Post your conlang's phonology

[cromulant]

touchy touchy

[Chagen]

Oh come on. I was thinking of rebooting Vzodyet anyway and I though I might have well answered that post.

Hearing "you don't know anything Semitic languages" is very aggravating when you've been checking out an arabic handbook and reading your school library's copy of The Cambridge Encyclopedia Of The World's Ancient Languages and checking out the Hebrew, Akkadian, and Ancient North/South Arabian sections everyday.

[احمکي ارش-ھجن]

I've been working on a tricon root language too and it's not a complete rip-off of Hebrew...
Here's the phoneme inventory:

Phonology
Vowels

The vowels /a e i o u/ are also distinguished by length. Stress is always placed on the long vowel.
Consonants

The stops /p b t d/ are distinguished between normal and uvularized.
Not listed in the table is that the voiceless stops /p t k q/ are also distinguished by pre-aspiration.

Syllable structure:
The syllable structure is at minimum (C)VC, though there can be a maximum of (C)CVC.

Allophony:

• The palatal fricative [ç] is an allophone of /h/ when it's net to /j/
• The phonemes /x χ/ become voiced [ɣ ʁ] when next to voiced stops.
• The phonemes /ʃ ʒ ʧ ʤ/ palatalize next to front vowels and /j/ and become [ɕ ʑ ʨ ʥ]
• The phoneme /i/ becomes rounded [y] when it is next to /ɹ~ɾ/ (this is due in part that I find it extremely difficult to pronounce a flap/tap without turning sound into [y] sound, but every other vowel is fine.

[GBR]

Introducing Eastern Lartau. Should have it's own thread soon, but wanted to run this by you guys first. No where near done on the phonology, but since (shock horror) phonology is maybe my least favourite part of conlanging, I've come pretty far.

Eastern Lartau is a tonal language with a relatively small phonetic inventory. Plosives, fricative and affricates contrast a labialised and unlabialised series, and there is rounding distinction in front vowels. There is also phonemic vowel length, with long vowels roughly equal in length to diphthongs.
Nearly all words in Eastern Lartau are monosyllabic. Disyllabic words are the result of compounding monosyllabic words or reduplication, and under-go tone sandhi.



Tones in long and short vowels:


I'm thinking of having a third set of tone realisations for diphthongs, but since the diphthongs are already a bit of a mess, I haven't started on this. Speaking of diphthongs (don't hate me, these are just my initial splurge):
<ao ai au oa ou ea eu ei> (Not sure on the exact phonetic realisation yet, but close to what you'd expect, given the orthography)

Low tone gains creaky voice in long syllables (long vowel/diphthong) and before a nasal.

Syllables are (C)V(N), where N is a nasal or /ɾ/. Here's the explanation from my notes:

Initials
All consonants permissible initially. Furthermore, /ŭ/ and /ĭ/ appear as semi-vowel initials, written <w> and <y> respectively, e.g. /ŭǒŋ/ wón tusk, /ĭɐ̰̀ː/ yaã five. Labialised consonants also undergo velarisation before rounded vowels, e.g. /t^wɣoː/ doo pig, pork, /s^wɣœ̌/ zeò leg.

Medials
Any vowel, long or short, and any diphthong is permitted medially. /u/ may not appear after any labiallised consonant.

Finals
Only /m ŋ ɾ/ are permitted finally.

Tone sandhi occurs in compound words thus:

[Nortaneous]

Woke up with a phonology fully formed in my head, not gonna do anything with it but

Code: Select all

m  n  ŋ  b  t  d  k  
mʼ nʼ ŋʼ pʼ tʼ lʼ kʼ ʼ
mʰ nʰ ŋʰ pʰ tʰ lʰ kʰ
m: n: ŋ: p: t: l: k: r  j  w
         ɓ     ɗ     rʼ jʼ wʼ
         h  s  l  x  rʰ jʰ wʰ

a i u ą į ų

only allowed consonant cluster is /sŋ/~/sn/~/st/~/ts/

high vs low tone I think

Orthography:
<m n ng/ŋ/ñ b t d k> + <’ h>
<mm nn ngg...>
<bh dh r’ y’ w’>
<h s l x rh yh wh>

[Pole, the]

What a pretty phonology you woke up with. I want to wake up with such phonologies, too!

[Kvan]

I'm toying around with a large inventory that is informed a bit by Australian Aboriginal languages mixed with Icelandic (sans the rounded front vowels), Kunjen is a pretty good approximation for the sound system I'm shooting for. Some touches of Laghuu too (save for any fricatives). The vowel system is a bit wonky but tenable, mostly influenced by Lakota plus rhoticization.

Consonants
Voiceless Nasals: /m̥ n̪̊ ɳ̊ ɲ̊ ŋ̊/
Voiced Nasals: /m n̪ ɳ ɲ ŋ/
Voiceless Unaspirated Stops: /p t̪ t͡s ʈ c k q/
Voiceless Aspirated Stops: /pʰ t̪ʰ t͡sʰ ʈʰ cʰ kʰ qʰ/
Voiced Stops: /b~^bm d~^dn ɖ~^ɖɳ g~^gŋ/
Voiceless Approximants: /l̥ ȷ̊ w̥/
Voiced Approximants: /l j w/
Voiceless Taps/Trills: /r̥/
Voiced Taps/Trills: /ɾ r ɽ/

Notes on the Consonants:

• There are no phonemic fricatives in this language however /t͡s/ has [s] as one of it's allophonic realizations, generally when /t͡s/ is part of a cluster.
• The phonemes /cʰ kʰ qʰ/ are often realized as [c͡ç k͡x q͡χ]
• The voiced stops /b~^bm d~^dn ɖ~^ɖɳ g~^gŋ/ are all underdifferentiated. In coda position they are (almost always) realized as prestopped nasals: /^bm ^dn ^ɖɳ ^gŋ/ and intervocalically and at the onset position they are voiced plosives /b d ɖ g/.

Vowels
Oral Vowels: /i u a ə/
Nasal Vowels: /ẽ õ ə̃ ɑ̃/
Rhotic Vowels: /ɚ ɑ˞/
Rhotic Nasal Vowels: /ɚ̃ ɑ̃˞/

Notes on the Vowels:
Rules of thumb on the geminates:

• Geminate closed vowels are actually realized as diphthongs: /i:/ > /iə̯/, /u:/ > /uə̯/
• Geminate /ə/ becomes either [ɜ:] or [æ:], perhaps conditioned by phonological surroundings or dialect (or both).
• Geminate /a/ becomes /ɑ:/
• Geminate vowels agree in nasalization and/or rhoticization
  • ə + ə̃ > ə̃:,ɜ̃:,æ̃:
  • ə + ɚ > ɚ:,ɜ˞:,æ˞:
  • ə + ɚ̃ > ɚ̃:,ɜ̃˞:,æ̃˞:
  • ə̃ + ɚ > ɚ̃:,ɜ̃˞:,æ̃˞:
• Two geminates cannot occur adjacently even at syllable boundaries.

Other tips and tricks:

• Vowels occuring near uvulars tend to be far more laxed/mid-centralized than one which are not adjacent to uvulars.

Some Quick Phonotactics
Syllable Structure:
A general conception of the underlying syllable structure is (N)(C)(A)V(C) where C translates to any consonant, N equates to the nasals and A is representative of the approximants.

Some important notes on the syllable structure:

• If there is a nasal onset in a cluster the nasal will be of the same place of articulation (homorganic, yes?) to the following consonant.
• Instances where there are two consonants at onset it can either be Nasal + C or C + Approximant. There are some more specific rules. In nasal initial consonant clusters, the second consonant cannot be a homorganic nasal, and only a few clusters of nasals are allowed (no initial geminates allowed). In instances of C + approximant, the original consonant cannot be a trill or approximant.
• Clusters of up to three consonants are allowable but are highly restricted to the aforemention Nasal + Homorganic Stop + Approximant.
• I'm still debating on how to resolve mixed voicing clusters (i.e. m̥b, nsl, etc.). I'm thinking they sum of them will all devoice but it requires a bit more thought.

I may have gotten out of hand there.

[Pinetree]

Hey all. It's been a while, but I mad another phonology that will probably not go anywhere. I had Gaelic and Japanese on my mind when I created it, but i referenced neither during the process.

Code: Select all

Consonants
	Fortis
		 N: m mʲ n ɲ ŋ ŋʲ
		-V: p pʲ t c k kʲ
		+V: b bʲ d ɟ g gʲ

	Lenis
		 N: ʋ̝̃ ʋ̝̃ʲ ɹ̝̃ j̝̃ ɰ̝̃ ɰ̝̃ʲ
		-V: f fʲ θ θʲ s ɕ x ç
		+V: v vʲ ð ðʲ z ʑ ɣ ʝ

	Approximant
		    ɹ ɹʲ j w ɥ

Vowels
	í ì---ú ù
	 |  |  |
	ɛ́ ɛ̀ ə ɔ́ ɔ̀
	  \ | /
	   á à

	Diphthongs
		iː ui̯ ɛi̯¹ ɔi̯ ai̯
		iu̯ uː ɛu̯ ɔu̯² au̯
		iɛ̯ uɛ̯ ɛː¹ ɔɛ̯² aɛ̯³
		iɔ̯ uɔ̯⁴ ɛɔ̯ ɔː² aɔ̯
		ia̯ ua̯ ɛa̯³ ɔa̯⁴ aː⁴

		¹rᶻᵈ as [eː]
		²rᶻᵈ as [oː]
		³rᶻᵈ as [æː]
		⁴rᶻᵈ as [ɒː]

		This grid applies to all combinations of tone. When diphthongs and long vowels occur, tone becomes a contour, either high, falling, low, or rising.

Syllable Structure
	(C)V(V)(N)
	Lenis consonants only occur word-initially and approximants only occur word-medially.

Sample Text
máɲɔ̀ ɕɛ̀n tʲɛ̂ɔ̯bɒ́ːn ɣánàŋʲ ðʲǽːpáɲa nə ʋ̝̃ətɛ́bə sóː gɒ̌ːŋ ɟɛ́mʲ ɕó sûi̯ ðʲɛ́mərə ðájən j̝̃úŋ sáməbà ʑáu̯ɥòː xə çɛ́ɲ ɟɔ̀

[Pole, the]

What do these mean:?

Code: Select all

      ¹rᶻᵈ as [eː]
      ²rᶻᵈ as [oː]
      ³rᶻᵈ as [æː]
      ⁴rᶻᵈ as [ɒː]

Also, why /uɔ̯⁴/ and not /aɔ̯⁴/?

[Pinetree]

What do these mean:?

Code: Select all

      ¹rᶻᵈ as [eː]
      ²rᶻᵈ as [oː]
      ³rᶻᵈ as [æː]
      ⁴rᶻᵈ as [ɒː]

Also, why /uɔ̯⁴/ and not /aɔ̯⁴/?

RealiZeD. Also, oops, nice catch. Meant to have that, but I must have missed it.

[roninbodhisattva]

Inspired by the African conlangs thread, here's an inventory for a North Atlantic inspired language:

Code: Select all

p      t      c      k      q      ʔ
b      d      ɟ      g     
mb    nd     ɲɟ     ŋg     ɴɢ
ɓ      ɗ
m      n      ɲ      ŋ
f      s      ʃ             χ
w      l      y 
       r  

i         u
e         o
ɛ         ɔ
     a

+ length distinction

[Turama]

This is the phonology of my not yet named conlang.
Nasals: m n
Plain stops : p t k q
Breathy stops: bʱ dʱ gʱ ɢʱ
Labialized stops: pʷ tʷ kʷ
Ejective stops: p' t' k' q'
Plain affricatives: ts tθ
Breathy affricatives: dzʱ dðʰ
Labialized affricatives: tsʷ tθʷ
Ejective affricatives: ts' tθ'
Plain liguids: s θ r l
Breathy liquids: zʱ ðʰ rʰ lʰ
Labialized liquids: sʷ θʷ rʷ lʷ
Ejective liquids: s' θ' r' l'
Approximants: j w

Vowels: i-a-u-(ə)

The most influece/ideas came from NC Languages, which have big phonemic inventories and very many consonant clusters and from Proto-Simitic. I know NC Languages have much bigger inventories, but I omited the velar-fricative and uvular-fricative series, because I plan to do daughter languages, and there should be enough place to develop. The breathy voiced consonants and (coaccidentally) the lack of velar-fricatives and uvular-fricatives came from PIE. The dental fricatives and the the vowel system came from (Proto-)Simitic.

Actually there aren't affricatives, but consonant clusters, but many consider these clusters as affricatives. (Why?)

This is the main phonology, which the roots made of. There definitively will be allophones.
What do you think about my phonology? Are these breathy/aspirated liquids possible? I know everything is possible, but I wonder if this regularity is common/natural. Do you have ideas how I could improve it?

PS: If you find some grammatical errors, please correct me.

[Dewrad]

Not a grammatical error, but it's "Semitic".

[Nortaneous]

/p b ɓ ⁿp ⁿb t d ɗ ⁿt ⁿd ʈ ɖ ⁿʈ ⁿɖ c ɟ ⁿc ⁿɟ k g ⁿk ⁿg kp gb ⁿkp ⁿgb/
/ɸʷ βʷ f v ⁿf ⁿv s z ⁿs ⁿz ʃ/
/ⱱ l r ɽ j w/
/ⁿbʙ ⁿdr/
/a̠ ɛ̠ ɔ̠ (ə̠) e o i ɨ u ã̠ ɛ̠̃ ɔ̠̃ (ə̠̃) ẽ õ ĩ n̩ ũ (m̩)/
four tones: low, low-mid, mid, high

+/-ret vowel harmony, but ɨ may give either a̠ or ə̠ -- ə̠ in noun compounds, a̠ otherwise.
n̩ is realized as ɨ̃ after a consonant, and can't appear after a liquid
schwa is rare outside vowel harmony
voiced stops > nasals before nasal vowel
m̩ can't appear with an onset or coda (could also analyze n̩ ɨ̃ as separate)

could do two orthographies:
African-style:
<p b ɓ mp mb t d ɗ nt nd ʈ ɖ nʈ nɖ c j nc nj k g nk ng kp gb nkp ngb> + <m n ɳ ñ ŋ ŋm>
<fh vh f v mf mv s z ns nz sh>
<vw l r ɽ y w>
<mbr ndr>
<a ɛ ɔ ə e o i ɨ u an ɛn ɔn ən en on in ɨn/n un m>
<à â a á>

or Yi-style:
<p b bb np nb t d d nt nd ch zh nch nzh q j nq nj k g nk mg kp gb nkp ngb> + <m n nr ny ng ngm>
<fu vu f v nf nv s z ns nz sh>
<vr l r rr y w>
<mb ndr>
<a ie uo e ee o i e u an ien uon en een on in en~n un m>
<p x 0 t>

I don't think I'm going to do anything with this so anyone who wants it can take it. I have never been able to get that particular no-nasals thing to work in the past. (Amqoli almost has no nasals, and I could get rid of them entirely were I so inclined, though if I did, phonetic nasals would be allophones of /w l/, and not produced by nasal vowels.)

[Nortaneous]

Another inventory that I'm not going to do anything with

pʰ p b mb tʰ t d nd tɬʰ tɬ dɮ ndɮ tsʰ ts dz ndz tʂʰ tʂ dʐ ndʐ tɕʰ tɕ dʑ ndʑ kʰ k g ŋg qʰ q ɴɢ
f v s z ɬ l ʂ ɻ ɕ j x ɣ w ħ ʕ
m n ɳ ɲ ŋ
æ ɑ e ɤ o i ɯ u

+ two or three tones

[8Deer]

Gonzhal ['ɣɔ̃zɐw]

Consonants
/b t d k g/ <b/v t d k g>
/s z h/ <s z h>
/m n/ <m n>
/l w j/ <l w y>

- /b g/ are [v ɣ] in all positions except after a nasal.
- /h/ is deleted after another fricative.
- The stop /d/ becomes [ɾ] between vowels.
- Voiced obstruents become voiceless word-finally.
- /l/ vocalises to [w] in syllable codas.
- There is also a syllabic nasal that can occur in unstressed syllables and is homorganic with the following consonant. It is usually transcribed /ɴ/.
- When two fricatives come into contact, the second hardens to a stop. Likewise, when two stops come into contact, the first becomes a fricative.

Vowels
/i ɨ e a o (u)/ <i u e a o (uw)>
/ai aɨ oi oɨ/ <ai au oi ou>

- The vowel derives from earlier /ɨw/, mostly from the vocalization of /l/, or from loanwords. In these cases, it is spelled <uw>. It also occurs as the word-final unstressed allophone of /o/, where it is written <o>.
- Stressed vowels are marked with an acute, unless word-initial. Stressed vowels in open syllables are long. Unstressed /i ɨ e a o (u)/ become [ɪ ɨ̞ ɛ ɐ ɔ (ʊ)]. Unstressed word-final /e a o/ become [i ə u].
- Vowels followed by a nasal and a fricative become nasalised and the nasal is deleted. These vowels may have a slightly different quality as well.
- /ɨ/ is “buzzed” [z̩] after an alveolar consonant. likewise becomes [β̩].

Sketch I wrote in a Syntax lecture inspired by the small consonant inventories thread and also by the Yi languages. Originally had no labials cause I was going for a more Iroquoian feel but I decided to add them for plausibility.

[Kvan]

Consonants
Unaspirated Plosives: /p t̪ c k/ <p t c k>
Aspirated Plosives: /pʰ t̪ʰ cʰ kʰ/ <ph th ch kh>
Voiced Plosives: /b d̪/ <b d>
Unvoiced Nasals: /m̥ n̪̊ ɲ̊ ŋ̊/ <mh/hm nh/hn ñh/hñ ŋh/hŋ>
Voiced Nasals: /m n̪ ɲ ŋ/ <m n ñ ŋ>
Unvoiced Trills: /r̥/ <rh>
Voiced Trills: /r/ <r>
Voiceless Approximants: /l̥ ʎ̥ w̥/ <lh/hl yh/hy wh/hw>
Voiced Approximants: /l ʎ w/ <l y w>

Vowels[size]
Monophthongs: /i ɨ u e ə o ɛ a/ <i î u e ô o ê a>
Diphthongs:

• Və Series: /iə̯ ɨə̯ uə̯ eə̯ oə̯ aə̯/
• Vi Series: /ai̯ əi̯ oi̯ ui̯/
• Vu Series: /au̯ ɛu̯ əu̯ iu̯ ɨu̯/
• Vɨ Series: /aɨ̯ əɨ̯ uɨ̯/

There are no real triphthongs but there are "W on-glides" which create a smaller closes class of faux triphthongs.

Maximum syllable structure is
C(r̥,r,w̥,w)V₁V₂C

[cromulant]

syllable structure is CV(C), codas banned word-finally

allowable cross-syllable clusters:
-geminates
-same POA/MOA, different phonation
-plosive + same POA, any phonation combination

the inventory is probably pretty vanilla, but you get contrasts like /pp pb bp bb/, /tt td dt dd/, /kk kg gk gg/, /ts tz ds dz/, /ss sz zs zz/, /pm bm/, tn dn/ etc. popping up all the time

maybe I should reanalyze all the clusters as segments (with a four-way voice contrast??) and say the phonotax is just CV no, keep them medial for now

[----]

So in that language would /ts/ be pronounced as an affricate or two different releases?

[Mornche Geddick]

Phonology for the Children of Apʰɯron in the CCC game.

Vowels : i e a o u ɯ and ɣ (the latter are close and mid back unrounded vowels).
Stops : p pʰ t tʰ k kʰ
Fricatives : s ʃ h
Nasals : m mʰ n nʰ
Liquids : r, l

Major pulmonic distinction is aspiration, not voicing, and it extends to the nasals. l is clear. r is a flap between vowels and a trill elsewhere. In some dialects the aspirate stops may be fricativised, in others non-aspirate stops may be voiced in between vowels.

[Hallow XIII]

Vowels : i e a o u ɯ and ɣ (the latter are close and mid back unrounded vowels).

The symbol you are looking for is ɤ.

[cromulant]

So in that language would /ts/ be pronounced as an affricate or two different releases?

For now I'll just say it's phonetically an affricate but there's no call to grant it phonemic status; it's just one of the allowable medial clusters.

That said, I may end up contrasting coda plosive + onset fricative with actual phonemic affricates, the former being realized as unreleased plosive + fricative. ([t̚s] and [t͡s] are phonetically distinct).

The language is called Yatdanna and the inventory is thus:

Code: Select all

p  b     t  d     c  ɟ     k  g     
f  v     s  z                         h
   m        n                
            l        j

a e i o u

written as in IPA except:
/ɟ j/ <j y>

Code: Select all

Medial clusters:

pp pb pf pv pm
bp bb bf bv bm
tt td ts tz tn tl
dt dd ds dz dn dl
cc cj
jc jj
kk kg
gk gg
ff fv
vf vv
ss sz
zs zs
mm
nn
ll

/h/ and /j/ can't geminate.

Two tones: H, L. High realized as 5 before voiceless consonants, 4 before voiced. Low realized as 3 before voiceless consonants, 2 before voiced. Vowels nasalize before nasals.

Stress falls on the rightmost closed syllable, no matter where in the word that syllable is. If there are no closed syllables, stressed falls on the final syllable. Stress manifests as an increase in vowel length; volume and vowel quality aren't much affected. Also in stressed syllables: high tone to 6 before voiceless, low tone to 1 before voiced.

[cromulant]

Roman Alphabet Conlang. Designed to have the Roman alphabet be designed to represent it. 21 vowels and 5 consonants 21 consonants and 5 vowels

/p t k q b d g/ <p t k q b d g>
/tS) dZ)/ <c j>
/f v s z x h/ <f v s z x h/
/m n/ <m n>
/r/ <r>
/l/ <l>
./j w/ <y w>
/i e a u o/ <i e a u o>

The only allowable syllable shape is CV. Stress is penultimate.

[CatDoom]

21 vowels and 5 consonants.

I think maybe you've got that backward.

Otherwise it looks like a fairly reasonable way to wring a phonology out of our wonderfully dysfunctional alphabet.

[cromulant]

21 vowels and 5 consonants.

I think maybe you've got that backward.

So I do. (I do that kind of shit a lot). However, that would make for an interesting design challenge...

Maybe something like

/t k m n r/
/i y e ø ɛ œ æ a ɨ ʉ ɘ ɵ ɜ ɞ ɯ u ɤ o ʌ ɔ ɑ/, C(r)V(C) syllable structure, etc...

But no, Roman Alphabet Conlang has a typologically normal number and balance of consonants and vowels.

The idea is to give my Romanization choices an in-world meaning. So for example, when I say Gac writes /ɲ ŋ/ as <cn kn>, or /d͡z/ as <ds> (because Gac has no /z/ so I don't want to use <z> in it's representation), etc., I am talking not just about the way I write it, but the way the fictional speakers, who have no connection with Earth, write it in the alphabet that was designed for Roman Alphabet Conlang, which did exist and is used in their world, and which happens to have 26 letters that map exactly to the 26 Roman letters (though of course they look different).