Random phonology/phonemic inventory thread [2011–2018]

[Ambrisio]

I'm trying to create a phonology that feels "classical", but not too much like Greek, Latin or Sanskrit (or English). Any suggestions?

[Egerius]

My suggestion: Put koiné Greek and Old French into a blender, I'm curious.

[Click]

Add retroflexes and labiovelars to Greek. Graeco-Latin-Sanskrit is what you get.

[Ear of the Sphinx]

My suggestion: Put koiné Greek and Old French into a blender, I'm curious.

Hmm, has anybody tried a Finnish-like descendant for French?

[Avo]

What even is a "classical" sounding phonlogy? The phonology of Classical Tibetan? Classical Maya? Classical Tamil?

[Xing]

/p t ʦ ʨ k/ <p t c ĉ k>
/b d ʣ ʥ g/ <b d j ĵ g>
/m n/ <m n>
/l/ <l>
/s ɕ h/ <s ŝ h>
/z/ <z>
/ʋ/ <v>

/iː/ <ii>
/ɛː/ <aa>
/ɵː/ <uu>
/ɐ/ <i>
/ɒ/ <a>
/o/ <o>


CV(C)

vaanhid sukhunjuv ŝancik kitil miitan zavĵaanin
/ʋɛːnhɐd sokhonʣoʋ ɕɒnʦɐk kɐtɐl miːtɒn zɒʋʥɛːnɐn/

bancuun siznaccuik hikutaa lataahtiĉiz
/bɒnʦɵːn sɐznɒʦːoɐk hɐkotɛː lɒtɛːhtɐʨɐz/

[Ambrisio]

Here's a possibility:

/a i u ə a: i: u: ə: e: o: y:/ <a i u ǝ ā ī ū ǝ̄ ē ō ч>
<ē ō> derive from historical *ai and *au, as in Sanskrit or Modern French. <ч> derives from historical *iu.

/p t ts k b d dz g ph th tsh kh/
/m n l r j w s v ð z ɣ h/

The orthography is somewhat complicated. The aspirated stops /ph th tsh kh/ are written <p t c k> at the onset of a stressed syllable, and <p' t' c' k'> otherwise. The unaspirated stops are written <pp tt cc kk> at the onset of a stressed syllable, and <p t c k> otherwise. The voiced stops are written <b d j g>.
The non-stop consonants are written <m n l r y w s v dh z gh h>.

For example,
<kīra> - /khi:ra/ 'wife'
<hadhēla> - /haðe:la/ 'cloud'
<sk'īra> - /skhi:ra/ 'to marry'

There are 8 pronouns -- and no number distinction!
<aha> - 1 clusivity (could mean 'I' or 'we')
<itu> - 2 clusivity ('you')
<inna> - 1/2 clusivity (inclusive 'we')
<ahum> - 1/3 clusivity ('we')
<ittum> - 2/3 clusivity ('you')
<innum> - 1/2/3 clusivity ('we', 'all of us')
<uma> - 3 clusivity ('he', 'she', 'it', 'they')
<ēnum> - empty clusivity ('nobody')

The possessive prefixes: ha-, ttu-, na-, hun/hum-, ttun/ttum-, nun/num-, un/um-, ēn/ēm-
<hak'īra> - 'my wife' (c.f. Finnish vaimoni)
<ttuk'īra> - 'your wife' (c.f. vaimosi)

Cases are expressed with prefixes, not suffixes as in most languages:
ni-: accusative
ki-: dative
(no prefix): genitive
hō-: instrumental
lə-: essive
...
These case endings attach to the possessive prefixes rather than the pronouns themselves (except for the third person and empty pronouns, with become -ma and -mē).
niha - me
kiha - to me
nihum - us
ləma - like him
...

I personally like this phonology. It feels like a nice cross between Dravidian and Akkadian, with overtones of Tolkien's Primitive Elvish.

[QuantumWraith]

/iː/ <ii>
/ɛː/ <aa>
/ɵː/ <uu>
/ɐ/ <i>
/ɒ/ <a>
/o/ <o>

What's going on with these vowels?

[Shemtov]

Here's a random phonology that I promise not to make into another project. I call it Welshapanese:
For various reseons, I'm presenting this as a syllabary chart:
a ɨ ə i u e o
pa pɨ pə pi ɸu pe po
ba bɨ bə bi βu be bo
ta tɨ tə tɕi tθu te to
da dɨ də dʑi ðu de do
ka kɨ kə ki xu ke ko
ga gɨ gə gi gu ge go
sa sɨ sə ɕi su se so
ha hɨ hə hi he ho
ɬa ɬɨ ɬə ɬi ɬu ɬe ɬo
ɾa ɾɨ ɾə ɾi ɾu ɾe ɾo
ja jɨ jə ji ju je jo
wa
na nɨ nə ni nu ne no
ma mɨ mə mi mu me mo
m̩ n̩

Mutations:
After adjectives, Noun-initial stops become fricatives, except after high vowels.
After syllabic nasals, word intial voiceless stops become corresponding voiceless nasals.
Intial syllabic nasals may become desylaabicised after round vowels
Thus:

kaxum̩sa wam̩ to m̩daɾu ɾoju kam̩ɬi

Is realized as:

kaxum̩sa wam̩ n̥omdaɾu ɾoju xam̩ɬi

[Xing]

/iː/ <ii>
/ɛː/ <aa>
/ɵː/ <uu>
/ɐ/ <i>
/ɒ/ <a>
/o/ <o>

What's going on with these vowels?

It started with a regular, three-vowel system with contrasting length. Then it underwent a nice little vowel-shift

[Ceresz]

The LLL is awful at phonological realism so here's an attempt at it. Presumably would be set somewhere in Western Europe.

/(p) b t d k g kʷ w/ <(p) b t d k g q w>
/ts̻ s̻ ts̺ s̺/ <tz z ts s>
/mː m nː n/ <m mh nn n>
/lː l rː ɾ j h/ <l ll r rr j h>
/a e o i u/ <a e o i u>
/ai au ei eu oi ou/ <ai au ei eu oi ou>

Note that Old Irish, Proto-Basque, and to a lesser extent Old English all have a pervasive fortis-lenis contrast of the sort exemplified here. Presumably, since it's pre-IE pretty much by definition, fortis consonants can't appear word-initially. The phonology depends on the location, which I don't care about since I'm not actually going to do anything with this; move it to northern France or something and it could have /m ṽ/ instead of /mː m/, a dental fricative [possibly in place of the apical-laminal sibilant contrast], and palatals. Keep moving it northward and the fortis-lenis contrast would decay [more likely phonotactically -- word-initial fortis plosives/affricates -- than phonologically; Old Irish still had the fortis-lenis sonorant contrast] and there'd probably be pervasive palatalization instead.

I find this very pleasing.

[hubris_incalculable]

This is my current project, code named "Fantasy Language, Quite Anglo-Germanic" (FLQAG)

Code: Select all

Phonology
	Consonants
		/m n/ <m n>
		/pʰ p tʰ t kʰ k/ <p b t d k g>
		/f v θ ð s z ʦ ç~h/ <f v þ ð ſ(s_#) s z h>
		/ɾ~ɹ l/ <r l>
		/j w/ <j w>
	Vowels
		/aː~ɑː æ eː ɛ iː ɪ oː~ɒː ɔ uː ʊ/ <a æ é e í i ó o ú u> 
		/aɪː~aeː~aɛ aʊː oɪː~oeː~oɛ/ <aí aú oí>
	Syllable structure
		(C)(r|l|j|w)V(r|l)(C)

Though, when I was working out the syntax (link), it turned out to be less Anglo-Germanic than I thought it would be.

[Click]

This is Chóng. Guess which natlang I based the phonology on.

/m n ŋ/ m n ng
/p pʰ t tʰ (ts tsʰ) tʂ tʂʰ (tɕ tɕʰ) k kʰ ʔ/ b p d t z c zh ch zy cy g k x
/f (θ) s (ɕ) ʂ x/ f th s sy h
/ʋ l ɽ j/ v l r y

/i e a o u/ ı e a o u
/ie ia io ue ua uo/ ıe ıa ıo ue ua uo

/a˧ a˧¹ a˦˥ a˨˩ (a˩˦ a˦˩)/ a ā á à ă â

¹ Unaffected by tone sandhi, unlike a˧.

[Nemesis]

My new kriepruno language:

/m n ɲ/
/p t ʦ c k kʷ/
/f θ s ɕ x xʷ/
/j w ɾ ɫ ʎ̝̊/
/æ ɑ e̞ ə o̞ i ɨ u/

I wanted to make a very simple language, with no voicing distinction, but yet with a great range of places of articulation. I added a rare consonant, a voiceless (aveolo-)palatal lateral fricative. As for the vowels, I wanted to have a medium number of vowels evenly distributed across the vowel chart.

[Click]

Looks nice, but I would personally substitute /ɕ/ with /ç/ to make the inventory a bit more symmetric.

[Ambrisio]

This is Chóng. Guess which natlang I based the phonology on.

/m n ŋ/ m n ng
/p pʰ t tʰ (ts sʰ) tʂ tʂʰ (tɕ tɕʰ) k kʰ ʔ/ b p d t z c zh ch zy cy g k x
/f (θ) s (ɕ) ʂ x/ f th s sy h
/ʋ l ɽ j/ v l r y

/i e a o u/ ı e a o u
/ie ia io ue ua uo/ ıe ıa ıo ue ua uo

/a˧ a˧¹ a˦˥ a˨˩ (a˩˦ a˦˩)/ a ā á à ă â

¹ Unaffected by tone sandhi, unlike a˧.

Mandarin Chinese?

[Incorruptus]

This is Chóng. Guess which natlang I based the phonology on.

/m n ŋ/ m n ng
/p pʰ t tʰ (ts sʰ) tʂ tʂʰ (tɕ tɕʰ) k kʰ ʔ/ b p d t z c zh ch zy cy g k x
/f (θ) s (ɕ) ʂ x/ f th s sy h
/ʋ l ɽ j/ v l r y

/i e a o u/ ı e a o u
/ie ia io ue ua uo/ ıe ıa ıo ue ua uo

/a˧ a˧¹ a˦˥ a˨˩ (a˩˦ a˦˩)/ a ā á à ă â

¹ Unaffected by tone sandhi, unlike a˧.

Mandarin Chinese?

Yes...that or some very close variant.

[Click]

This is Chóng. Guess which natlang I based the phonology on.

[...]

Mandarin Chinese?

Correct.

[thetha]

Proto-Henaic:

S - /p t c k ʔ/
A - /pʰ tʰ cʰ kʰ/
F - /s x h/
L - /m n l r w j/

/i e ɛ a ə ɔ o u/

PH has a weakly tonal vowel system. Monosyllabic words can take the tones /à ā á/, and polysyllabic words either follow the form HLHLHL... or LHLHLH...
Initial clusters include s+L, S+s, S/A+w/j, L+w/j (besides stuff like /ww/ and /jj/ of course), and any combination A+S that doesn't involve two stops at the same POA. Syllable codas are everything except A, h, and r.

This is the ancestor of Taahu and Tuyuk. Official sound changes haven't been developed yet.

[Chagen]

I have no idea what this is for. But, uh, hypothetical proto-lang inventory, grammar is triliteral root style:

/m n ŋ/
/p b t d k g q/
/s~z ɬ~ɮ h~ɦ/
/tɬ dɮ/
/t' k'/
/j w/

/a i u/
/aː iː uː/

(C)(C)V(C)(C)

Phonemic gemination in between vowels and word-finally.

Stress (which matters more for the proto-lang) is a little complex:

All words can have two points of stress. First the stress "looks" for a long vowel in the word. If there is one, is automatically gives it stress. It then looks for another long vowel; that will also get stress, and the word is fully stressed.

If there is only one long vowel in the word, the next short vowel will immediately be stressed. A short vowel will never be stressed before a long vowel--in other words, when a long vowel takes stress, it blocks off ALL syllables before it from taking stress.

If there are no long vowels, the first syllable will automatically be stressed, then the second, unless it has the same vowel as the first, or it has two consonants one of which is a plosive, then it moves onto the next (if that other one also has the same vowel as the first, or two consonants with one plosive, it is stressed). Disyllabic words simply stress both syllables without question, unless the last syllable has a long vowel, in which it and only it will take stress.

The two stress positions cannot be more than two syllables apart. Even if it normally would violate the rules, a syllable can be stressed if it will prevent the two stress positions from being two-plus syllables apart. Thus, with some example words, long vowels being represented with doubling:

múú-si-gáá

gda-wáá-qá

qú-ŋdi-sá

saļļáá

Vowels not stressed become schwas and drop in the daughter, and thus a triliteral root system begins to form. The daughter lang's reflexes would be:

mūzgā

gdwāqa

quŋdza

sļā

I've made very little changes to this language, but I might want to say that I've added /e o/, with long variants.

There's also an archiphoneme that exists solely to explain why certain plural inflections act the way they do, it takes on the vowel quantity of the vowel right next to it.

Also: Verbal roots are always CCC. Nouns can be CCVC, CVCC, CVCVC, CVCV, CVC (rare), CV (also rare), VCCVC, VCVC, VCCV, and maybe a few others but that's all the basic kinds.

/w j/ cause some weird-as-hell samprasarana to happen when the tri-con craziness begins. They show up as /u: i:/ when next to a consonant, which can completely fuck everything sideways.

Unlike Semitic langs, vowels are very important at least when it comes to nouns--instead of vowel changes, the vowels a noun has stay but simply get shifted around.

I guess, say, from the root W-M-G, one could get the word wumūg, with a plural of āūmūg, but aint allowed so we get ūmūg? That looks nice, some little samprasarana right there.