Post your conlang's phonology

[Přemysl]

Are there languages that differentiate post alveolars and alveolo-palatals?

Ubykh.

Leave it to those crazy Caucasians. Say those funky phonemes white guy, say those funky phonemes right.

[Herr Dunkel]

I don't really know if I've posted, but here I come!

(b) t d k~kʲ ʔ
s z ʃ ʒ ç~x~χ~h
r m n

a e~ɛ u̟~u~ŭ i~i̠~ɪ ɔ~ɔ̆ ɞ͍~ə~ɜ
ɑː~äː~ɐː ɛː wː~uː jː~iː ɔː ɞː~ɞ͍ː

[p~b~f~ɸ~β~p͡f~p͡ɸ] /b/ is used only in toponyms and hydronyms of non-Dnaric origin. It has no stable phone seeing that it's very distinct from /m/ [m]
The tildes represent allophones, which are a bitch to explain and are seriously not really important. The vowels are grouped into two rows: long and short. First short then long. There are six of each.

ɒy uy uʌ yɑ~ya jɞ
ai ui ua ia iæ

The diphthongs are spelled as below. When the vowels come in contact due to affixing (/a/ + /i/) they result in the dipthong as spelled (=/ɒy/)

eʉa aɨɞ uʉɞ
eia aie uiæ

Same crap.

Long vowels, diphthongs and triphthongs are "heavy"; this plays a vital role in some very complex stress rules.

The following combinations (Excluding diph- and triphthongs, ofc) are the legal types:

/md/ /mk/ /mʔ/ /ms/ /mz/ /mʃ/ /mʒ/ /mx/ /mr/ /mn/ /nd/ /nk/ /nʔ/ /ns/ /nz/ /nʃ/ /nʒ/ /nx/ /nm/

/nr/ /st/ /sd/ /sk/ /sʔ/ /sʒ/ /sx/* /sm/ /sn/ /sr/ /zt/ /zd/ /zk/ /zʔ/ /zm/ /zn/ /zr/ /xt/ /xd/ /xk/ /xʔ/

/xs/ /xʒ/ /xm/ /xn/ /xr/ /rt/ /rd/ /rk/ /rʔ/ /rs/ /rʃ/ /rx/ /rm/ /rn/ /tʔ/ /ts/ /tz/ /tʃ/ /tʒ/ /tx/ /tm/ /tn/ /dʃ/

/dʒ/ /dn/ /dm/ /dr/


/ae/ /au/ /aɔ/ /aɛ/ /ua/ /uɛ/ /uɔ/ /uɞ/ /ɛi/ /iɛ/ /iu/ /iɔ/ /ɔa/ /ɔɛ/ /uɛ/ /uɔ/ /uɞ/ /ɞa/ /ɞɛ/ /ɞi/ /ɞɔ/ /ɞu/


The language has the following allowed syllable structures: (C)CV(C) or (C)(C)VC or (C)Cʔ (or a very rare (C)(C)V(C) - that is, with no required consonant and plausibly open)

[----]

Here's a little rough phonology I just came up with today. I'm gonna make it into a whole language.

/m n nˤ/ - m n n
/p pˤ t tˤ k ʔ~ʡ/ - p p t t k k
/s sˤ χ χˤ h/ - s s r r h
/l lˤ j ɰ/ - l l y g

/i ɯ e a/ i u e a

-All vowels can be long (marked with an acute); pharyngealization is marked with an underdot on the following vowel. Vowels that are long and succeed a pharyngealized consonant are marked with a hook (ả). /ʡ/ is represented in the orthography as a pharyngealized accompaniment of /k/. /lˤ/ can sometimes lose coronal articulation and become /ʕ~ʢ/; vowels following /lˤ/ can also be strident (accompanying epiglottal trill)

[Nortaneous]

Saanich ripoff that I may or may not do something with:

Code: Select all

p  t  tθ  ts      tʂ  (k) kʷ  q  qʷ  ʔ      p t tx ts    tz c co k ko h
pʼ tʼ     tsʼ tɬʼ tʂʼ ɕ   kʷʼ qʼ qʷʼ        b d       dl dz ç ĝo g go
      θ   s   ɬ   ʂ       xʷ  χ  χʷ             x  s  ll z    ħo j jo
      ð       l       j   w                     r     l     y w
             
i       i
  ə o     e u
    a       a

...Oh, I forgot to add nasals, didn't I. That seems to happen a lot. /m n N N_w/ I guess, and perhaps glottalized resonants with limited distribution?

edit: NEW AND IMPROVED FAR SUPERIOR ORTHOGRAPHY FOR GREAT GLORY. VERILY, THIS IS THE MOST QUALITY PRODUCT I HAVE EVER, UH, PRODUCED FOR CONLANGIGGNGNGNN

Code: Select all

p  t  tθ  ts      tʂ  (k) kʷ  q  qʷ  ʔ      P T Ⱦ     Ć C  Ȼ Ḵ Ḱ ,
pʼ tʼ     tsʼ tɬʼ tʂʼ ɕ   kʷʼ qʼ qʷʼ        B D     Ṯ J S̸  Q K ₭
      θ   s   ɬ   ʂ       xʷ  χ  χʷ             Ŧ S Ƚ Ś    W̱ X X̱
      ð       l       j   w                   Đ     L   Y  W
m  n                          ŋ  ŋʷ         M N              Ṉ N̸
             
i       I
  ə o     E U
    a       O

(nb: my favorite part is the <N̸>, more people should use that)

[cybrxkhan]

(Cross-posted from CBB)

Here's a recent rehash of the Proto-Beseric (formerly Proto-Beseran) phonology inventory, with all the wonderful consonant cluster madness still in (though not as bad as Salish's or Georgian's or what not).

CONSONANTS
Stops: /pʰ p bʱ b tʰ t dʱ d cʰ c ɟʱ ɟ kʰ k ɡʱ g qʰ q/ <ph p bh b th t dh d ch c jh h kh k gh g qh q>
Fricatives: /f v s z ʃ ʒ ʁ h/ <f v s z sh zh x h>
Affricates: /͡tsʰ ͡ts ͡dz/ <tsh ts dz>
Nasals: /m n ɲ ɴ/ <m n nh ng>
Approximants: /j w r/ <y w r>
Laterals: /l~ɬ/ <l>

VOWELS
/a ɛ ə~ɤ i u/ <a e o i u>

Syllable Structure

(S)(St)(F)(N)(A)V(A)(N)(F)(St)(S)

Where St is a plosive, F is a fricative, N is a nasal, A is an approximant or lateral and S is an alveolar fricative.


Random possible words:

Pmryedz
Shnglwornfq
Thwongx
Skhwangz
Frulnsh
Crilmg
T'shnhyawmk
Spmrewndzs
Sblengk
Xnhultsh
Fnrolmd
Dzvyewxb
Yals
Tilq
Bweg
Mralts
Sjhnglongx
Ngwerf

[Kvan]

As a distraction from Leþwin I'm diddling around with a language set on the same world but from a different lineage than Leþwin.

Taps/Trills: ɺ ʀ~ʁ <r ř>
Nasals: m n <m n>
Laterals: l~ɬ tɬ <l c>
Plosives: p k͡p b ɡ͡b t d k g <p kp b gb t d k g>
Fricatives/Approximants: w~β ʰs ʰsʷ s sʷ ɥ x xʷ <v s s̥ z z̥ y h h̥>

Vowels:
ɪ ʏ ɛ ɑ <i u e a>

The language is fairly agglutinative and contains contrasts for length in both vowels and in consonants. I'm figuring I'll make some use of labio-palatization. Like Leþwin I think stress is determined by syllable weight with a proclivity for penultimate stress.

[Přemysl]

Komite is based on the 6 languages of the UN done as a committee. It is half auxlang and half jokelang.

/p/ <p> /t/ <t> /tʃ/~/tʂ/ <c> /k/ <k>
/f/ <f> /s/ <s> /ʃ/~/ʂ/ <x> /x/ <h>
/m/ <m> /n/ <n>
/r/~/ʁ/ <r> /l/ <l> /j/ <y>
/a/ <a> /e/ <e> /i/ <i> /o/ <o> /u/ <u>
CV(N/R)

Russian votes for p t k f s x h m n r l y a e i o u
French votes for p t k f s x m n r l y a e i o u
Spanish votes for p t c k f s h m n r l a e i o u
Arabic votes for t k f s x h m n r l y a i u
English votes for p t c k f s x m n l y a i u
Mandarin votes for p t c k f s x h m n l a i u

p passes with a vote of 5-1
t passes unanimously
c stalls with 3-3, French and Arabic and Russian ask for an epenthetic vowel concession, passed unanimously as amended.
k passes unanimously
f is passed with a vote 5-1, Arabic is told to grow up and learn a labial
s passes unanimously
x passes 5-1
h passes 4-1-1 with English abstaining
m and n are passed unanimously
r passes 4-2, French laughs in English's general direction
l passes unanimously
y passes 4-2
a, i, and u passed unanimously
e and o stall at 3-3, Chinese asks for one or the other which is turned down, after the Great Diphthong Compromise the measure is passed 4-2 with Chinese and English changing their votes.

[Nortaneous]

k͡p ɡ͡b

yessssssss

also, why are all the sibilants preaspirated?

[Přemysl]

Random possible words:

Pmryedz

Oh God, I've undergone a sound change.

[cybrxkhan]

Random possible words:

Pmryedz

Oh God, I've undergone a sound change.

That's be /pmrjedz͡/ in one swoop.

[derkins]

Gwóntu /ŋʷʊ̃tu/

/m mʲ mʷ n ɲ ŋ ŋʷ/
/p pʲ pʷ t tʷ c k kʷ ʔ/
/t͡s t͡sʷ t͡ɕ/
/ɸ ɸʲ ɸʷ s sʷ ɕ x xʷ/
/w (r) j l ʎ ʟ ʟʷ/
/i e æ ɜ ɑ o u/

The following vowels are present in most accents, but conservatives tend to keep their historic forms. Each group will typically either be wholly present or wholly absent.
/iː~ɪi eː~ɛi uː~ʊu oː~ɔu/
/ɪ̃ ʊ̃ ã/
/ɪu ɛu au ʊi ɔi ai/

Syllable structure is CV(C). The majority of accents merge /r/ into /ʔ/. In accents that preserve /r/, it is a trill word initially and a tap elsewhere. All consonants can appear syllable initially. Intervocalically, all consonants become lightly voiced and there is a distinction at morpheme boundaries in compound words. The language has phonemic lexical stress, and none of the vowels carry any weight. The phonemes /i e ɜ o u/ reduce to [ɪ ɛ ə ɔ ʊ] when unstressed or in closed syllables and /æ ɑ/ merge to [ɐ] when unstressed or in closed syllables. In accents that do not have the new vowels, /m n ŋ ʔ (r) l ʟ/ can appear in the syllable coda. Nasal assimilation is absent in the language; instead, epenthetic plosives are inserted. So, for example, /komta/ becomes [kompta]. These epenthetic plosives are usually absent in accents with nasal vowels.

Phonemes with one spelling in most accents
/m mʲ mʷ n ŋ/ <m mj mw n g>
/p pʲ pʷ t tʷ k kʷ ʔ/ <p pj t tw k kw h>
/t͡s t͡sʷ t͡ɕ/ <c cw cj>
/ɸ ɸʲ ɸʷ s sʷ x xʷ/ <f fj fw s sw x xw>
/(ɾ) l ʟ/ <d l r>
/i e æ ɜ ɑ o u/ <i e ø v a o u>

Phonemes with multiple spellings
/ɲ/ <nj gj>
/ŋʷ/ <nw gw>
/c/ <tj kj>
/ɕ/ <sj xj>
/w/ <w dw hw>
/j/ <j dw hw>
/ʎ/ <lj rj>
/ʟʷ/ <lw rw>
/iː~ɪi/ <id ih>
/eː~ɛi/ <ed eh>
/uː~ʊu/ <ul ur>
/oː~ɔu/ <ol or>
/ɪ̃/ <im in ig em en eg>
/ʊ̃/ <um un ug om on og>
/ã/ <øm øn øg vm vn vg am an ag>
/ɪu/ <il ir>
/ɛu/ <el er vl vr>
/au/ <øl ør al ar>
/ʊi/ <ud uh>
/ɔi/ <od oh vd vh>
/ai/ <ød øh ad ah>

Additionally, some accents have merged /iː~ɪi eː~ɛi uː~ʊu oː~ɔu/ with /i e u o/ and/or /iː~ɪi uː~ʊu ɪu ʊi/ with /ʲi ʷu ʷi ʲu/. So in a way I lied about some of the phonemes only having one spelling. In writing, stress is marked by an acute accent. Not long ago there was a spelling reform that changed historic <Cl Cr Cd> to <Cw Cw Cj>. The reform has been mostly successful, but it is not uncommon for people to use the old orthography for various purposes.

[Kvan]

k͡p ɡ͡b

yessssssss

also, why are all the sibilants preaspirated?

Perhaps I should have a contrasting series? I was debating on doing it, but I parsed 'em out.

[Kvan]

Gwóntu /ŋʷʊ̃tu/

/m mʲ mʷ n ɲ ŋ ŋʷ/
/p pʲ pʷ t tʷ c k kʷ ʔ/
/t͡s t͡sʷ t͡ɕ/
/ɸ ɸʲ ɸʷ s sʷ ɕ x xʷ/
/w (r) j l ʎ ʟ ʟʷ/
/i e æ ɜ ɑ o u/

The following vowels are present in most accents, but conservatives tend to keep their historic forms. Each group will typically either be wholly present or wholly absent.
/iː~ɪi eː~ɛi uː~ʊu oː~ɔu/
/ɪ̃ ʊ̃ ã/
/ɪu ɛu au ʊi ɔi ai/

Syllable structure is CV(C). The majority of accents merge /r/ into /ʔ/. In accents that preserve /r/, it is a trill word initially and a tap elsewhere. All consonants can appear syllable initially. Intervocalically, all consonants become lightly voiced and there is a distinction at morpheme boundaries in compound words. The language has phonemic lexical stress, and none of the vowels carry any weight. The phonemes /i e ɜ o u/ reduce to [ɪ ɛ ə ɔ ʊ] when unstressed or in closed syllables and /æ ɑ/ merge to [ɐ] when unstressed or in closed syllables. In accents that do not have the new vowels, /m n ŋ ʔ (r) l ʟ/ can appear in the syllable coda. Nasal assimilation is absent in the language; instead, epenthetic plosives are inserted. So, for example, /komta/ becomes [kompta]. These epenthetic plosives are usually absent in accents with nasal vowels.

Phonemes with one spelling in most accents
/m mʲ mʷ n ŋ/ <m mj mw n g>
/p pʲ pʷ t tʷ k kʷ ʔ/ <p pj t tw k kw h>
/t͡s t͡sʷ t͡ɕ/ <c cw cj>
/ɸ ɸʲ ɸʷ s sʷ x xʷ/ <f fj fw s sw x xw>
/(ɾ) l ʟ/ <d l r>
/i e æ ɜ ɑ o u/ <i e ø v a o u>

Phonemes with multiple spellings
/ɲ/ <nj gj>
/ŋʷ/ <nw gw>
/c/ <tj kj>
/ɕ/ <sj xj>
/w/ <w dw hw>
/j/ <j dw hw>
/ʎ/ <lj rj>
/ʟʷ/ <lw rw>
/iː~ɪi/ <id ih>
/eː~ɛi/ <ed eh>
/uː~ʊu/ <ul ur>
/oː~ɔu/ <ol or>
/ɪ̃/ <im in ig em en eg>
/ʊ̃/ <um un ug om on og>
/ã/ <øm øn øg vm vn vg am an ag>
/ɪu/ <il ir>
/ɛu/ <el er vl vr>
/au/ <øl ør al ar>
/ʊi/ <ud uh>
/ɔi/ <od oh ød øh>
/ai/ <ød øh ad ah>

Additionally, some accents have merged /iː~ɪi eː~ɛi uː~ʊu oː~ɔu/ with /i e u o/ and/or /iː~ɪi uː~ʊu ɪu ʊi/ with /ʲi ʷu ʷi ʲu/. So in a way I lied about some of the phonemes only having one spelling. In writing, stress is marked by an acute accent. Not long ago there was a spelling reform that changed historic <Cl Cr Cd> to <Cw Cw Cj>. The reform has been mostly successful, but it is not uncommon for people to use the old orthography for various purposes.

Lateral crazy, nasal heavy, diphthongy, rounded and palatized, I love these Austronesiany languages. So nice.

[derkins]

Thanks!
I always found rounded and palatalized consonants interesting and I was hoping to derive them in a somewhat believable way but at the same time make it hard for them to be analyzed as simple sequences. Can anyone tell me if this is at least a little realistic?

I messed up one of the spellings. It should be /ɔi/ <od oh vd vh>

[MisterBernie]

I keep fiddling around with Asuāneica:

Consonants:

Code: Select all

Labial        p b       m  (w)
Labiodental        f v 
Dental        t d  θ ð  n
Alveolar           s z 
Retroflex     ʈ ɖ  ʂ ʐ  ɳ  ɻ
Palatal       c ɟ       ɲ  (j)
Velar         k g  x ɣ  ŋ
Uvular               ʁ
Glottal       ʔ

/t/ and /d/ are generally dental, /θ/ and /ð/ interdental.
[w] and [j] generally only occur intervocally; [w] commonly as an allophone of /v/; [j] as an epenthetic hiatus vowel.
Syllable-initial consonants and consonant clusters may be labialised. This gives another place for [w], as initial [pw, bw, mw] are possible as a result of labialisation.
Two affricates [ʈ͡ʂ] and [ɖ͡ʐ] can occur initially as a result of /cr/, /ɟr/.

Vowels:

Code: Select all

	       Front	Back
Close     i  y       u
Close-mid ɛ  ø       o
Near-ope  æ			
Open              ɑ  ɒ

All vowels can be long or short, although [ɒː] is extremely rare.
Permissible dipthongs are [aɪ̯ ~ æɪ̯], [ɑʊ̯], [oʊ̯], [øʏ̯], [yø̯].

The general syllable structure is (C)(r)V(r)(C).
/r/ can be the nucleus of a syllable.

Some sandhi stuff:

• A double consonant is simplified, and the preceding vowel is lengthened.
• A palatal touching /ɻ/ becomes a retroflex (plosives are turned into affricates). If a palatal plosive precedes /ɻ/, the r drops out unless it is in the absolute coda, in which case it becomes syllabic (for example, [mɑc] + DAT -r > [mɑʈʂɻ̍], + DAT-DU -ru > [mɑʈʂu])
• A dental touching /ɻ/ becomes the corresponding retroflex.
• A dental following historical /i/ (i.e., <i>) also turns into the corresponding retroflex (this is mainly important for the accusative, where nouns ending in historical - have inherited -[iɳ], but nouns in - from later developments have regular -[in], for example [ˈmɑɻi] > [ˈmɑɻiɳ], but [ˈmiɻje] → [ˈmiɻi] > [ˈmiɻin]). Basically, it's a diachronic rule that has not yet been levelled one way or the other by analogy).
• Any vowel touching , [j] or [ɲ] is fronted; /e/ is raised instead. [ɲ] shifts to [n].
• /e eː/ and following /o oː/ merge into [ø øː].
  /e eː/ and following /u uː/ merge into [øʏ̯].
• /i iː/ and following /o oː/ merge into [yø̯].
  /i iː/ and following /u uː/ merge into [y yː].
• [o] followed by any other vowel than becomes [ʷV] (for example, [ˈɑto] + PL -a > [ˈɑtʷa])
  [ø] becomes [ʷV+front] (for example, [mø] > [mʷæ])
• followed by a vowel becomes [ʷV] (for example, [ˈmɑnu] > [ˈmɑnʷa])
  [y] becomes [ʷV+front] (for example, [my] > [mʷæ]).

So, the questions:
1) Does this make sense so far?
2) How should I represent /ʁ/ orthographically?
In general,

Code: Select all

p  b  f  v  m     -  p  b  f  v  m
t  d  θ  ð  n     -  t  d  þ  ð  n
ʈ  ɖ  ʂ  ʐ  ɳ  ɻ  -  ṭ  ḍ  ṣ  ẓ  ṇ  r
c  ɟ  s  z  ɲ     -  c  j  s  z  ñ
k  g  x  ɣ  ŋ     -  k  g  x  ġ  ŋ
ʔ                 -  q

Historically, /ʁ/ is the outcome of /ɢʲ/ and /ɢʱ/ (and /ʔ/ is from /q/), but I'm not fond of a second g-with-diacritic, and something like <q̇> seems... unintuitive.
3) I've been toying with the idea of having initial nasals in a stressed syllable surface as [mb, nd, ɳɖ, ɳc, ŋg] - would that be plausible?

And FWIW, it's supposed to be relatively close to Baranxe'i, kinda Dutch to Baranxe'i's German

[Earthling]

Ř?

[Nortaneous]

ḫ, ƣ, ṙ, ŗ, ɛ. ʀ

(ɛ is a pharyngeal approximant in Kabyle or something and ʀ capitalizes to Ʀ)

[finlay]

ṛ for the retroflex, r for the uvular. Might as well; then you have a whole line of retroflexes marked with dots (except you seem to have missed ṇ from the grid).

the other latin letters you haven't used are L, H and W. You could use any of them at a stretch, although a variant of R seems like a more natural choice.

[MisterBernie]

ṛ for the retroflex, r for the uvular. Might as well; then you have a whole line of retroflexes marked with dots (except you seem to have missed ṇ from the grid).

Fixed the missing <ṇ>.
And hmm, that would make it nicely symmetrical; otoh, [ɻ] (or maybe it'll be [ɽ], that'd fit nice with the semi-Indianness of the langs) is much more frequent than ʁ (it's the reflex of both the proto-lang's /r/ and /l/), so using plain <r> for that would make it more convenient to type.

<h> currently serves as a hiatus marker (and thus is either [w] or [j], depending on the surrounding vowels' roundedness).
So yeah, I guess some r-with-diacritic will have to fulfil that role because the orthography is already confusing enough without adding <w>-as-a-rhotic

[Nortaneous]

Or yogh? Would work with thorn and eth, anyway.

[GrinningManiac]

My as-yet-unnamed language. Phonology is influenced by a love of Hindi, Polish and for the French R

Consonants:

Code: Select all

Bilabial         p b       m  
Labiodental      v 
Alveolar         s z       l
Postalveolar     t͡ʃ  
Retroflex        ʈ ɖ       ʂ ʐ      ɳ  
Palatal          ɲ 
Velar            ŋɡ        k g   
Uvular            ʁ

Vowels:

Code: Select all

	       Front  	Centre	Back
Close     iː                   uː
Near-close      ɪ   
Mid                   ə			
Open                          ɑː

[finlay]

no plain t/d? that's pretty rare. also hindi v is [ʋ].

[MadBrain]

[C] [/j,w/] V [/j,w/] T

C: Consonant
- p b t d k g
- v s ʒ ʁ
- m ɱ n ɲ ŋ ɴ
- w j

V: Vowel
- i ĩ ɨ ɨ̃ u ũ
- e ẽ ə ə̃ o õ
- a ã

T: Tone
- High, medium, low

[Maoti]

An eskimo-aleut inspired phonology I came up with for a conlang I am working on that I haven't named yet:

Consonants:
Plosive: /p b t d k g q ɢ/
Fricative: /f v s z ɬ ɮ ʃ x ɣ χ h/
Nasal: /m n ŋ ɴ/
Approximant: /l j w ʁ̞/

Vowels: /i a u/

Allophony: The high vowels /i/ and /u/ become allophonically lowered to [e] and [o] respectively when occurring before a uvular consonant (/q/, /ɢ/, /χ/, /ɴ/ or /ʁ̞/), and this change is indicated in the orthography.

Phonotactics:
Syllable structure is CV(C). All words therefore begin in a consonant. About 50% of words end in a consonant and 50% end in a vowel. Words never begin in a uvular consonant. Only the consonants /m n ŋ ɴ l ʁ̞/ and more rarely /p t k g q ɢ s x χ/ can occur in the syllable coda. There are no diphthongs. The consonant /j/ does not occur before the vowel /i/, and /w/ does not occur before /u/. Apart from these exceptions, any consonant can occur before or after any vowel.

[roninbodhisattva]

A revised inventory for Hiirawə:

Consonants /p t k q ts n w ɾ s h ʔ/
Vowels /i i: u u: a a: ɛ ə/