The CBB

IIRC, the tendency in languages that have long diphthongs, it's the "most prominent" part of the diphthong is the one that is lengthened, e.g. [ai̯] vs. [a:i̯] (you can see this occur allophonically in Icelandic diphthongs, for example, but also in languages like Thai where the difference is phonemic). There do seem to be languages where the "less prominent" part of the diphthong is also lengthened, but this tends to occur along with lengthening of the most prominent part, e.g. [ai̯] vs. [a:i̯] vs. [a:i:] but rarely vs. [ai:] (this would tend to occur as the long counterpart to a rising diphthong [a̯i] instead, although I've only seen this on the Wikipedia article for Swabian German, so I don't know whether this is actually what this is. I've seen [ɐ̯i] and [ə̯i] reported for various Portuguese and Croatian dialects).

Today, I thought about a language where for every segment there is only one syllable it can occur in, i.e. there are no two syllable that share a segment. A toy example would be a language that has /p t k/ and /a i u/: this would only allow the syllables /pa ti ku/. Of course one can use much more segments and maybe than one could say that the language does not have segments in some way?

A Southwest Wanian language:
/p t k /
/m n ɲ/
/s ʃ x h/
/l/
/ɹ w j/

/i e u o ɐ æ ɒ/
/ĩ ẽ ũ õ ɐ̃ æ̃ ɒ̃/

/˥ ˧ ˩/

Creyeditor wrote: ↑14 Jun 2018 21:51 Today, I thought about a language where for every segment there is only one syllable it can occur in, i.e. there are no two syllable that share a segment. A toy example would be a language that has /p t k/ and /a i u/: this would only allow the syllables /pa ti ku/. Of course one can use much more segments and maybe than one could say that the language does not have segments in some way?

Huh, this was much trickier than I expected, and I kind of cheated by using nasality and secondary articulations:

nɨ̃ ɲĩ ŋɯ̃ ɴã
tɨ c͜çi kɯ qa
nʷʉ̃ ɲʷỹ ŋʷũ ɴɒ̃
tʷʉ c͜çʷy kʷu ɴɒ

Not a very pretty inventory. And still too small. Not sure if anything like this could exist in a natural language - it seems that, even with languages with many phonological restrictions and patterns of harmony and whatnot, a fundamental feature of human language is multiple possibilities for combining a specific segment with other segments.

Here's a vowel inventory I came up with for palatal harmony. It is based on a four vowel system like Nahuatl, but with a front-back distinction.

/æ ɑ e ɤ i ɯ ø o/

/p b t d ṯ ḏ c ɟ k q ɢ ʔ/ <p b t d tš dž tj dj c k g q> p b t d ţ ḑ ť ď k q g ʾ
/f v s z ʃ ʒ ç ʝ x ɣ χ ʁ ħ ʕ h ɦ/ <f v s z š ž sj zj ch ğ kh gh ħ ḥ h ḫ> f v s z š ž ś ź x ǥ ḫ ƣ ḥ ʿ h ĥ
/m n ṉ ɲ ŋ ɴ/ <m n nš nj nc nk> m n ņ ň ŋ ṅ
/ɾ ɾ̝/ <rd řd> r ʀ
/r r̝ ɺ͝r/ <r ř lr> ɍ ř ḻ
/l ʋ j jʱ/ <l w j jh> l v j ȷ̈

/i y ɨ ʉ ɯ u ɪ ʏ ᵻ ᵾ ɯ̽ ʊ e ø ɘ ɵ ɤ o ɛ œ ɜ ɞ ʌ ɔ æ œ̈ ɜ̈ ɞ̈ ʌ̈ ɔ̈ a ɶ ɐ ɶ̈ ɑ ɒ/ <i y ï ü ı u ie yø ïë üö ıə uo e ø ë ö ə o ea øę ëä öą əâ oå ae ęø ëä ąö âə åo a ę ä ą â å> i ü î û ï u ì ȕ ȋ ȗ ȉ ù e ö ê ô ë o è ȍ ȇ ȏ ȅ ò à ȁ ȃ y̑ ʉ ɵ a ä â ŷ å ơ
/a aː aˁ aːˁ/ <a ah aḥ ahḥ> a ā ą ą̄
First one is more multi OS. Second one would be official romanization.
also
/p t k ts m n f s h j r i e u o a/
/aː/
p t k c m n f s h j r i e u o a
ā
/p t k ʔ m n ŋ ʦ ʧ ʨ f s ʃ ɕ x h ᵐb ⁿd ᵑg l j w r i y ɯ u ɨ e ɤ o œ ə a/ p t k : m n ng c č q f s š x gh h mb nd ngg l j w r i û ū u ī e ō o ē ā a
In loanwords /b d g v/ b d g v
/p pʼ t tʼ k kʼ ɓ ɗ ɠ m n ŋ v ð z ɣ ɦ l j ɰ ɥ w r ɮ ð͡l i ɨ u e ɤ o ə aː ɑ ɒ/ p p̓ t t̓ k k̓ b d g m n ŋ v ð z ƣ h l ĭ ẅ ÿ ŭ r ƚ ⱡ i ɨ u e ă o ə a ɑ ɒ
Languages of Dhott Islands
Plymwþ
/p b t d c ɟ k g q ɢ ʡ/ p b t d ǩ ǧ k g ḳ ġ ȝ پ ب ت د ګ ڰ ک گ ق ݠ ع
/m n ŋ ɲ/ m n ŋ ñ م ن ڠ ڽ
/f v s z ʃ ʒ ç ʝ x ɣ χ ʁ ħ ʕ h ɦ ɧ/ f ƀ s z š ž ç j ĥ ĝ ḫ ƣ ḣ ǥ h ħ ꜧ ف ڤ س ز ش ژ ݜ ڙ خ غ خّ غّ ح ݞ ه ݟ ݽ
/ts dz tʃ dʒ/ c ʒ č ǯ څ ڄ چ ج
/l ɫ j w/ l ḷ j ů ل ݪ ي و
/r r̝ ʀ/ r r̃ γ ر ݛ ݝ
/ɬ ɮ ɭ̊˔ ɭ˔ ʎ̝̊ ʎ̝ ʟ̝̊ ʟ̝/ θ δ ϛ ς ψ λ φ ϣ ص ظ ض ط ڞ ڟ ٿ ڷ

/iː ɨː eː æ/ ī ȳ ē ǣ إِي إٍي إٖي أٙا
/ɯː uː ɤː oː ɑː/ w̄ ū v̄ ō ā أٌو أُو أٗا أٗو آ

/i ɨ e æ/ i y e æ إِ إٍ إٖ أٙ
/ɯ u ɤ o ɑ/ w u v o a أٌ أُ أّٗ أٗ أَ

Dhott Islandic
/p pʰ b t tʰ d ʈ ʈʰ ɖ c cʰ ɟ k kʰ g/ प फ ब त थ द ट ठ ड च छ ज क ख ग
/ɓ ɗ ᶑ ʄ ɠ/ भ ध ढ झ घ
/m mʱ n nʱ ɳ ɳʱ ɲ ɲʱ ŋ ŋʱ/ म म़ न ऩ ण ण़ ञ ञ़ ङ ङ़
/ɸ β f v s z θ ð ʂ ʐ ʃ ʒ x ɣ χ ʁ ħ ʕ ɦ/ प़ ब़ फ़ भ़ स ज़ थ़ ध़ ष ष़ श श़ क़ ग़ ख़ घ़ ह़ अ़ ह
/l ɭ ʎ/ ल ळ य़
/ʋ ɹ j ɥ/ व ॹ य ॺ
/ɾ ɽ/ र ड़
/r ɽr/ ऱ ढ़

/i iː e eː æ~ɛ æː~ɛː/ इ/कि ई/की ऎ/कॆ ए/के ऍ/कॅ ऐ/कै
/u uː o oː ʌ ʌː ɔ ɔː ə aː/ उ/कु ऊ/कू ऒ/कॊ ओ/को ॳ/कऺ ॴ/कऻ ऑ/कॉ औ/कौ अ/क आ/का

/p t̪ (t) ʈ c k k͡p/
/m n̪~n ɳ ɲ ŋ ŋ͡m/
/w l ɻ j/
/r/

The alveolar stop and dental nasal are restricted in distribution; the dental nasal only occurs as the word-initial equivalent of the alveolar nasal while the alveolar stop only occurs intervocalically as the geminated counterpart of /r/. Additionally the apical consonants (alveolar and retroflex) are restricted to word-internal position.

/i u/
/a/

Syllable structure is Strictly (C)CV, where clusters only occurs interocalically and are restricted to geminate stops/nasals, nasal + homoorganic stop and lateral + stop.

Stress is rigidly word-initial.

/p t t͡ʃ kʷ q ʔ/
/s ç xʷ χ h/
/m n j w ʁ/
/m̥ n̥/

The voiceless fricatives other than /s h/ pattern with the voiced sonorants in the same way the voiceless nasals pattern with the voiced ones.

/i ɨ u/
/ɛ a ɔ/

Syllable structure is CV(C), where the only permissible final consonants are /s h/ and the voiced sonorants. h + voiced sonorant & voiced sonorant + h become voiceless sonorant/fricatives.

---

/b t t͡s k ʔ/
/t’ k’/
/f ç h/
/m n ɾ ɲ ŋ/
/mˀ nˀ/

The glottalisation contrast is restricted to word-initial position, and does not occur in affixes.

/i ɨ u/
/(e) ə o/
/a/

/ə/ is realised as /e/ and /ɨ/ is merged with /i/ when preceding palatal consonants.

Syllable structure is CV(C), where final consonants are restricted to /h ŋ/ and gemination of an immediately following consonant and /ŋ/ assimilates to the POA of a following consonant.

/m n w/
/p t k/
/ɸ s x/
/a i u/

/w/ is [ŋ] in coda

DesEsseintes wrote: ↑30 Jun 2018 16:19 /m n w/
/p t k/
/ɸ s x/
/a i u/

/w/ is [ŋ] in coda

That w > ŋ change reminds me of some similar changes that happened in various Italian, Romanch, and Francoprovencal varieties where it sometimes went all the way to [g] or [k]. Also coda j > k. I also read a thing by Robert Blust recently where he talked about these kinds of fortitions being especially characteristic of esoterogenous/inward-oriented languages spoken by smaller, insular communities, like groups in New Guinea or in isolated Alpine valleys.

Porphyrogenitos wrote: ↑01 Jul 2018 07:08 That w > ŋ change reminds me of some similar changes that happened in various Italian, Romanch, and Francoprovencal varieties where it sometimes went all the way to [g] or [k]. Also coda j > k. I also read a thing by Robert Blust recently where he talked about these kinds of fortitions being especially characteristic of esoterogenous/inward-oriented languages spoken by smaller, insular communities, like groups in New Guinea or in isolated Alpine valleys.

That's interesting to know. I might have a use for this in a project currently.

I thought about a tone languages, wich has high, low and underspecified moras, underlyingly. Then a high tone would spread to the right, as long as the target mora is toneless. This would happen inside a word and to the first syllable of the next word. The default tone for underlyingly toneless syllables is low. Also there is downstep between underlying high tones.

/lághába ngamòna/ -> [lá↓ghábá ngámònò]
/ròbèmaú juujanák/ -> [ròbèmaú júùjànâk]
/wunúcak mèkína/ -> [wùnúcák mèkíná]
/aù nu´k maa/ -> [àù nǔk máá]
/gé manágha`m/-> [gé má↓nághâm]

Porphyrogenitos wrote: ↑01 Jul 2018 07:08

DesEsseintes wrote: ↑30 Jun 2018 16:19 /m n w/
/p t k/
/ɸ s x/
/a i u/

/w/ is [ŋ] in coda

That w > ŋ change reminds me of some similar changes that happened in various Italian, Romanch, and Francoprovencal varieties where it sometimes went all the way to [g] or [k]. Also coda j > k. I also read a thing by Robert Blust recently where he talked about these kinds of fortitions being especially characteristic of esoterogenous/inward-oriented languages spoken by smaller, insular communities, like groups in New Guinea or in isolated Alpine valleys.

That's quite interesting. I am thinking of applying it to my romlang as well. How exactly might the chain work tho?

DesEsseintes wrote: ↑30 Jun 2018 16:19 /m n w/
/p t k/
/ɸ s x/
/a i u/

/w/ is [ŋ] in coda

I thought this allophony might maybe be explainable historically, if we assume a proto-phoneme *g. It became [ŋ] in codas (which is a very natural sound change) and [w] otherwise (also natural). I wonder what became of *b and *d.

Some minimalistic languages
/t k n s x j l i ɯ ɤ æ a/ t k n s x j l i u o e a
/p b t k g ɢ ʡ ʔ f θ s x ɣ h ɦ ɧ m n ɲ ŋ j w ɰ ɥ l ɾ i iː y yː ʉ ʉː ɯ ɯː u uː e eː ø øː ɵ ɵː ɤ ɤː o oː ɛ ɛː œ œː ɞ ɞː ʌ ʌː ɔ ɔː æ æː a aː ɑ ɑː ɒ ɒː/ p b t k g ḡ ʿ ʾ f ǧ s x ƣ h ɦ ꜧ m n ɲ ŋ j w ẅ ÿ l r i ī ü ǖ y ȳ ǫ ǭ u ū e ē ö ȫ ȯ ȱ ė ė̄ o ō ĕ ē̆ ö̆ ȫ̆ ȯ̆ ȱ̆ ė̆ ė̄̆ ŏ ō̆ æ ǣ a ā ą ą̄ å å̄
/p t kʷ ʔ f ɕ xʷ m n ŋʷ l j w̹ i u e o ə ɑ/ p t k q f s h m n g l ĭ ŭ i u e o w a
/p t ʈ c k ʔ ɸ s ʂ ç x h m n ɳ ɲ ŋ j w l r ɮ i u ə a/ p t ⱦ c k ' f s ꞩ ş x h m n ꞥ ņ ŋ j w l r ꝣ i u e a

I can see the minimalism!

Classical Momucharusumuko, native name Momčalsumai, a dead language spoken in the Momčalsum Empire in what is now Southern Fuhe. As the Momčalsum Empire wrote about Ki (which they called Ruwah) indepently from the Proto-Fuheans during the Ki Wars, the Fuheans preserved their books in the original language (as they wrote about "Phonetics and Ruwah"), with a "Fuhekoan Reading". This is the phonology as actually read by Momčalsum people:
/p pʰ b t tʰ d t͡ʃ t͡ʃʰ d͡ʒ k kʰ k͡p k͡pʰ/ < p ph b t tʰ d č čh ǰ k kh q qh>
/m n/ <m n>
/s z h ɦ/ <s z h hh>
/ɾ/ <r>
/ʙ r / <v ŕ>
/l w j/ <l w j>

/i u a ai au/ <i u a ai au>
Marginal syllabics, only occuring in verbs and nominalized verbs:
/e ei eu o oi ou ə əi əu m̩ n̩̩ ʙ̩ r̩ / <e ei eu o oi ou y yi yu m n v ŕ>

Phonotactics: Affixes containing an aspirate are deaspairated if the closest root suffix is also an aspirate. /ei oi eu ou əi əu ai au/ become [ej oj ew ow aj aw əw] if followed by a vowel.
(C)(C)V(C)(C)

Permitted initial sequences :
/bd db zb zd sp spʰ st stʰ sk skʰ ps ks bz pl pʰl bl kl kʰl pr pʰr br kr kʰr tʙ tʰʙ dʙ kʙ kʰʙ sl zl / and any stop and nasal followed by /j w/
Permitted final sequences:
/sp st sk zp zd rp rb rt rd rk lp lb lt ld lk ls lz mp mb nt nd nk [ŋk] ps ks bz/

Something short and sweet: I suppose one could add ejectives at some PoAs, and replace /p/ with /b/ (as is all the rage), but for now I’ll leave it like that.

I like having both of /θ q/ in a 13-consonant inventory.

Oh vowels! Uhm, /a i o/?

/p t~ɾ t͡ʃ k/
/tʰ kʰ h/
/p’~ɓ t’~ʔ t͡ʃ’ k’ (ʔ)/
/f~v s~z/
/m n ʁ/

The ejective alveolar is realised as a simple glottal stop in the onset of unstressed syllables by many speakers; in the other sounds with /~/ the sound on the right is the intervocalic realisation and the one on the left the elsewhere case. The glottal stop could be argued to be a separate phoneme in addition to this allophony, since in coda position it represents a historic neutralisation of several ejective stops.

/i i u/
/iə uə/
/ɛ ɛ̃ œ ɔ ɔ̃/
/a/

Syllable structure is CV(F), where F is restricted to one of /ʔ h ̃ ʁ/. Of these, the nasal feature only occurs with the nasal vowels (and these vowels do not occur before other coda consonants), While the voiced uvular fricative is restricted to occurring after /iə uə œ/.

Finally finished with Pigeonese phonology
/p b ɓ t d ʈ ɖ c ɟ k g ɢ ʔ/ p b ꞗ t d ŧ đ ꞓ ɟ k g ȝ ɂ
/m n ɳ ɲ ŋ ɴ/ m n ꞥ ɲ ŋ ṅ
/ts dz ʈʂ ɖʐ ʧ ʤ cç ɟʝ kx gɣ ɢʁ/ c ȷ ȼ ɉ č ʒ ǩ ǧ ḵ ḡ g̈
/ɸ β f v θ ð s z ʂ ʐ ʃ ʒ ç ʝ x ɣ χ ħ ʕ ʢ h ɦ ɧ/ ꞙ ꞵ f v ϑ δ s z ꞩ ƶ š ž ɏ j ꝁ ǥ ꭓ ḥ ḩ ꜧ h ɦ ɥ
/ʋ w ɰ j ɥ/ ʋ (written as p or b after voiceless or voiced consonants respectively, and after vowels, ŭ) ƣ y/ĭ ŷ
/r ɾ ɽ ɽr ɻ/ ṙ r ɍ ɽ ṟ
/l ɫ ɭ ʎ/ l ɫ ƚ ȳ
/ɬ ɮ ꞎ ɭ˔ ʎ̝̊ ʎ̝/ ɬ ꝇ ⱡ ŀ ļ ľ

/i y e ø ɛ æ/ i ü é ö è æ
/ɨ ʉ ɪ̈ ʊ̈ ɘ ɵ ə ɵ̞ ɜ ɞ/ î û ị ụ ȩ ɵ e ẹ ĕ ă
/ɯ u ɯ̽ ʊ ɤ o ʌ ɔ ɑ/ ư u ṳ ů ơ o ə ŏ a
/ɑː ɑ̃ ɑ̃ː/ ā ą ą̄
/ɪ̈̃ ʊ̈̃ ɘ̃ ɵ̞̃ ɯ̽̃/ are written as: ĩ ũ ẻ ở ử
The letters q, w, and x are used in loanwords only. Standing for /kʷ~q ɣʷ ks/
Same with
/qʷ ʀ r̃ w kʃ ʨ ɕ ʑ ʥ/ ꝗ ꝛ ꝝ ꝡ x̌ ć ś ź ǰ
Now working in this too
/p t k/ <p t k> p t k
/t͡s/ <ts> c
/f θ s h/ <f c s h> f j s h
/w l ʎ j/ <w l lj j> w l ł y
/r/ <r> r
/ɬ/ <ls> z
/m n ɲ ŋ/ <m n nj g> m n q g

/pː tː kː/ <pp tt kk> pp tt kk
/t͡sː/ <tts> cc
/fː θː sː hː/ <ff cc ss hh> v x ss ħ
/wː lː ʎː jː/ <ww ll llj jj> wh ll lh ji/jï
/rː/ <rr> rr
/ɬː/ <lls> zz
/m n ɲ ŋ/ <mm nn nnj gg> mm nn ñ ŋ

/i u ʊ e o ʌ a/ <i u ụ e o ọ a> i u ü e o ö a

/iː uː ʊː eː oː ʌː aː/ <ii uu ụụ ee oo ọọ aa> í ú ű é ó ő á

/a˩ a˧ a˥ aˀ˥ a̰˩ a˩˥ a˥˩/ <a¹ a a² a³ a⁴ a⁵ a⁶> aɩ a aз aƨ aч aƽ aƅ
/ˈa˩ ˈa˧ ˈa˥ ˈaˀ˥ ˈa̰˩ ˈa˩˥ ˈa˥˩/ <á¹ á á² á³ á⁴ á⁵ á⁶> Stressed written with ː after

/puː˥˩ˈʦːi˥ko˩ŋːi˧ ŋːḭŋ˩kiŋ˥˩ˈɬʊː˩ taˀm˥ˈɬo˥˩ kθuj˥˩ lʌː˥/

<Puu⁶ttsí²ko¹ggi ggi⁴gki⁶glsụ́ụ́¹ ta³mlsó⁶ kcu⁶j lọọ².>
Púƅcciзːkoɩŋi Ŋiчgkigƅzűɩː taƨmzoƅː kjuƅy lőз.
The second doesn’t have too many digraphs, so they like it better.
That’s actually a protolanguage.
The evolution is like this
Geminated fricatives and stops become voiced. Tones simply drop out and become finals.
So like this one so far
/p b t d ʈ ɖ c ɟ k g q ʡ ʔ/ p b t d ƭ ɗ ƙ ɠ k g q ƣ ʾ
/f v θ ð s z ʂ ʐ ɕ ʑ ç ʝ x ɣ χ ʁ ħ ʕ ʜ ʢ h ɦ ɧ x͝f ɣ͝v/ f v θ δ s z ȿ ɀ ṡ ż ʃ ȥ x γ ẋ ǥ ḥ ʿ ḫ ḣ h ḧ ḩ ƕ ƿ
/m̥ m n̥ n ɳ̊ ɳ ɲ̊ ɲ ŋ̊ ŋ ɴ̥ ɴ/ ṃ m ṇ n ṇ́ ń ɲ̧̇ ɲ ŋ̇ ŋ ṁ ṅ
/l̥ l w̥ w j̊ j ɰ̊ ɰ ɥ̊ ɥ/ ḷ l ẉ w ỵ y ẇ ẅ ẏ ÿ
/ʦ ʣ ʨ ʥ ʧ ʤ t͝ɬ d͝ɮ/ c j ĉ ĵ č ǰ ċ ȷ̈
/ɬ ɮ ꞎ ɭ˔ ʎ̝̊ ʎ̝ ʟ̝̊ ʟ̝/ ľ ł l̈ r̈ ḹ ḽ ḻ ẕ
/r̥ r ʀ̥ ʀ ᴙ̥ ᴙ/ ṛ r γ̇ ҕ ṝ r̄

/iː yː ʉː ɯː uː ɪ ʏ ᵾ ω ʊ eː øː ɵː ɤː oː ɛ œ ɞ ʌ ɔ ə æ æː aː ɒ ɒː/ ī ȳ ʉ̄ ɯ̄ ū i y ʉ ɯ u ē ø̄ ɵ̄ ǝ̄ ō e ø ɵ ǝ o a æ ǣ ā ɔ ɔ̄
The capital of ǝ is Ǝ.
Here’s another descendent
/p t c k ʔ/ <p t c k q> p t q k ʔ
/f s ʃ x ħ h ɦ/ <f s š x ḥ h g> f s ś x ħ h ɦ
/m n ṉʲ ɲ ŋ ŋʷ/ <m n nj ň ŋ ŋw> m n ń ñ ŋ ŋ̊
/w ɥ j/ <w jw j> w ɥ j
/l ɫ/ <l ł> l ɫ
/ɽ/ <d> ɽ
/r/ <r> r
/pʼ tʼ cʼ kʼ t͡sʼ t͡ʃʼ/ <p' t' c' k' s' š'> ṗ ṭ q̇ ḳ c̣ č̣

/i iː y yː ɯ ɯː u uː ʊ ʊː eː ø oː ɛ ʌ̃ ɔ ɔ̃ æ̃ a ã/ <i ii y yy ı ıı u uu ù ùù ee ø oo e į o ǫ ę a ą> i ī y ȳ ȯ ȱ u ū ư ư̄ ē ȫ ō e ö o ŏ ĕ a ă