Post your conlang's phonology

[Click]

My try at an Australian phoneme inventory. This is not a phonology, because I don't want to spend half a hour on it.
C:
/m n̪ ɳ ɲ ŋ/
/p t̪ ʈ c k/
/l̪ ɭ ʎ j w/
/r̪ ɽ/

V:
/i iː u uː/
/a aː/

I think that Chagen didn't look at phoneme inventories of Australian languages. He just read they don't have any fricatives and made a phoneme inventory without the fricatives.

[ol bofosh]

I've played around with Alahithian's phonology. A lot of the changes are just retroflexes.

/a e i o u/ <a e i o u>
/m ɳ ŋ/ <m n ñ>
/p b t d ʈ ɖ k g ʔ/ <p b t d th dh k g h>
/ɭ w j r/ <l w y r>
/f v s z ʂ ʐ x h/ <f v s z sh zh x h>
/ʈʂ ɖʐ/ <c j>

/ʔ/ only appears when <h> is word final.

Not happy about the effect of /ɭ/, think I might just change it to /l/ instead.

[Pinetree]

So, I messed with the phonology and orthography of my new and unnamed language to make it a little easier to use (and also a little more Finnic, which was my original intent for the phonology):

Phoneme Inventory



Allophony

- [q ɢ qː χ χː] are allophones of /k g kː x xː/ (in that order):
- Prior to or following a back vowel (/kok/ is pronounced [qoq])
- Prior to a fronting diphthong (/xua/ is pronounced [χua])
- Following a backing diphthong (/iug/ is pronounced [iuɢ])
- /ɾ/ is in free allophony with [r ɽ ɹ ɻ]
- [θ] is an allophone of /s/ following a vowel
- [θː] is an allophone of /sː/ following a vowel
- [ð] is an allophone of /z/ following a vowel
- [æ] is in free allophony with /a/

Phonotactics

- Syllable structure is (O)(S)V(V)(S)(O), where O denotes obstruents, and S denotes sonorants.
- Long consonants may not follow one another in a word. In the case of affixes, in which the affix and the word to which it attaches create two abutting long consonants, the second long consonant is lenited to its voiced short equivalent
- Front vowels and back vowels may not occur within the same word, excepting fronting and backing diphthongs, which may occur in either case.

Stress

- Stress is fixed at the penultimate position.

[Lyra]

WIP Proto-Eastern

/p pʰ pʔ b t tʰ tʔ d k kʰ kʔ g ʔ
s sʰ sˤ z h ɣ
l ɾ
m n/
<p ph p· b th t· d k kh k· g ·
s sh s· z h x
l r
m n>

/i u
e o
a/
<i u
e o
a>

/au, ai, ua, ia/
/awa, aja/
<au, ai, ua, ia>
<aua, aia>

Allophony:
ʔ_
/e/ → /e̞/
/a/ → /ä/
/o/ → /o̞/

glottals no release in coda

Phonetactics:
word construction is (C)V(C)
suffixes V(C)
preffixes (C)V
/ʔ/ separates vowels.

Stress on core syllable.

*I know that <'> for glottals is not the most beautiful or elegant way of doing it, but this is just for ease of writting... maybe I could use 'j'?

~Lyra

[Click]

suffixes V(C)
affixes (C)V

Lyra, suffixes are also affixes. Did you mean prefixes, maybe?

I know that <'> for glottals is not the most beautiful or elegant way of doing it, but this is just for ease of writting... maybe I could use 'j'?

Don't use ‹j›. I suggest to use ‹·›. It is quite nice IMO.

[Lyra]

suffixes V(C)
affixes (C)V

Lyra, suffixes are also affixes. Did you mean prefixes, maybe?

lol yes, xD
Keep correcting me, I'm horrid at nomenclature.

~Lyra

[Wattmann]

Something random:

/p t k q/
/s h/
/m n r/

/a i u/

Vowels are seperated by /h/, morpheme structure is strictly CV(C)CVC(C)...V(C), both /m/ and /n/ allophonically become |N| before plosives, /kq/ yields /qq/, /pq/ yields /pk/, /rh/ yields /s/, /qi/ yields /ki/, /ku/ yields /qu/, /hi/ yields /si/.
|N| realizes at the POA of the following consonant, geminates are allowed intervocalically.

[brandrinn]

OK, here's my dilemma.
I have a phonology, but it's kind of... flat and artificial. A real phonology is like a series of semi-transparent layers. You can see the sound changes and phonological processes going on. They look like a snapshot of a whirlpool, rather than a completed product. So I am asking for assistance from the ZBB. If anyone is willing, I would like you all to do one or both of the following:
a) Make sound changes from this phonology, extending into the future. I may incorporate them, fast-forwarding the phonology a few generations.
b) Make sound changes from the past that would lead to this phonology. This will help me think about the diachronics of the current system.
Couldn't I just do all this myself? Yes, I've been doing that. But I'm at that stage where you take all the commas out, then read again and put the commas back in. I'm in a rut, and I think getting some input from others would help immensely. I promise not to outright copy anyone's idea without their permission. Anyway, here it is:

Code: Select all

                              labial   coronal   dorsal
stiff voice plosive:           p_k      t_k        k_k
slack voice plosive:           p_t      t_t        k_t
stiff voice fricative:                  s_k        x_k
slack voice fricative:                  s_t
nasal:                          m        n
approximant:                    P        l          j

Code: Select all

                  front   central   back
high:               i       }        (u)
mid-high:                   I\
mid-low:            E                 O
low:                        a

Initial clusters of the pattern sC and kC are permissible (except ss and kk). Coda consonants are restricted to t, l, and a homorganic nasal N.

Most roots are two syllables, but the more I generate, even with derivational morphology, the more they just look made up. It doesn't feel like a living system. So please, if you would, play my little game and make up some anterior and posterior sound changes. I think it will help me get my mind out of the rut it's been in. Thank you.

[communistplot]

Latest version of the Proto-Yagalo-Atarean Phoneme Inventory:

Consonants
Nasal
/m mʲ n nʲ nʷ ŋ ŋʲ ŋʷ/ <m mj mw n nj nw ŋ ŋj ŋw>
Stop
/p pʲ t tʲ tʷ k kʲ kʷ ʔ/ <p pj t tj tw k kj kw q>
Fricative/Approximant
/s̞ s̞ʲ s̞ʷ ɬ ɬʲ ɬʷ x̞ x̞ʲ x̞ʷ h hʷ/ <r rj rw s sj sw x xj xw h hw>
Lateral
/l lʲ lʷ/ <l lj lw>

_________________________

Vowels
Front
/i y e ø/ <i ü e ö>
Centre
/ɨ ə a/ <ï ë a>
Back
/u o/ <u o>
There's also phonemic length contrast on vowels.

[Bristel]

Tugen Luliti (Lilithian)

This is supposed to be a tri-literal root language, based somewhat heavily on Hebrew. I'm currently struggling through creating "binyanim" (roots) for the language. The name is inspired from a fictional group in a comic book my friend is working on called the "Order of the Snake". I was inspired by the mythological story of Lilith, and Hebrew was the first linguistic inspiration for it, along with other Semitic languages.

Phonology

/p b t d k g ʔ/ <p b t d k g ˀ>
/m n ŋ/ <m n ng>
/f v θ ð s z ʃ ʒ x ɣ h/ <f v th dh s z š ž kh gh h>
/t͡s t͡ʃ d͡z d͡ʒ/ <ts tsh dž dž>
/r l j w/ <r l y w>

/a aː e eː i iː y yː o oː u uː/ <a ā e ē i ī y ȳ o ō u ū>

Allophony

n > ŋ /_k,g
p,t,k > f,θ,x /V_V
b,d,g > v,ð,ɣ /V_V
s,ʃ > z,ʒ /_C[+voice]
b,d,g > p,t,k /_C[-voice]
p,t,k > b,d,g /_C[+voice]
v > f /_C[-voice]

Phonotactics

CV(V)(C)C

(having trouble with phonotactics, mostly because I'm not sure what the phonotactics should be in a triliteral language)

[Wattmann]

OK, here's my dilemma.
I have a phonology, but it's kind of... flat and artificial. A real phonology is like a series of semi-transparent layers. You can see the sound changes and phonological processes going on. They look like a snapshot of a whirlpool, rather than a completed product. So I am asking for assistance from the ZBB. If anyone is willing, I would like you all to do one or both of the following:
a) Make sound changes from this phonology, extending into the future. I may incorporate them, fast-forwarding the phonology a few generations.
b) Make sound changes from the past that would lead to this phonology. This will help me think about the diachronics of the current system.
Couldn't I just do all this myself? Yes, I've been doing that. But I'm at that stage where you take all the commas out, then read again and put the commas back in. I'm in a rut, and I think getting some input from others would help immensely. I promise not to outright copy anyone's idea without their permission. Anyway, here it is:

Code: Select all

                              labial   coronal   dorsal
stiff voice plosive:           p_k      t_k        k_k
slack voice plosive:           p_t      t_t        k_t
stiff voice fricative:                  s_k        x_k
slack voice fricative:                  s_t
nasal:                          m        n
approximant:                    P        l          j

Code: Select all

                  front   central   back
high:               i       }        (u)
mid-high:                   I\
mid-low:            E                 O
low:                        a

Initial clusters of the pattern sC and kC are permissible (except ss and kk). Coda consonants are restricted to t, l, and a homorganic nasal N.

Most roots are two syllables, but the more I generate, even with derivational morphology, the more they just look made up. It doesn't feel like a living system. So please, if you would, play my little game and make up some anterior and posterior sound changes. I think it will help me get my mind out of the rut it's been in. Thank you.

I could help, if you provided some IPA ~_~
I'm not good in XYZ-Sampa

[Click]

I could help, if you provided some IPA ~_~
I'm not good in XYZ-Sampa

You can convert to the IPA using this.

[Wattmann]

Hmm.

Stiff voice might have arisen from glottalisation...

k͡xʼ > x' > x̰
t͡sʼ > s' > s̰

(t͡s >) s > s̤
(k͡x >) x > x̤

pʼ > p̰ :: p > p̤
tʼ > t̰ :: t > t̤
kʼ > k̰ :: k > k̤

Perhaps that earlier glottalisation affected vowels in some ways? Perhaps, the nasals had glottalisation as well, and that gave rise to some weird-ass vowel alternations in inflection?
Just a thought, just a thought...

[Tropylium]

For brevity, I'll use voiced/voiceless symbols here for stiff/slack voice. The first thing to note is the lack of (slack) /x/: this suggests that the voice distinction is not a recent innovation. I gess Wattman's ex-glottalization hypothesis makes some sense…

In the vowel system, what does "(u)" mean? Is this a rare vowel? If so, then probably there was a change u ʉ at some point. If there are example words it might be possible to work out what the conditions for retention of /u/ are.

Beyond that, it's a pretty clean 2×3 system that doesn't really require an immediate explanation. You say you don't like the words you've generated, so I won't ask to see some and try to work backwards…

You say nothing about prosody, so I'll arbitrarily assume penultimate stress accent.

OK then, let's crank this onwards a little.
• /ʉ/ would likely push /ɪ̵/ towards [ə], and pull /ɔ/ upwards. To spice things up a little, let's say /ɔ/ actually splits, and we end up with /u/ by default, but still /ɔ/ if /ʋ/ precedes, or if another open vowel (/ɛ a ɔ/) follows. Including /ɔ/ that was itself raised, so eg. *kjɔnɔ kjɔnu.
• Word-final /ɛ/, by some degree of analogy, also raises to /i/.
• I assume "prefixal" /s k/ and coda /t/ assimilate in voice? But what, then, happens before the sonorants? Let's assume slack voice is the default. Next, in stiff-voiced clusters, /gb gd gz gɣ zb zd zg zɣ db dd dg dɣ/ an epenthetic shwa is inserted (without changing the stress pattern), with the exception of /dz/ which will be pronounced as an affricate.
• Lenition is always a fun way to wreck some havoc in a system:
• intervocalic (l)d (l)g (l)ʔ, unless preceding a stress'd vowel
• mb nd ŋg b d g, to phonemicize the new /ʔ/
• s ɹ if not adjacent to another obstruent; ns ls sl n l l. Initial ɹm ɹn ɹʋ ɹj persist.
• More marginal phone(me)s: tj nj nz ts nts ndz (NB dz ts are implicitly pre-existing clusters); also mʋ mp.

A good tricks to pull off next (or interweave into the previous) might be syllable-closure conditioned vowel shifts. But this should do for starters.

[Nortaneous]

I'll do things the opposite way, taking stiff voice as the default, and writing them with voiceless symbols and slack-voiced consonants with voiced symbols, as in Javanese. That seems to be easier, since there's /x/ but no /ɣ/; it's harder to justify, so I'll just assume it for the protolang. Also, I'm calling that mid vowel /ɘ/.

The starting inventory, then, is /p t k b d g s x z m n ʋ l j i ʉ u ɘ ɛ ɔ a/.

Where'd the /ɣ/ go? It turned into /h/, which turned preceding /p t k s x/ into /b d g z h/ (phonation is realized on the vowels more than on the consonants, so breathy-voicing a vowel could easily make this happen) and then dropped, or turned into a glide after a high vowel. So: *ɣata *tuɣa *xiɣu ata duʋa iju.

/ɘ/ looks like it wants to be either a high or a mid vowel. Let's make it assimilate: it becomes a high vowel word-initially and after high vowels, and a mid vowel after mid and low vowels. Note the possibilities for alternation here: throw in some cliticizing, and *mɘra, *ka mɘra mira, kmera. If there's no/marginal /u/, do the same for /ɔ/: *mɔra, *ka mɔra mura, kmɔra.

Might as well keep the syllable structure consistent: k > x / #_C, followed by some sort of enforcement of the sonority hierarchy: metathesis before plosives, aspirated plosives (if you want a PIE-style phonation inventory, this is the way to go), epenthetic vowels, etc. Or enforce CVCV instead and nuke them, expand the phonation distinction to liquids, etc.

That's all I can see in it now, but I might look at it again later.

[Vuvuzela]

Tugen Luliti (Lilithian)

This is supposed to be a tri-literal root language, based somewhat heavily on Hebrew. I'm currently struggling through creating "binyanim" (roots) for the language. The name is inspired from a fictional group in a comic book my friend is working on called the "Order of the Snake". I was inspired by the mythological story of Lilith, and Hebrew was the first linguistic inspiration for it, along with other Semitic languages.

So would this be the language Lilith spoke?

Phonology

/p b t d k g ʔ/ <p b t d k g ˀ>
/m n ŋ/ <m n ng>
/f v θ ð s z ʃ ʒ x ɣ h/ <f v th dh s z š ž kh gh h>
/t͡s t͡ʃ d͡z d͡ʒ/ <ts tsh dž dž>
/r l j w/ <r l y w>

I'm going to assume the italicized was a mistake, but was the bolded as well?

/a aː e eː i iː y yː o oː u uː/ <a ā e ē i ī y ȳ o ō u ū>

Does this work like some afrasian languages, where long vowels can be an inherent part of the root, but short vowels are entirely up to grammar?

Allophony
n > ŋ /_k,g
p,t,k > f,θ,x /V_V
b,d,g > v,ð,ɣ /V_V
s,ʃ > z,ʒ /_C[+voice]
b,d,g > p,t,k /_C[-voice]
p,t,k > b,d,g /_C[+voice]
v > f /_C[-voice]

Small nitpick: these seem to be rules for assimilation, rather than allophony, since all of your "allophones" are actual phonemes in the language. Don't get me wron, they're alright assimilation rules (although part of me wonders why dental and velar spirants don't assimilate for voice), but I'd be interested in seeing your actual allophony (that is, phonemes having multiple different realizations that are not, themselves, phonemes), especially for the vowels.

(having trouble with phonotactics, mostly because I'm not sure what the phonotactics should be in a triliteral language)

In a triliteral root language, phonotactics should be rules that govern the way phonemes are allowed to go together to form words. Same with any language.
The difficulty with saying "these are common features of triliteral root languages, phonologically and grammatically" is that all languages which use triliteral roots belong to one family, Afro-asiatic, and it may be difficult to differentiate between features those languages have because they're triliteral and features those languages have because they're Afrasian. The best I could say is that is should allow a wide variety of clusters, but only a limited size, but I could be totally wrong about that. So do whatever you want.

[8Deer]

Həilh

/p b t tʷ d dʷ k kʷ g gʷ ʔ/ <p b t tw d dw k kw g gw '>
/ts tsʷ dz dzʷ tʃ tʃʷ dʒ dʒʷ/ <ts tsw dz dzw tš tšw dž džw>
/β ɬ ɬʷ ɮ ɮʷ s sʷ z zʷ ʃ ʃʷ ʒ ʒʷ xʷ ɣ ɣʷ/ <v lh lhw l l s sw z zw š šw ž žw xw ɣ ɣw>
/w/ <w>
/m n nʷ/ <m n nw>

/i i: e e: ə ə: a a: o o:/ <i ī e ē ə ə̄ a ā o ō> + nasalization marked by <Vn>

Allophony:
Some dialects replace /xʷ/ with /ɸ/.
Some speakers replace /ɮ ɮʷ/ with /l lʷ/
Some speakers merge /tsʷ dzʷ/ with /tʃʷ dʒʷ/, although this is considered substandard. (All this is vague, I know, still a WIP).
/ə/ is /ʊ/ before rounded consonants and /ɪ/ before unrounded alveolar consonants.

Messed up another phonology with sound changes and came up with this. Original phonology was somewhat Uto-Aztecan-ish. I wanted to create a Berber feel with this one but instead ended up with something completely different. Not really sure about phonotactics yet, other than that clusters don't occur initially. Also, it needs a better orthography.

Example word with some inflections:
<wōts> "child"
<wəɣəzw> "children"
<nwōts> "my child"

[Ghostfishe]

Kôbogzrak
/b g k l m n r w z/
/a â e ê i î o ô u û/
Circumflex indicates tenseness. Or something like that. I'm not convinced it's entirely consistent but that seems to have been the general idea when I started.

eg. Zârak ga grôzgunin zînenak grôz
Sarnath the city is big

Generally double letters are contracted into one unless they are vowels, in which case the weaker of the two is dropped. Or if the second one is part of a suffix, a z is shoved in between them.
In its very earliest incarnation it was designed to be part of a game, something similar to Bibble. But then I changed my mind and decided I wanted to use it for other purposes. I'm gradually revising it little by little to make it less cheap.

[Tropylium]

Kôbogzrak
/b g k l m n r w z/

If you're going to have only four obstruents, /b z g k/ is pretty wack, assuming that this is supposed to sound naturalistic at all. At least I'd expect the sibilant to be unspecified for voice rather than uniformly voiced… also having both /l r/ but no corresponding stops would seem likely to push the latter towards [d] (at least outside of consonant clusters).

[brandrinn]

Wow, very helpful, guys! It really gave me a lot to think about.

[Ghostfishe]

assuming that this is supposed to sound naturalistic at all.

Yeah it's not really. "Craplang" might be more appropriate than "conlang"?
I have other conlangs that I treat more seriously but Kôbog is mostly just for fun. And given the kind of atmosphere it was meant to compliment, "pretty wack" is probably more appropriate.

[Vuvuzela]

Kôbogzrak
/b g k l m n r w z/
/a â e ê i î o ô u û/

I can't tell if you created this phonology just to break every phonological universal ever, or to speculate what it would be like if you spoke Hawai'ian and Black Speech simultaneously. Either way, good job.

[Nortaneous]

beep boop fucking with insular kett

/p_h (t_h k_h) p t k/ <p th kh b t k>
/p\ f v s (S Z) X h/ <w f v s - - h ḩ>
/m n N/ <m n ng>
/l 4 j H w/ <l r i y u>
/a e/ <a e>

/t_h k_h/ appear mostly in loanwords, but also as contractions of /th kh/ sequences. /S Z/ appear mostly in dialects, which aren't written phonemically, but if I add them to standard I guess they'd be <sh j> or something.

also /ej eH ew/ are [Ij YH 8w] so it could be analyzed as having five vowels, except then high vowels couldn't appear before word-final consonants (or maybe CC clusters, i know nothing about the phonotactics of the damn thing) and that is kind of dumb.

actually the only attested word-final CC clusters where the first C isn't a semivowel or identical to the second C are mb mh nt lt nc (and mh appears in only one word) so i guess i'm going to go with that

[Grunnen]

Now I came up with a phonology that I think must be almost identical to some native American language's phonology, but I can't find such a language. So my question is if anyone recognises this as being very similar to ...:

Code: Select all

p  t       k kʷ       ʔ 
pʰ tʰ      kʰ kʷʰ
   ts  tʃ 
   tɬ  
f  s   ʃ   x xʷ  χ χʷ h
   ɬ 
m  n       ŋ ŋʷ 
   l             ʀ ʀʷ 
w      j  

[WeepingElf]

Now I came up with a phonology that I think must be almost identical to some native American language's phonology, but I can't find such a language. So my question is if anyone recognises this as being very similar to ...:

Code: Select all

p  t       k kʷ       ʔ 
pʰ tʰ      kʰ kʷʰ
   ts  tʃ 
   tɬ  
f  s   ʃ   x xʷ  χ χʷ h
   ɬ 
m  n       ŋ ŋʷ 
   l             ʀ ʀʷ 
w      j  

A nice consonant inventory that is indeed reminiscent of indigenous American languages, especially of western North America. Now choose a set of vowels and build a phonology on that.