Post your conlang's phonology

[Haplogy]

/p t k/ p t c
/b~β~w/ w
/m n ŋ/ m n g
/f θ h/ f þ h
/j~l~ʎ/ ll

/ɪ ʏ ɛ œ ɑ o ʊ/ i y e ø a o u
/iː~ɛː yː~œː uː~ɔː ɑː/ é ǿ ó á

[Sexendèƚo]

Consonants
/m n ɲ ŋ/
/p t k ʔ b d g/
/ts tʃ cç dz dʒ ɟʝ/
/f s ʃ ç x h/
/r rʲ ʋ j/
/l ʎ/
Vowels:
[+ATR / -low] group: /e i o u ɘ̆/ /ei ɘʉ/
[-ATR / -high] group: /a ɒ ɛ ɔ ɜ̆/ /aɛ aɒ/


- +/- ATR / vowel height harmony
- Fricatives become voiced between vowels.
- Approximant/nasal+h cluster is realized as voiceless approx: /ʋh/→[ʋ̥] /mh/→[m̥], approximants devoice before voiceless stops: /rt/→[r̥t], nasals never appear before voiceless stops.
- ʔ+plosive/affricate is often realized as an ejective: /ʔk/→[ʔk']
- Voiced affricates/stops can only appear medially but do contrast with voiceless ones.
- the palatal series doesn't appear finally except /j/, and can only appear in [+ATR] roots. When present in a suffix palatals change the harmonic class of following vowels to [+ATR]
- glottal series /ʔ h/ derives from earlier uvulars. Can only appear in [-ATR] roots, and when present in a suffix changes the harmonic class of following vowels to [-ATR]
- anything can geminate

I would appreciate any comments or suggestions on what might be interesting to do/could be improved

[Shemtov]

/t t' tˡ t̪ t̪' t̪ˡ c c' cˡ ʡ ʔ/ <t t' tl d d' dl c c' cl q '>
/n n̪ ɲ/ <n nh nj>
/θ s s̪ ħ h/ <th s z hh h>
/ɬ ɬ̪/ <l thl>
/j/ <j>
/ǀ ǃ ǂ/ <x ! k>
/ǀʰ ǃʰ ǂʰ/ <xh !h kh>
/ᵑǀ ᵑǃ ᵑǂ/ <xn !n kn>
/ˀǀ ˀǃ ˀǂ/ <x' !' k'>
/ǁ ǁ̪/ <!l xl>


/i ɨ ɯ/ <i y u>
/e̞ ə ɤ̞/ <e a o>
/ai aɯ ae̞ aɤ̞/ <ai au ae ao>
All plain vowels can occur long, shown by doubling the vowel.

High-mid vowel harmony
Tones (Only contrastive on affixes and particles):
/˧ ˥/
High tone is marked by an acute accent.

[Αυτοβοτα]

I'm concerned that this inventory may be too scattered. OTOH, the languages that inspired it are similarly over the place.

Code: Select all

                Lab     Alv     Lat     Pal     Vel     Uvu     Glot
Nasal           m       n               ɲ       ŋ
Plos    vl              t                       k               ʔ
Plos    vd      b       d
Affr    eject           t͡sʼ     t͡ɬʼ                     q͡χʼ
Affr    vd              (d͡z)           d͡ʒ
Fric    vl      f       s       ɬ       ʃ       x~χ             h
Fric    vd      (v)     z                       ɣ~ʁ
Approx                          l       j       w
Vibrant                 r

The ejectives are weak and may simply be a phonetic detail of voiceless affricates.
The uvular has the allophonic range [q͡χʼ~q͡χ~qʼ~q].
f v / [+voice]_[+voice]
/z/ varies with [d͡z]. The latter is favored after sonorant consonants.
ŋ ŋː or ŋg / V_V
ŋ ŋg / #_
ŋ ɴ / _[+uvular]
n n̪ˠ / _[+labial / +velar / +uvular]
Generally nasals don't assimilate to the POA of a following consonant.
l ɫ / _[+velar / +uvular]
w ɰ / [+labial]_ ; e.g. /bwa/ [bɰa~bˠa]

Code: Select all

                Short           Long
                Front   Back    Front   Back
Close           i~ɪ     u~ʊ     iː      uː
Close-Mid       (e)     (o)
Open            a~æ     (ɑ)     aː
Diphthong       aj      aw

The mid vowels are allophones of the short close vowels. They lower in closed syllables and after /x ɣ q͡χʼ/ and the lowering spreads to the front of the word (usually) or to the whole word (rarer). /a/, the diphthongs, and the long vowels are not affected by lowering and may block the spread of lowering.
a ɑ / adjacent to x ɣ q͡χʼ

Syllable Structure: (N)C(w) { V(C)C(w) } V(C)(C)

The only "prenasalized" word-inital clusters allowed are [mb nd ŋg] (where [ŋg] is phonemically /ŋ/).
The only final consonant clusters allowed are of the type sonorant + obstruent.
Word-finally, a long vowel or diphthong may only be followed by a single consonant.
I'm looking to have aggressive agglutination so, beyond what I've written above, I'm unsure of what cluster constraints to enforce.
That's the maximal structure, but C(w)V is the most common.

Looking to have some minor ablaut, maybe consonant gradation.

I'm inspired by Arabic, South Arabian languages, and Somali.

[Nortaneous]

I haven't finished my current conlang yet, but it'll probably end up with something like this:

/p t ts (tʂ) tɕ k/
/(b) d dz (dʐ) dʑ g/
/s ʂ ɕ/
/z ʑ/
/β j l ɻ/
/m n ɲ ŋ/
/a e ə o i ɨ u/ + length

Maximum syllable structure: CCCV(ɻ N)

[2+3 clusivity]

/m mʱ n nʱ/
/p pʰ b bʱ t tʰ d dʱ k kʰ g gʱ/
/ts tsʰ dz(ʱ)/
/sʰ zʱ/ (really just [s] and [z] with right spreading delayed VOT, i.e. /sʰtV/ -> [stʰV], [zʱtV] -> [zdʱV])
/ʋ ʋʱ l lʱ ɻ ɻʱ y yʱ ɦ/ (y is, of course, not IPA's y but its j, also see also below)

Short vowels and syllabic units: /ʋ̩, ʋ̩ʱ, l̩, l̩ʱ, z̩ʱ, ə, əʱ, ɻ̩, ɻ̩ʱ, ɨ, ɨʱ, y̩, y̩ʱ, ɦ, a, a/
long vowels and syllabic units /ʋ̩:, ʋ̩:ʱ, l̩:, l̩:ʱ, z̩:ʱ, ə:, ə:ʱ, ɻ̩:, ɻ̩:ʱ, ɨ:, ɨ:ʱ, y̩:, y̩ʱ, ɦ̩:, a:, a:/

Syllable structure:

(F)(C)V(C), where F is [sʰ] or [zʱ] and where (F)(C) or (C).(F/C) cannot be a series of fricatives.

Syllabification occurs right to left

Allophones:

(1) F becomes retroflex following /r, u, K/
(2) coronal POA spreads recursively left to right from the first fricative or affricate to the next fricative or affricate unless blocked by a nasal
(3) F + coronal in onset -> retroflex cluster, e.g. /$sʰtV/ -> /$ʂʈʱV/
(4) all coronals following a retroflex become retroflex
(5) F + C -> spreads VOT
(6) /#F[unvoiced]C[voiced]V/ -> [#F[unvoiced]C[voiced]V], e.g. /#smV/ -> /#zmV/
(7) coronal stops before /ɨ(:)/ -> affricate to [TS(ʰ)]
(8) velar stops before /ɨ(:) affricate, unvoice, and pre-velarize to [kʟ̘̊(ʰ)]
(9) /i or y/ post-alveolarize all coronals and velar stops, nasal stops, affricates, and fricatives. Stop also become affricates.
(10) coronal POA then spreads recursively right to left from the last fricative or affricate to the next fricative or affricate unless blocked by a nasal
(11) nasals take the POA of following C.
(12) retroflex POA spreads recursively left to right from the first retroflex nasal stop or /ɻ(ʱ)/ to the next /V_V/ or /VN#/ nasal.

[Nortaneous]

Short vowels and syllabic units: /ʋ̩, ʋ̩ʱ, l̩, l̩ʱ, z̩ʱ, ə, əʱ, ɻ̩, ɻ̩ʱ, ɨ, ɨʱ, y̩, y̩ʱ, ɦ, a, a/
long vowels and syllabic units /ʋ̩:, ʋ̩:ʱ, l̩:, l̩:ʱ, z̩:ʱ, ə:, ə:ʱ, ɻ̩:, ɻ̩:ʱ, ɨ:, ɨ:ʱ, y̩:, y̩ʱ, ɦ̩:, a:, a:/

uh

[2+3 clusivity]

Short vowels and syllabic units: /ʋ̩, ʋ̩ʱ, l̩, l̩ʱ, z̩ʱ, ə, əʱ, ɻ̩, ɻ̩ʱ, ɨ, ɨʱ, y̩, y̩ʱ, ɦ, a, a/
long vowels and syllabic units /ʋ̩:, ʋ̩:ʱ, l̩:, l̩:ʱ, z̩:ʱ, ə:, ə:ʱ, ɻ̩:, ɻ̩:ʱ, ɨ:, ɨ:ʱ, y̩:, y̩ʱ, ɦ̩:, a:, a:/

uh

In re the inventory, the use of /ʱ/ with vowels rather than under diaereses, or something else?

Slathering a length and voicing distinction across a three quality vertical vowel system coupled with a byzantine set of nasals, approximants and a voiced fricative that can stand in the nucleus seems pretty fun to me. The more extended allophony is going to be hilarious -- hell, a third of the traditional vowels already affricate and pre-velarize velars to [kʟ̘̊(ʰ)]!

[Max1461]

A phonology I've just come up with, tell me what you think.

Inventory

/p t ʈ tʃ k/ <p t rt tgs c(h)>
/b d~ɾ/ <b d(h)>
/m n ɳ ŋ/ (m n rn ng)
/f s x/ <ps s h>
/z ʝ~ɣ/ <z gs>
/l j w/ < l y w>
/ɪ ʊ ə ɛ ɑ/ <i u a e o>
/iː uː aː/ <ii uu aa>
/à á/ <a á>

Phonotactics

Maximal (C)(C)V(C) syllable structure, with initial clusters restricted to / voiceless fricative + voiceless stop/, /nasal + stop/, /nasal + fricative/, / consonant + glide/, or one of /pk tʃk dn dɳ sf tʃf kf/. Final consonants must be one of /p t ʈ tʃ k n/.

Allophones

/d/ > [ɾ] between vowels
/ɣ/ > [ʝ] before front vowels
stops never released in final position, /t k/ neutralize to [ʔ]
/ə/ > [ɚ] adjCent to retroflexes

[Jonlang]

/p b t d k g/ p b t d c g
/f v θ ð s h/ f v th dh s h
/ɬ/ lh
/ɾ/ r
/w ɹ j/ w r y
/l/ l
/n ŋ/ n ng

rh - aspirated /ɾ/ or /ɹ/
qu - same as English queen.

/a ɒ ɑː/ a o á
/ε ʌ/ e u
/ɪ/ i
/i: u:/ í ú

/aɪ/ ai
/aʊ/ au
/εɪ/ ei

EDIT: Added /j/ (y)

[TaylorS]

NW European language isolate with French-influenced orthography:

/m n ɲ/ m n gn
/p t tʃ k/ p t ch c
/b d dʒ g/ b d j g
/f s ʃ x/ f s sh h
/j l r~ʁ/ y l r

/ ɛ e i/ ai e i
/a œ ø y/ a eu oi u
/ ɔ o u/ au o ou

Fricatives are voiced between vowels.
Voiceless stops are lightly aspirated.
Syllable-initial voiced stops are slightly implosive is stressed syllables.
Speakers under 40 tend to use the uvular R.

Naifou eu Nauned
[ˈnɛvu œ ˈnɔned]
nai-fou- eu=Nauned
1SG.PST-go to=Nantes
"I went to Nantes"

[schyrsivochter]

My conlang no. 4:

/p b t d k g/ <p b t d k g>
/p͡f b͡v t͡s t͡ʃ d͡z d͡ʒ/ <pf bw ts tc dz dj>
/f v ʃ ʒ ʁ h/ <f w c j r h>
/j/ <ı>
/m n ŋ/ <m n q>
/l/ <l>

/ɪ ɨ ʉ ʊ/ <ị y ÿ ụ>
/ɤ o ə e ø/ <o ö a e ë>
/ɛɪ̯ ɔʊ̯ ɛa̯ ɔɑ̯/ <i u ę ǫ>

No short-long vowel distinction. The vowels ị, ụ and a are considered short as they derive from short vowels; all others are long – they derive from long vowels or diphthongs.
h is only pronounced initially; otherwise either, after i (Edit: i, e or ë) derives from earlier /ç/, or, after any long vowel, from /x/; it is also used to mark any hiatus.
c and j may be pronounced [ɕ ʑ] before front vowels.
ı may be pronounced finally.
ę and ǫ may be pronounced [æ ɒ] finally.

[Mâq Lar]

/p͡f b͡v

Even though German has /pf/, non-alveolar affricates are really rare. That doesn't mean you can't have them, just wondered if you knew. Similarly with /ʁ/ as the only rhotic; it's a very German/French choice, especially as there are no other uvulars.

You seem to have a few odd choices for romanisation. <v> for /w/ is understandable as a German speaker, but might be confusing if you're writing for English speakers (you can have different orthographies for different audiences). Using <j> for /ʒ/ and then <ı> for /j/ seems confusing, as does <q> for /ŋ/. In the vowels, you use the 'plain' vowel letters <i o u> for less cardinal sounds, and represent the more usual /ɪ o ʊ/ with letters with diacritics. It all seems a bit unintuitive and unclear to me. Generally, a romanisation should be as clear as possible - easy to read and as intuitive as possible.

As for the phonology itself, I like it. Although it seems odd that there are so many high and mid vowels and no low vowels.

Any example words?

[Mâq Lar]

A phonology I've just come up with, tell me what you think.

Inventory

/p t ʈ tʃ k/ <p t rt tgs c(h)>
/b d~ɾ/ <b d(h)>
/m n ɳ ŋ/ (m n rn ng)
/f s x/ <ps s h>
/z ʝ~ɣ/ <z gs>
/l j w/ < l y w>
/ɪ ʊ ə ɛ ɑ/ <i u a e o>
/iː uː aː/ <ii uu aa>
/à á/ <a á>

I like this - kind of Australian sounding (apart from the voicing contrast). Not too keen on the orthography though: why /tʃ k/ <tgs c(h)>? Why not just <c k>? (And why the optional h? Ditto with <ps gs> for /f ʝ~ɣ/; why not just <f g>?

Also, what is /à á/?

[WeepingElf]

/p͡f b͡v

Even though German has /pf/,

... and is currently losing it. Teachers are fighting an uphill battle against that, saying things like Es heißt 'Pferd', nicht 'Ferd'. But non-coronal affricates are indeed rare.

[Mâq Lar]

/p͡f b͡v

Even though German has /pf/,

... and is currently losing it. Teachers are fighting an uphill battle against that, saying things like Es heißt 'Pferd', nicht 'Ferd'.

Didn't know that!

[schyrsivochter]

/p͡f b͡v

Even though German has /pf/,

... and is currently losing it. Teachers are fighting an uphill battle against that, saying things like Es heißt 'Pferd', nicht 'Ferd'. But non-coronal affricates are indeed rare.

You’re wrong. Even though some dialects tend to pronoune /pf/ as [f], they do so only word-initially. In the middle or end of a word, it is retained, e.g. stopfen ‘to stuff’, Apfel ‘apple’. Nobody would pronounce these [ʃtɔfn̩], [afl̩] instead of [ʃtɔp͡fn̩], [ap͡fl̩].

[WeepingElf]

/p͡f b͡v

Even though German has /pf/,

... and is currently losing it. Teachers are fighting an uphill battle against that, saying things like Es heißt 'Pferd', nicht 'Ferd'. But non-coronal affricates are indeed rare.

You’re wrong. Even though some dialects tend to pronoune /pf/ as [f], they do so only word-initially. In the middle or end of a word, it is retained, e.g. stopfen ‘to stuff’, Apfel ‘apple’. Nobody would pronounce these [ʃtɔfn̩], [afl̩] instead of [ʃtɔp͡fn̩], [ap͡fl̩].

True.

But the point of this exercise was that non-coronal affricates are uncommon. There are not many languages with such phonemes, and some people even claim that German /pf/ was not a phoneme but merely a cluster.

[Qxentio]

Jagárh
Plosives: /p b t d k g q/ [c ɟ] (allophones of k and g, respectively)
Fricatives: /f v θ ð s z ʃ ʒ χ/(x is supposed to be uvular)
Nasals: /m n ŋ/ and their allophones [ɱ ɲ ɴ]
Liquids: /l̪ r ʀ/

Short vowels: /a ɛ ɪ ɔ ʊ ə/
Long vowels: /ɒː eː iː oː uː/

You’re wrong. Even though some dialects tend to pronoune /pf/ as [f], they do so only word-initially. In the middle or end of a word, it is retained, e.g. stopfen ‘to stuff’, Apfel ‘apple’. Nobody would pronounce these [ʃtɔfn̩], [afl̩] instead of [ʃtɔp͡fn̩], [ap͡fl̩].

I grew up in the Eastphalian dialect area and now live in the Westphalian dialect area. I have noticed that /pf/ is slowly disappearing from speech around these parts, at least in the most colloquial forms. So yes, I would pronounce "stopfen" as [ʃtɔfn̩]. Maybe the [f] would be a tad longer than usual. It can also go the other way though: Kopf is often realized as [kɔp].

[schyrsivochter]

Short vowels: /a ɛ ɪ ɔ ʊ ə/
Long vowels: /ɒː eː iː oː uː/

It’s interesting to have the long form of /a/ be /ɒː/, but then again, why not? The rest of the phonology seems pretty average to me; but that’s perfectly fine. If your grammar is interesting enough, the phonology doesn’t matter that much.

Anyway, have some tea and pickles!

[Travis B.]

Short vowels: /a ɛ ɪ ɔ ʊ ə/
Long vowels: /ɒː eː iː oː uː/

It’s interesting to have the long form of /a/ be /ɒː/, but then again, why not? The rest of the phonology seems pretty average to me; but that’s perfectly fine. If your grammar is interesting enough, the phonology doesn’t matter that much.

Anyway, have some tea and pickles!

Well Hungarian contrasts [ɒ] and [aː] and Persian contrasts [æ] and [ɒ], so I do not find anything odd about this kind of contrast.

[opipik]

Mungchi:

/p pʰ b t tʰ d k kʰ/ <p ph b t th d k kh>
/m n ɲ ŋ/ <m n ny ng>
/r/ <r>
/s z/ <s z>
/ʦ ʦʰ ʣ ʧ ʧʰ ʤ/ <ts tsh dz ch c·h j>
/w ʍ l j/ <w wh l y>

/i y e ø ə ɐ a o u/ <i ʉ e ø ɨ a a꞉ o u>
/a˧ a˥ a˩ a˩˥ a˥˩/ <a ā a̱ á à>

[Qxentio]

It’s interesting to have the long form of /a/ be /ɒː/, but then again, why not? The rest of the phonology seems pretty average to me; but that’s perfectly fine. If your grammar is interesting enough, the phonology doesn’t matter that much.

As Travis B. wrote, Hungarian has a similar contrast and it was one of the influencing languages for Jagárh. /ɒː/ fit well into the rather regular vowel system of the language, and I wanted to explain it as being realized in many different ways among dialects. Only the upper class actually realizes the long a as /ɒː/, and everyone else rolls their eyes at them for that.

Anyway, have some tea and pickles!

Thanks! I'm actually having some tea right now.

[Max1461]

I like this - kind of Australian sounding (apart from the voicing contrast). Not too keen on the orthography though: why /tʃ k/ <tgs c(h)>? Why not just <c k>? (And why the optional h? Ditto with <ps gs> for /f ʝ~ɣ/; why not just <f g>?

Also, what is /à á/?

Thanks for the input! The romanization is designed to have a direct letter-by-letter correspondence with the native orthography, and those are the digraphs/trigraphs therein. The native orthography is reasonably etymological, and an <h> (along with its native equivalent) is placed between <c> or <d> and a front vowel to indicate a lack of some as-yet undetermined historical palatalization of those consonats in those positions. /à á/ is a somewhat nicer looking way of denoting /˦ ˨/, the high and low tones, respectively.

[Max1461]

I like this - kind of Australian sounding (apart from the voicing contrast). Not too keen on the orthography though: why /tʃ k/ <tgs c(h)>? Why not just <c k>? (And why the optional h? Ditto with <ps gs> for /f ʝ~ɣ/; why not just <f g>?

Also, what is /à á/?

Thanks for the input! The romanization is designed to have a direct letter-by-letter correspondence with the native orthography, and those are the digraphs/trigraphs therein. The native orthography is reasonably etymological, and an <h> (along with its native equivalent) is placed between <c> or <d> and a front vowel to indicate a lack of some as-yet undetermined historical palatalization of those consonats in those positions. /à á/ is a somewhat nicer looking way of denoting /˦ ˨/, the high and low tones, respectively.