔ>
/m m: n n: ɲ ɲ: ŋ ŋ:/ <m mm nn ñ ññ ŋ ŋŋ>
/ɸ s ç h/ <f s ś h>
/ɾ r:/ <r rr>
/l l:/ <l ll>
/w j/ <w y>

/i u ɐ a/ <i u ă a>
/i: u: a:/ <í ú á>
/ai au ɐi ɐu/ <ai au ăi ău>
/iɐ uɐ/ <iă uă>

Phonotactics: CV(C)
Permitted finals: /p b t d k g m n ŋ s l/
/k͡p g͡b/ can only occur root-initial.

I'm thinking of a divide into two major branches: One where plain voice stops devoice, and the unvoiced stops merge with the fricatives, leaving to a stop series like /p ⁿb t ⁿd c ⁿɟ k ⁿg k͡p ʔ/, and one where unvoiced stops become voiced intervocalically, plain voiced, to fricatives, and prenasalized to nasalized vowel+fricative, plus epithetic vowels allowing for old stops to keep their values despite being intervocalic in the daughter.lang

[Creyeditor]

You have heard of extensive vowel syncope but have you heard of extensive consonant lenition elision (ECLE)?
Imagine a CV language and every other consonant is deleted. You could link this to stress. If every other vowel has (primary and then secondary) stress, counting from the right, then you could say a consonant is deleted if it precedes an unstressed vowel. Here are some examples of ECLE:

/'ka.ta/ -> [ka:] 'bus'
/kata/-/na/ -> /ka.'ta.na/ -> [ka.ta:] 'bus-DEF'
/kata/-/na/-/di/ -> /,ka.ta.'na.di/ -> [ka:.nai] 'bus-DEF-LOC'
/kata/-REDUPLICATION -> /,ka.ta.'ka.ta/ -> [ka:.ka:] 'bus-PL

As a productive system that wouldn't probably last very long. But as a morphophonological process that is very interesting. Maybe I'll cope that in a small extend.

You are probably right. It probably boils down to many morphemes having two metrically-conditioned allomorphs. And feel free to use the idea, of course

[sangi39]

You have heard of extensive vowel syncope but have you heard of extensive consonant lenition elision (ECLE)?
Imagine a CV language and every other consonant is deleted. You could link this to stress. If every other vowel has (primary and then secondary) stress, counting from the right, then you could say a consonant is deleted if it precedes an unstressed vowel. Here are some examples of ECLE:

/'ka.ta/ -> [ka:] 'bus'
/kata/-/na/ -> /ka.'ta.na/ -> [ka.ta:] 'bus-DEF'
/kata/-/na/-/di/ -> /,ka.ta.'na.di/ -> [ka:.nai] 'bus-DEF-LOC'
/kata/-REDUPLICATION -> /,ka.ta.'ka.ta/ -> [ka:.ka:] 'bus-PL

As a productive system that wouldn't probably last very long. But as a morphophonological process that is very interesting. Maybe I'll cope that in a small extend.

You are probably right. It probably boils down to many morphemes having two metrically-conditioned allomorphs. And feel free to use the idea, of course

I mean, liaison in French has been around, I think, for around 300 years, and I don't know how fragile the French system is at the moment, so something like this system might last for at least a little bit. I'm sure there's an Austronesian language with a roughly similar (tried finding it while I was at work, but no lucky)

I'd imagine what might happen, is that the grammatical endings might go from agglutinative to synthetic, e.g. -/'na.di/ > -/'nai/ vs. -/na/ > -/a/ and -/di/ > -/i/ which are then attached to the root, whose form is dependent on whether it's stressed or unstressed, which feels at least somewhat more "stable"

[Arayaz]

/t tˤ k kʷ/ t tx k kw (tx instead of t with a hacek because it renders better)
/n nˤ/ n ň
/s̺ sˤ/ s š
/r j w/ r y w

/i y u e ø o a/ + high tone
i ü u e ö o a í ű ú é ő á

[Arayaz]

Also:

/t k/ t k
/ⁿb ⁿd/ mb nd
/ɓ ɗ/ b d
/m n/ m n
/w/ w

/ɨ ~ u ~ ɯ/ i
/ɛ ~ ɵ ~ ə/ e
/ɒ ~ ɑ ~ ʌ/ o

All vowels can be long.
The first listed vowel phoneme is the default realization. The vowels can be backed after /k/ or /w/, with or without rounding depending on the speaker. The low vowel is rounded by default, and unrounds in various contexts, again depending on the speaker/dialect.

[Shemtov]

I'm thinking of putting this language, Gǂàamɱ́, in the World of Fuhe. I'm still working on the morphosyntax, so I'm posting the phonology here:
/p b ɓ t tʰ d ⁿd ɗ t͡s t͡sʰ d͡z ⁿd͡z t͡s' t͡ʃ t͡ʃʰ d͡ʒ t͡ʃ' ⁿd͡ʒ t͡ɬ t͡ɬʰ t͡ɬ' k k͡x g k' kʷ kʷ' q q͡χ ʔ ʔʲ ʔʷ/ <p b b' t th d nd d' c ch ż nż c' č čh ž č' tl tlh tl' k kh g k' kw kwh q qh ʔ ʔy ʔw>
/m n mʱ m̥ʰ n nʱ n̥ʰ nʲ nʲʱ ŋ ŋʱ/ <m mh hm n nh hn ny nyh ng ngh>
/f θ ð s z ɬ ʃ ʒ ç ʝ χ ʁ χʷ ħ ʕ h ɦ/ <f þ ð s z hl sh zh ś ź x gh xw ħ 3 h ɦ>
/r rʱ r̥ʰ rʲ rʲʱ/ <r rh hr ry ryh>
/l lʱ lʲ lʲʱ/ <l lh ly lyh>
/j w/ <y w>

/ǀ ᶢǀ ǀʰ ᶢǀʱ ⁿǀ ⁿǀʱ |'/ < ǀ gǀ hǀ ǀh nǀ nǀh ǀ'>
/ǃ ᶢǃ ǃʰ ᶢǃʱ ⁿǃ ⁿǃʱ ǃ'/ < ! g! h! !h n! n!h !'>
/ǁ ᶢǁ ǁʰ ᶢǁʱ ⁿǁ ⁿǁʱ ǁ'/ ǁ gǁ hǁ ǁh nǁ nǁh ǁ'
/ǂ ᶢǂ ǂʰ ⁿǂ ǂ' / <ǂ gǂ hǂ nǂ ǂ'>

/i e o a ɨ̃ ə̃/< i e o a ę ą>
/i: e: o: a:/ <i: e: o: a:>
/ai oi ei ae ao eo aɨ̃ õɨ̃ ẽɨ̃ ə̃ĩ ə̃õ ə̃ẽ/ <ai oi ei ae ao eo aę oę eę ąi ąo ąe
/m̩ ŋ̩/<ɱ ŋ>

/˩ ˧ ˥ ˩˧ ˧˥ ˥˧ ˧˩/ (shown on /a:/ <aa>) <àà aa áá àa aá áa aà>
Contour tones only occur on long vowels and diphthong.

Syllable structure: CV

/o/ is realized as [ u ] if the nearest mora not blocked by "Strong Obstruents" (all obstruents except <ʔy ʔw ś ź ħ 3 h ɦ>) has as <i> or <ę> as the vowel, except in the same diphthong and after <kw kwh ʔw f xw>.
/jo/ is realized as [jʉ]
/wi we/ are realized as /wy wø/
/a/ is realized as [ɛ]/ after <č čh ž č ʔy ny nyh sh zh ś ź ry ryh ly lyh y.>
/a/ is realized as [ɔ]/ after <m mh hm p b b ' kw kwh ʔw f xw>

[Vlürch]

Just something random, not sure if I'll do anything more fleshed-out with this but I might chuck it somewhere in one of the two con-archipelagos I've been focusing all of my recent conlanging on.

/m n̪ ɳ~ŋ/ <m n ñ>
/p b t̪ d̪ ʈ~k ɖ~g (ʔ)/ <p b t d k g Ø>
/s̪ ʂ~x/ <s h>
/f v θ ɻ~ɰ/ <f v z r>
/ɬ/ <l>

/a ɘ i ɔ u/ <a e i o u>
/aː ɘː iː ɔː uː/ <á é í ó ú>

/a/ is [a̠]
/ɘ/ is [ɨ] after /n̪ t̪ d̪ s̪ θ/ and word-initially (ie. after /ʔ/ in the first syllable), [ə] elsewhere
/i/ is [ɪ] after /f v θ ɬ/, [i​] elsewhere
/ɔ/ is [ʊ̞] after /f v θ/, [ɔ] elsewhere
/u/ is a [ʋ̩ʷ] after /f v θ/, [ɯ̹] elsewhere

/ɳ~ŋ ʈ~k ɖ~g ʂ~x ɻ~ɰ/ are:
[ɳ ʈ ɖ ʂ ɻ] before /a/
[ɳˠ ʈˠ ɖˠ ʂˠ ɻˠ] before /ɘ/
[ɲ t͡ɕ d͡ʑ ɕ j] before /i/
[ɳʶ q ɖʶ χ ʁ] before /ɔ/
[ŋ k g x ɰ] before /u/

Strict CV syllable structure, including in potential loanwords.

Stress is on the first syllable with a long vowel (in words with only short vowels, the first syllable). There's a pitch accent where a stressed syllable has a rising tone, the syllables before it have a low tone, the syllable directly after it has a falling tone while later syllables have a low tone, except final syllables in polysyllabic words where neither the last nor penultimate syllable are stressed have secondary stress with a high tone.

Probably agglutinative with a VSO word order where verbs have prefixes but nouns have suffixes, adjectives follow nouns and don't inflect at all, while in compounds the head noun comes last. It'd mostly have prepositions, only a few postpositions. Just for the sake of making it weirder, it'd have allocutive agreement based on both gender and moiety with the speakers having facial tattoos to mark which moiety they belong to. It'd be spoken on a small island where the population is divided between two moieties that mostly interact with each other for trade (one group could be hunter-gatherers and the other agriculturalists) and intermarriage once every however many years while otherwise being endogamous. I guess there'd be grammatical gender based on some kind of animist belief system that divides everything in existence into male and female energies or whatever. There'd be two dialects with some subtle phonological and lexical differences, and grammatically they might group some things into different genders too. Also, gendered and moietied(?) pronouns, the latter of which would be different between the dialects when referring to members of the other moiety.

Some gibberish example words:
ñuú [ŋɯ̹̀ˈʔɯ̹̌ː]
hikañe [ˈɕǐʈâ̠ɳˠə̀]
harógagi [ʂà̠ˈʁɔ̌ːɖâ̠ˌd͡ʑí]
zúrimávulu [ˈθʋ̩̌ʷːjîmà̠ːvʋ̩̀ʷˌɬɯ̹́]
fohoikóba [fʊ̞̀χɔ̀ʔìˈqɔ̌ːbâ̠]
venaséñi [və̀n̪à̠ˈs̪ɨ̌ːɲî]

[Arayaz]

Trying a four-feature consonant inventory:

+/- voiced
+/- dorsal
+/- labial
+/- continuant

Code: Select all

         -D, -L   -D, +L   +D, -L   +D, +L
-C, -V     t        p        k        kʷ
-C, +V     d        b        g        gʷ
+C, -V     s        f        x        xʷ
+C, +V     n        m        j        w

[DesEsseintes]

Latest toy thing. I apologise for the lazily spaced table.

Code: Select all

   m              n       ɲ       ŋ
pʰ p  b        tʰ t d          kʰ k g
   ɸ      θ ð    s z      ʃ       ɣ       h
   w             l̥ l      j

The vowels - untabulated (I’ve no standards these days) - are:

a ɛ ei i
ɔ ou u
ʌ ɤi ɯ

Syllable structure is (C)V-ish, but syllables can be nasalised. glottalised or post-aspirated (read: there are codas but I won’t admit it.)

I’m calling it Eidho for now.

[Creyeditor]

The consonant inventory looks like the notes to a jingle

[Porphyrogenitos]

Inspirations: Tlingit, Pali, Bengali

/t c k kʷ/ t c k kw
/tʰ cʰ kʰ kʷʰ/ th ch kh kwh
/d ɟ g gʷ/ d j g gw
/ɾ j w/ r y w
/ɾ̥ j̊ w̥/ rh yh wh
/h/ h

/a e i o u/ a e i o u
/aː eː iː oː uː/ aa ee ii oo uu
/ai̯ au̯ ei̯ eu̯ oi̯ ou̯/ ai au ei eu oi ou

Syllable structure is (C)V(C); word-finally coda C may only be h; word-medially it may only be geminate lengthening of the following consonant. h cannot be geminated. Long vowels and diphthongs may only occur in open syllables.

Geminate consonants are represented orthographically by doubling of the first letter in a digraph; e.g. tth, ggw, kkwh, rrh.

A long vowel in a syllable that becomes closed due to morphology/phonology becomes short; diphthongs shorten in this manner:

ai au ei eu oi ou → e o i u i u

When morphology/phonology produces a cluster with h + a following consonant, the h assimilates completely to the following consonant and causes it to become aspirated/voiceless. Other would-be clusters are variously resolved through complete assimilation/gemination, deletion, or epenthesis.

Diphthongs do contrast with y and w; for instance, khaio is longer than kayo. Contrasts like raua vs. rauwa are possible, but marginal.

Foreign nasals are adapted in loanwords as voiced stops, sometimes as r, y, or w. Foreign labials are adapted in loanwords as labiovelars. Foreign sibilants are usually adapted as palatal stops; foreign non-sibilant fricatives may be adapted as h.

I would say that geminates cannot be explained by underlying h even though the two are technically in complementary distribution, since unaspirated and voiced geminates like kk, dd, and yy can never be formed by assimilation of h.

[WeepingElf]

I'd expect the labiovelars to evolve into labials in such a system.

[DesEsseintes]

Inspirations: Tlingit, Pali, Bengali

/t c k kʷ/ t c k kw
/tʰ cʰ kʰ kʷʰ/ th ch kh kwh
/d ɟ g gʷ/ d j g gw
/ɾ j w/ r y w
/ɾ̥ j̊ w̥/ rh yh wh
/h/ h

/a e i o u/ a e i o u
/aː eː iː oː uː/ aa ee ii oo uu
/ai̯ au̯ ei̯ eu̯ oi̯ ou̯/ ai au ei eu oi ou

Lovely. What do you plan to do with ìt morphologically?

I, incidentally, harbour no expectations towards your labiovelars.

[Porphyrogenitos]

I'd expect the labiovelars to evolve into labials in such a system.

Maybe! But I was taking inspiration from the North American languages that stably lacked labials.

Lovely. What do you plan to do with ìt morphologically?

I, incidentally, harbour no expectations towards your labiovelars.

Oh, I'm not terribly sure. I might just take a cue from Pali and Bengali again and make it suffixing-fusional with some clitics.

There could be alternations triggered by assimilation of /h/ and other consonants, e.g.

khar- 'go'
khar- + -wa → kharuwa 'to go'
khar- + ta → khatta 'I go, am going'

jah - 'eat'
jah- + -wa → jawwha 'to eat'
jah - + -ta → jattha 'I eat, am eating'

There will definitely be vowel-shortening alternations:

gai '1s.ᴘᴀᴛ'
gairu '1s.ᴀɢᴛ'
geh '1s.ᴅᴀᴛ'
gekkhe '1s.ʟᴏᴄ'

Possibly rhythmic gradation of the voiced stops, assuming a Latin-like stress rule:

thora 'knife' + -ji 'ᴅɪᴍɪɴ' → thoraji 'paring knife' (stress on penult)
aagwe 'bird' + -ji 'ᴅɪᴍɪɴ' → aagweyi 'birdie' (stress on first syllable)

[Omzinesý]

p t tˠ k
t͡s t͡sˠ t͡ɕ
s sˠ ɕ
v z ð ɹˠ
m n nˠ
l lˠ j

i u
ə
a


A Hebrew-style morphology
Stress on the final syllable but a single ɘ is unstressed.