Random phonology/phonemic inventory thread

[DesEsseintes]

Having grown a bit tired of Núta after the relay*, I’ve been toying with TLFKAT over the last week or so. I have many ideas about the morphology, but in the beginning I was convinced that I was satisfied with the phonology, which hadn’t changed since mid-2016.
*I still fully plan on returning to it.

How wrong I was thinking I could revisit a lang without a Phonology overhaul!

So here it is: TLFKAT 4.1 - Epiglottal Epiphany

Code: Select all

p    p͡m   p͡ʂ   p͡ʟ̝̊   m        w
p’   p͡m’  p͡ʂ’  p͡ʟ̝̊’  
t    t͡n   t͡s   t͡ɬ   n     s  l  ɬ
t’   t͡n’  t͡s’  t͡ɬ’
tʷ   t͡nʷ  t͡sʷ  t͡ɬʷ  nʷ    sʷ lʷ ɬʷ
tʷ’  t͡nʷ’ t͡sʷ’ t͡ɬʷ’
          t͡ʃ              ʃ  j
          t͡ʃ’            
k    k͡ŋ   k͡ʂ   k͡ʟ̝̊   ŋ  x  ʂ  ʟ  ʟ̝̊
k’   k͡ŋ’  k͡ʂ’  k͡ʟ̝̊’           ɰ
kʷ   k͡ŋʷ  k͡ʂʷ  k͡ʟ̝̊ʷ  ŋʷ xʷ ʂʷ ʟʷ ʟ̝̊ʷ
kʷ’  k͡ŋʷ’ k͡ʂʷ’ k͡ʟ̝̊ʷ’
q    q͡ɴ        q͡ʟ̝̠̊   ɴ  X        ʟ̝̠̊
q’   q͡ɴ’       q͡ʟ̝̠̊’
qʷ   q͡ɴʷ       q͡ʟ̝̠̊ʷ  ɴʷ Xʷ       ʟ̝̠̊ʷ
qʷ’  q͡ɴʷ’      q͡ʟ̝̠̊ʷ’
ʔ                      ħ     ʕ   
ʔʷ                     ħʷ    ʕʷ  

/ɰ/ is in the wrong place in the table because I forgot to include it and had to slot it in somewhere. (In fact, I should probably place /w/ right beneath /ʟʷ/ shouldn’t I?)

The vowels are a height-based /a ə ɨ/ trio and realisations are determined by surrounding consonants. I won’t go into the details for now.

So what has changed?

1. Aspiration has been eliminated as a phonemic contrast. The main reason for this was the romanisation. I was using h which resulted in rather unwieldy oddities like qnhw. Using b d g for unaspirated stops didn’t feel satisfying either, and made the language look like Navajo, thus ruining the TLFKAT aesthetic.
2. Labialised coronals are now in again.
3. I’ve reintroduced voiced velar lateral approximants. Oddly, this means that TLFKAT distinguishes between lateral and plain approximants at the velar PoA only, which I think is interesting. I intend to make heavy use of the velar laterals. Note that I didn’t add lateral approximants at the uvular PoA (see below)
4. Epiglottals! (Or are they pharyngeals? I can’t say I know the difference.) I speak rudimentary Arabic so for now I’ve just added the epiglottals from Arabic along with labialised variants. Now the interesting thing here is that I can have the voiced epiglottal fricatives function as the voiced counterparts to the uvular lateral fricatives since there are no uvular lateral approximants. I think this is neat.

Des Pondérables:

• I haven’t included /h/ and I don’t think I will.
• I like the fact that the retroflex affricates have velar stop components, but I wonder of /q͡ʂ/ might not be nicer? Or even distinguish all of /t͡s k͡s̠ q͡ʂ/?
• I am tempted to expand the epiglottals a bit, by adding epiglottal stops perhaps? I’m not sure though as I’m not very confident producing such sounds.

Comments, questions, suggestions?

[Parlox]

Comments, questions, suggestions?

I have to say i like this inventory a lot! And i really like /ʟ/, and all the other laterals, especially /ʟ̝̠̊ʷ/. I can't say anything about how realistic it is(I wouldn't know).

[DesEsseintes]

Comments, questions, suggestions?

I have to say i like this inventory a lot! And i really like /ʟ/, and all the other laterals, especially /ʟ̝̠̊ʷ/. I can't say anything about how realistic it is(I wouldn't know).

Thank you! Yes, the profusion of laterals is meant to be the most salient feature of the phonology. You are also right that I’m definitely stretching it a bit when it comes to naturalism.

[Parlox]

Thank you! Yes, the profusion of laterals is meant to be the most salient feature of the phonology. You are also right that I’m definitely stretching it a bit when it comes to naturalism.

Some languages are crazier, doesn't ǃXóõ have over 150 consonants?

Also, i love the aesthetics of the languages you make. I'm a sucker for any north-american inspired conlang.

[shimobaatar]

Thanks for taking a look. It's really helpful to get an outside opinion on this one. As for /k͡xʰ/, I was mostly thinking of that glottal realisation because I seem to pronounce it that way, but now I think about it, I guess it's because my dialect of English doesn't really, you know, do an intervocalic /h/. [k͡xəh] really does make more sense.

I'm going for (C)(J)V(C), where J is any glide. Onsets can be any any consonant but /ŋ/, and codas can be any nasal or stop. I'm still working out which glide-vowel combinations are allowable, but it's particularly limiting for back vowels.

No problem at all! I look forward to seeing where this will go.

---

1. Epiglottals! (Or are they pharyngeals? I can’t say I know the difference.) I speak rudimentary Arabic so for now I’ve just added the epiglottals from Arabic along with labialised variants. Now the interesting thing here is that I can have the voiced epiglottal fricatives function as the voiced counterparts to the uvular lateral fricatives since there are no uvular lateral approximants. I think this is neat.

I typically see /ħ ʕ/ called "pharyngeals".

And I agree. That is neat.

Des Pondérables:

• I haven’t included /h/ and I don’t think I will.
• I like the fact that the retroflex affricates have velar stop components, but I wonder of /q͡ʂ/ might not be nicer? Or even distinguish all of /t͡s k͡s̠ q͡ʂ/?
• I am tempted to expand the epiglottals a bit, by adding epiglottal stops perhaps? I’m not sure though as I’m not very confident producing such sounds.

• I don't see why you would need to, especially now that the aspirated consonants are gone.
• I'd personally go for distinguishing all three.
• If it's important to you to be able to pronounce the language, then I'd say probably don't add the epiglottal stop.

[Porphyrogenitos]

A vowel harmony idea I just came up with. Plausible? Not sure. It seems intuitive.

harmony class I: /i ə/
harmony class II: /ɨ o/
neutral: /a/

[DesEsseintes]

A vowel harmony idea I just came up with. Plausible? Not sure. It seems intuitive.

harmony class I: /i ə/
harmony class II: /ɨ o/
neutral: /a/

Start off with /a e i o u/ and harmony between /e i/ and /o u/, then e u → ə ɨ.

And yes, it feels good.

[Porphyrogenitos]

Start off with /a e i o u/ and harmony between /e i/ and /o u/, then e u → ə ɨ.

And yes, it feels good.

Thanks - and that makes sense, though is e > ə so common? I guess something sort of similar happens with the NCVS.

[gestaltist]

A vowel harmony idea I just came up with. Plausible? Not sure. It seems intuitive.

harmony class I: /i ə/
harmony class II: /ɨ o/
neutral: /a/

Start off with /a e i o u/ and harmony between /e i/ and /o u/, then e u → ə ɨ.

And yes, it feels good.

Isn't it just a vertical 3-vowel system with front-back harmony? I'd posit /a ə i/ and then harmonized backing to [o ɨ]. In any case, looks and feels reasonable.

[Vlürch]

Something inspired by Austronesian languages, Japanese and Arabic.

/m n/ <m n>
/b k q ʔ/ <b k q c>
/ʔʷ/ <p>
/t͡s t͡ʃ/ <z j>
/s/ <s>
/ɸ j w ʕ h/ <f y w ɔ h>
/r~l/ <r~l>
/ɬ/ <x>

/a i u/ <a i u>
/aː iː uː/ <ā ī ū>

Allophony:
/ma mi mu/ [mæ mɪ mʊ] <ma mi mu>
/maː miː muː/ [mɒː miː muː] <mā mī mū>
/na ni nu/ [nɒ ɲɪ nʊ] <na ni nu>
/naː niː nuː/ [nɑː ɲiː nuː] <nā nī nū>
/ba bi bu/ [pɑ pɪ pʊ] <ba bi bu>
/baː biː buː/ [bɑː biː buː] <bā bī bū>
/ka ki ku/ [kæ cɪ kʊ] <ka ki ku>
/kaː kiː kuː/ [kʰɑː c͡çiː kʰuː] <kā kī kū>
/qa qi qu/ [qɑ qəɪ̯ qəʊ̯] <qa qi qu>
/qaː qiː quː/ [q͡χɒː q͡χə̯i q͡χə̯u] <qā qī qū>
/ʔa ʔi ʔu/ [ʔʌ ʔe ʔo] <ca ci cu>
/ʔaː ʔiː ʔuː/ [ʔɑː ʔiː ʔuː] <cā cī cū>
/ʔʷa ʔʷi ʔʷu/ [ʔʷə ʔʷəɪ̯ ʔʷoʊ̯] <pa pi pu>
/ʔʷaː ʔʷiː ʔʷuː/ [ʔʷɒː ʔʷyː ʔʷuː] <pā pī pū>
/t͡sa t͡si t͡su/ [t͡sæ t͡sɨ t͡sʊ] <za zi zu>
/t͡saː t͡siː t͡suː/ [t͡sɑː t͡sə̯i t͡suː] <zā zī zū>
/t͡ʃa t͡ʃi t͡ʃu/ [t͡ʃə t͡ɕɪ t͡ʂʊ] <ja ji ju>
/t͡ʃaː t͡ʃiː t͡ʃuː/ [t͡ʃæː t͡ɕiː t͡ʂuː] <jā jī jū>
/sa si su/ [ʃə ɕɪ ʂʊ] <sa si su>
/saː siː suː/ [sæː siː suː] <sā sī sū>
/ɸa ɸi ɸu/ [ɸʷə ɸʲɪ fʊ] <fa fi fu>
/ɸaː ɸiː ɸuː/ [fæː fʲiː ɸʷuː] <fā fī fū>
/ja ji ju/ [jæ jɪ jʉ] <ya yi yu>
/jaː jiː juː/ [jæː jiː jyː] <yā yī yū>
/wa wi wu/ [wæ ɥɪ wʊ] <wa wi wu>
/waː wiː wuː/ [wɒː ɥiː wuː] <wā wī wū>
/ʕa ʕi ʕu/ [ʕə ʕɨ ʕʊ] <ɔa ɔi ɔu>
/ʕaː ʕiː ʕuː/ [ʕæː ʕiː ʕʉː] <ɔā ɔī ɔū>
/ha hi hu/ [ħæ çɪ xʊ] <ha hi hu>
/haː hiː huː/ [xɑː çiː χoː] <hā hī hū>
/ɾa ɾi ɾu/ [ɾæ ɾʲɪ lʊ] <ra ri lu>
/ɾaː ɾiː ɾuː/ [læː ɾʲiː ɾuː] <lā rī rū>
/ɬa ɬi ɬu/ [ɬə ʎ̝̊ɪ ʟ̝̊ʊ] <xa xi xu>
/ɬaː ɬiː ɬuː/ [ʟ̝̊ɑː ʎ̝̊iː ɬʉː] <xā xī xū>

Word-initially, vowels have allophones depending on the vowel in the following syllable:
Back vowel: /a i u/ [ɑ e u], /aː iː uː/ [ɑː iː uː]
Front vowel: /a i u/ [æ i o], /aː iː uː/ [æː iː yː]

Syllables are (C)V(N).
C = any consonant
V = any vowel
N = any nasal

Word-final /n/ has the allophones [n ŋ ɲ] after central, back and front vowels respectively.

Short vowels that are not diphthongs have voiceless allophones between two voiceless consonants.

Random meaningless words:
ixāna [eʟ̝̊ɑːnɒ]
curīɔū [ʔoɾʲiːʕʉː]
hayāfu [ħæjæːfʊ]
urakīwa [oɾæc͡çiːwæ]
ilājim [ilæːt͡ɕɪm]
ainaxapi [æinɒɬə̥ʔʷəɪ̯]
lāqākin [læːq͡χɒːcɪɲ]
ɔajūbiji [ʕə̥t͡ʂuːpɪ̥t͡ɕɪ]
zaranān [t͡sæɾænɑːŋ]
yāfan [jæːɸʷən]
sikulām [ɕɪ̥kʊlæːm]

[Frislander]

/p t t͡ʃ k ʔ/
/β ð ɣ/
/m ɾ/

/i ə/
/e a/

There is a highly morphologised process of vowel reduction, whereby /e/ > /i/ and /a/ > /ə/ in unstressed syllables.

Syllable structure is CV(ʔ, m, ɾ). The nasal /m/ assimilates to the POA of a following oral consonant, and furthermore the cluster /mɾ/ undergoes double assimilation to [nn]. Additionally the VCV sequences /aβə/, /əβa/ > [oː] and /əβə/ > [uː] respectively, while the intervocalic sequences with /ɣ/ are resolved as follows:

Code: Select all

  a   ə   e   i
a -   aː  -   aɣi
ə aː  əː  əɣe əɣi
e -   ejə -   eː
i ija ijə eː  iː

Stress is mobile and can appear in different places in the word depending on the particular conjugation, with corresponding vowel reduction (see above).

[DV82LECM]

Eathut /i'atʔut/ ['jat:on] (formally Iachor):

/p~ɸ~β~m~b t~s~l~n~d~ɾ k~x~ɣ~ŋ~g ʔ~h/ <p t c h>
/i~e~j u~o~w a/ <e u a>

I've ALWAYS wanted to go minimalist somehow and showcased Iachor, a language that was inspired by an attempt to rip-off the Jetsons' for inspiration ("Eep Opp Ork Ah Ah"). Anyway, I ran into a dead-end for the last attempt and picked it up 2 days ago and I adore what I have. There is a shit-ton of allophony and it ALL works. The best part is, in the new formula, my original idea for the phrase "I love you" still works.

Epá put caha.
[i'pa poɾ 'kaʔa]
1p. 2p. "love"

[Porphyrogenitos]

Many conlangers who try to include clicks in their conlangs end up including a series of plain/tenuis click that looks like this:

/ʘ ǀ ǂ ǃ ǁ/

There's nothing inherently wrong with this, but in terms of naturalism, every language that features clicks makes use (sometimes exceedingly generous use) of the contrasts available in the velar coarticulation inherent to clicks. I've occasionally been interested in using clicks, and although I am at least in principle capable of articulating a fair number of click coarticulations, it's nonetheless rather difficult for me, and I'd prefer to include a smaller click inventory. But for the sake of naturalism, I'd at least try to include some coarticulation distinctions.

Anyways, all of that is a preface to this tiny click inventory I might use sometime. (And yes, the bilabial click is actually the rarest of the clicks, but I like it and it's easy to articulate.)

/k͡ʘ k͡ǃ/
/ŋ͡ʘ ŋ͡ǃ/

(I thought it would be weird to only have one non-bilabial click PoA, but apparently Sesotho and a couple other Bantu languages have just one click PoA that varies between alveolar, dental, and lateral, so I guess it's not that strange.)

[Frislander]

Here's two sister languages.

Language 1

/p t k ʔ/
/b d/
/t͡s’ t͡ɬ’ q͡χ’/
/s ɬ h/
/m n/
/w j/

/i iː u uː/
/e eː ə ɤː o oː/
/aː/
/ie uo/
/ea oa/

The only vowel which occurs in unstressed syllables is /ə/ and this is also the only vowel which may be found unstressed. Syllable structure is CV(F), where F can only be one of /ɬ h n/, of which /h/ is restricted to this position.

---

Language 2

/t k kʷ ʔ/
/t͡s/
/f s x xʷ h/
/m n/
/w ɾ j/

/i iː u uː/
/a aː/

The length distinction between vowels only occurs in the stressed syllable. Syllable structure is (C)V(F), where F is one of /n ɾ/, or alternatively any consonant may occur geminated intervocalically.

---

Cognate sets

[ˈɬo̯abə] : [ˈɾawa]
[ˈt͡siːt͡ɬ’ə] : [ˈxiːti]
[ˈpohq͡χ’ə] : [ˈfukʷːa]
[ˈtɤːɬt͡s’ə] : [ˈhaːɾt͡su]
[ˈt͡s’ɪ̯eʔən] : [ˈkaʔun]
[ˈkedə] : [ˈxʷi.a]
[ˈwuːjə] : [ˈwuːju]
[ˈmeːhsəɬ] : [ˈmaːxːiɾ]
[ˈbe̯anə] : [ˈwana]
[ˈq͡χ’iɬdə] : [ˈkʷiɾi]

[DesEsseintes]

Language 2

/t k kʷ ʔ/
/t͡s/
/f s x xʷ h/
/m n/
/w ɾ j/

/i iː u uː/
/a aː/

The length distinction between vowels only occurs in the stressed syllable. Syllable structure is (C)V(F), where F is one of /h n ɾ/, or alternatively any consonant may occur geminated intervocalically.

Lovely.

[DV82LECM]

Here's two sister languages.

Language 1

/p t k ʔ/
/b d/
/t͡s’ t͡ɬ’ q͡χ’/
/s ɬ h/
/m n/
/w j/

/i iː u uː/
/e eː ə ɤː o oː/
/aː/
/ie uo/
/ea oa/

The only vowel which occurs in unstressed syllables is /ə/ and this is also the only vowel which may be found unstressed. Syllable structure is CV(F), where F can only be one of /ɬ h n/, of which /h/ is restricted to this position.

---

Language 2

/t k kʷ ʔ/
/t͡s/
/f s x xʷ h/
/m n/
/w ɾ j/

/i iː u uː/
/a aː/

The length distinction between vowels only occurs in the stressed syllable. Syllable structure is (C)V(F), where F is one of /h n ɾ/, or alternatively any consonant may occur geminated intervocalically.

---

Cognate sets

[ˈɬo̯abə] : [ˈɾawa]
[ˈt͡siːt͡ɬ’ə] : [ˈxiːti]
[ˈpohq͡χ’ə] : [ˈfukʷːa]
[ˈtɤːɬt͡s’ə] : [ˈhaːrt͡su]
[ˈt͡s’ɪ̯eʔən] : [ˈkaʔun]
[ˈkedə] : [ˈxʷi.a]
[ˈwuːjə] : [ˈwuːju]
[ˈmeːhsəɬ] : [ˈmaːxːiɬ]
[ˈbe̯anə] : [ˈwana]
[ˈq͡χ’iɬdə] : [ˈkʷiɾi]

I adore these. Would love to see the progenitor and soumd changes. Got one small issue with your cognate sets, though: ['meːhsəɬ] : [ˈmaːxːiɬ]

Where is [ɬ] in the SECOND phonology? Was it supposed to be [ˈmaːxːiɾ]?

[CarsonDaConlanger]

Consonants
Stop:/p t t' k k'/ p t t' k k'
Nasal:/m n ŋ/ m n ŋ
Fricative:/s x/ s h
Approximant:/β ɹʷ j/ v w y
Tap:/ɾ/ r

Vowels
/i u/i u
/e o/e o
/a/a

(C)V(V)(m,n,ŋ)

Fricatives are voiced intervocally.
x>ç/_E
All vowels are pure.

[Frislander]

I adore these. Would love to see the progenitor and soumd changes.

Well, if you say so...


*p *t *ts~s *k *kʷ *ʔ
*t’ *ts’ *k’ *kʷ’
*b *d
*m *n
*l *j *w

*i *u
*ē *ea *ō *oa
*a~ā

Stress is word-initial, and in unstressed syllables only *i, *u, *a occurred, and of those only *i, *u occurred in stressed syllables as well.

Syllable structure was CV(F), where F was restricted to *ʔ, *n and *l.

Sound changes

To Daughter 1

• Plain velars are palatalised before front vowels
• Stressed *i *u were lowered to /e o/ respectively before *a in the second syllable
• Stressed *ē *ea *ō *oa were raised to /iː ie uː uo/ before *i or *u in the second syllable
• *ā was raised to /eː/ before *i in the second syllable and /ɤː/ before *u in the second syllable
• *i, *a and *u were reduced to /ə/ in an unstressed syllable
• * was strengthened to /ɬ/
• *t’ **k’ *kʷ’ were affricated to /t͡ɬ’ q͡χ’ q͡χʷ’/
• Postalveolar sibilants merged with the alveolar sibilants.
• Plain sibilant affricate *ts was lenited to /s/
• Labialisation was lost
• Coda *ʔ was lenited to /h/

To Daughter 2

• Plain and voiced oral stops/affricate were spirantised to fricatives (i.e. *p *t *ts *k *kʷ *b *d became /f θ s x xʷ v ð/)
• Ejective stops/affricates were lenited to plain stops
• *l was rhoticised to a flap /ɾ/
• The diphthongs *ea, *oa merged as a short /a/
• *ʔ plus consonant became a geminate, and word-final *ʔ was lost
• *θ debuccalised to /h/, *v lenited further and merged with /w/, and *ð was deleted entirely

Got one small issue with your cognate sets, though: ['meːhsəɬ] : [ˈmaːxːiɬ]

Where is [ɬ] in the SECOND phonology? Was it supposed to be [ˈmaːxːiɾ]?

Yeah, that's an error.

[Porphyrogenitos]

A language with the following permissible syllables:

Code: Select all

mi mʷy mɨ mʷu me mʷø mə mʷo mæ mʷɒ
ni nʷy nɨ nʷu ne nʷø nə nʷo næ nʷɒ
pi pʷy pɨ pʷu pe pʷø pə pʷo pæ pʷɒ
ti tʷy tɨ tʷu te tʷø tə tʷo tæ tʷɒ
ki kʷy kɨ kʷu ke kʷø kə kʷo kæ kʷɒ
   βʷy    βʷu    βʷø    βʷo    βʷɒ
si sʷy sɨ sʷu se sʷø sə sʷo sæ sʷɒ
hi     hɨ     he     hə     hæ
ɾi ɾʷy ɾɨ ɾʷu ɾe ɾʷø ɾə ɾʷo ɾæ ɾʷɒ
ji     jɨ     je     jə     jæ

Edit: Maybe with nasalization, or a permissible final /N/, or tone or pitch-accent. Yes, rather Japonic, I know. There may or may not be some kind of harmony process.

[nunbit]

This phonology is pretty unusual. Is it too unnaturalistic?

/t k/
/m n ŋ/
/v θ s/
/ɾ j/

/i ɯ/
/o/
/æ/

(C)V(n, ɾ, θ, s)