Hmmm...

[TurkeySloth]

I have the my conlang's sound inventory formatted with /phoneme(s)/ or [allophone(s)] above related <orthography(ies)> (cf. /m/ [ break ] <m>). When I post the table here, do I need to put each entry in its own set of identifiers (cf. /p b/ [ break ] <p b> or /p/ /b/ [ break ] <p> <b> and [m~n~ŋ] [ break ] <ɱ> or [m] [n] [ŋ] [ break ] <ɱ>)?

Edit: Renamed to widen scope; previous name reflected above question about proper notation. Ironically, this was the pre-posted topic's original name.

[missals]

It doesn't matter.

A linguist writing a phonological description would put them in a table and the brackets wouldn't really be necessary anyways since they'd say something to the effect of "there are 15 phonemic consonants in _______, as well as one major allophone (fig. 2)" that would make it clear what the table was depicting

[zompist]

Yeah, just make it clear what you're presenting. If you say "The consonantal phonemes of Yangish are these:", then people know what they're looking at.

[TurkeySloth]

My dislike of geminate moraic fricatives brings up an interesting interaction involving words like nakev [ˈnɑ̟.kɛ̠ɮ], meaning military. Whereas words like the name Ohyn [ˈœ̠.hʊ̃] maintain the moraic, logically, words like nakev can't because of gaining an <o> as an adjective and the CV(m) syllable structure. Nakev's nominative adjectival form is nakeso [ˈnɑ̟.kɛ̠.ɕœ̠], currently. Would it be wiser of me to create a fourth ("historical") adjectival declension specifically for nouns ending in <v> [ɮ] or leave said nouns as a special subset of third declension? For context, the name Miɱ [mɪŋ] becomes Miɱo [ˈmɪŋ.œ̠] as a nominative adjective because [ŋ] must be moraic.

[Salmoneus]

I don't understand your question. I think you may be using the word 'moraic' idiosyncratically.

[TurkeySloth]

I knew I was forgetting something. Anyways, I'll use nakev/nakeso as examples of my meaning for moraic (@ as the mora marker).

Nakev: na—ke@v
Nakeso: na—ke—so

[Tropylium]

I'm not sure where there is anything geminate in here. Also, have you explained anywhere what your third declension in general looks like? (Certainly not in this thread so far.)

Taking a wild guess, you perhaps mean that (1) you want to retain a moraic v in your adjective for 'military', but (2) your adjective ending does not have an established consonant-initial variant (nakeso probably involves a mutation of v to s?), and (3) you do not want to apply gemination to gain nakevvo? If this is on the right track: yes, you could surely introduce a CV allomorph for the adjectival ending just for final /ɮ/ — historical explanations would be easy to propose, e.g. maybe /ɮ/ comes from earlier word-final /ld/, and when a vowel-initial suffix is added, the /d/ resurfaces, and so you'd end up with something like nakevdo. If this kind of a thing needs to be described as "a fourth declension" is purely a matter of taste however, since "Nth declension" is only a traditional Latin grammar term, not anything that has a strict modern linguistic definition.

[TurkeySloth]

Before standardization into the CV(m) structure, the word for "military" was nakesh, which because nakes after a merger of [s], [ʃ], and [ɕ]. They lateralied that to [ɬ] to avoid confusing the sounds before introducing <v> for the moraic fricative. Later, they merged it with [ɮ]. Now that I think about it, the smoother transition to the moraic fricative is introducing it and <v> simultaneously, which removes the need for lateralization and voicing as <v> would then represent moraic variants both [ɕ] and [ʑ]. For the record, historical geminates, such as Attamis, don't exhibit the lenition expected of plosive+<a> because the fricative would gemiante under those circumstances.

The consonant-initial formation comes from pairs like Ohyn/Ohynno, where the consonant geminates.

Ohyn: o—hy@n
Ohynno: o—hy@n—no

[TurkeySloth]

Currently, my conlang's base vowel system is /ɑ̟(ː) ɛ̠(ː) ə̹(ː) ɪ(ː) œ̠(ː) ʊ(ː)/, each of which has five possible tonal nasalizations bringing the total number of vowel phones to sixty (60). Does any natural language have [ə] or [ə̹] in stressed positions? I included [ə] because I may switch [ɑ̟] and [ə̹] to [ɒ̟] and [ə], respectively.

[Nortaneous]

Does any natural language have [ə] or [ə̹] in stressed positions?

yes

[Zaarin]

Does any natural language have [ə] or [ə̹] in stressed positions?

To elaborate on Nortaneous' post, some of the Modern South Arabian languages, French (insofar as French has stress), Catalan, Mandarin, Adyghe (and other languages of the Caucasus), and a number of Indo-Iranian languages to name a few. If nasalized counts, one may add quite a few of the Iroquoian languages.

[TurkeySloth]

Thanks.

As you may know from the CBB, my conlang, currently, spirantizes plosives before [ɑ̟] and [ʊ]. However, I'm changing the process regarding alveolar plosives to affrication to preserve distinct [t] and [d] in all phones relevant to *t and *d. While I know the causes may differ because of differing processes involved, do any causes have to remain consistent throughout both series (cf. [tə̹] → [t͡sə̹] while [d]+stress → [d͡z])?

[Nortaneous]

That's an unusual change. When affrication/spirantization is conditioned by specific vowels, it's usually either spread of the frication of the fricated vowel onto the consonant (as happened in some Sino-Tibetan and Bantu languages) or palatalization. In Tohono O'odham you have all back vowels triggering palatalization. In some Naga languages (a subgroup of ST), fricated vowels de-fricated after affricating labials, so in Angami you have /p/ [pf] before schwa and so on.

In Tocharian, PIE *d is reflected as /ts/ in most environments, and zero otherwise. It isn't clear what the "otherwise" entails - possibly *d > 0 before resonants, or before front vowels, or something. Tocharian had enough analogical leveling that there are difficulties in reconstruction. But this is an unusual sound change. (I think there's a tendency for *ejectives* to affricate, so maybe this is evidence for the glottalic theory...)

[mae]

-

[TurkeySloth]

That's an unusual change. When affrication/spirantization is conditioned by specific vowels, it's usually either spread of the frication of the fricated vowel onto the consonant (as happened in some Sino-Tibetan and Bantu languages) or palatalization. In Tohono O'odham you have all back vowels triggering palatalization. In some Naga languages (a subgroup of ST), fricated vowels de-fricated after affricating labials, so in Angami you have /p/ [pf] before schwa and so on.

In Tocharian, PIE *d is reflected as /ts/ in most environments, and zero otherwise. It isn't clear what the "otherwise" entails - possibly *d > 0 before resonants, or before front vowels, or something. Tocharian had enough analogical leveling that there are difficulties in reconstruction. But this is an unusual sound change. (I think there's a tendency for *ejectives* to affricate, so maybe this is evidence for the glottalic theory...)

Okay. Having read this, I'm very intrigued by fricative vowels, which seem like the logical route through which to get the changes I want (cf. [tʃ͡ə̹] → [t͡sə̹] and [kx͡a] → [xɑ̟]). However, how can I explain the bilabial/labiodental and velar affricates disappearing while the alveolar sibilants remain?

[Nortaneous]

In Tohono O'odham you have all back vowels triggering palatalization...

You mean high vowels, right?

oops. yes

Okay. Having read this, I'm very intrigued by fricative vowels, which seem like the logical route through which to get the changes I want (cf. [tʃ͡ə̹] → [t͡sə̹] and [kx͡a] → [xɑ̟]). However, how can I explain the bilabial/labiodental and velar affricates disappearing while the alveolar sibilants remain?

The problem is that fricative vowels are usually 'super-high' -- higher than the high vowels.

If your fricative vowel has coronal (rather than labial) frication, you might not have frication of the labials and velars to begin with -- see Mandarin and Nuosu. Nuosu also has 'retracted' (or as we say here, [+AAAAARRRGH]) fricated vowels, written yr ur, but I'm not sure where retraction comes from, or even if the contrast in y u vs. yr ur has the same origin as in the non-fricated vowels, where +/-AAAAARRRGH maps to low/high. The +AAAAARRRGH fricated vowels can only occur in the mid tone (or the falling tone, a variant of the mid tone), whereas the low vowels can mostly occur with any tone.

Coronally and labially fricated vowels are written /ɿ ʮ/ in the Sinological tradition. These symbols are not technically in IPA, but probably should be. (The problem is that they're more useful as *phonemic* symbols than as *phonetic* ones, but everyone uses IPA for phonemics now, so...)

Developing /ɿ/ isn't hard -- you could even develop /ɿ/ without /ʮ/. IIRC, there's a Mandarin dialect that had i ie > ɿ i. Developing a +/-AAAAARRRGH contrast on /ɿ/ is a little harder, and I'm not sure what the precedent would be for it (AFAIK the historical phonology of Nisoish is not well understood), but if you get that, /ɿ ɿ̠/ would affricate preceding coronal plosives, and wouldn't necessarily affricate other ones. You could then shift them to /ɨ ɑ/, but retain the coronal frication.

(It's been suggested that Japanese tu du > tsɯ dzɯ happened as a side effect of u > ɿ > ɯ. This would also explain the occasional reduction of mu > ɴ, which is really a placeless moraic nasal. The IPA symbol for a uvular nasal can be used for it because phonemic uvular nasals are vanishingly rare.)

There would be a few other side effects of this. If your language has consonant clusters, you might get reduction of sɿ > s in some environments. You'd also have to figure out what would happen to resonants preceding ɿ -- they'd probably become syllabic, but there are plenty of ways to get rid of syllabic resonants later. (See PIE.)

[TurkeySloth]

Okay. Having read this, I'm very intrigued by fricative vowels, which seem like the logical route through which to get the changes I want (cf. [tʃ͡ə̹] → [t͡sə̹] and [kx͡a] → [xɑ̟]). However, how can I explain the bilabial/labiodental and velar affricates disappearing while the alveolar sibilants remain?

The problem is that fricative vowels are usually 'super-high' -- higher than the high vowels.

If your fricative vowel has coronal (rather than labial) frication, you might not have frication of the labials and velars to begin with -- see Mandarin and Nuosu. Nuosu also has 'retracted' (or as we say here, [+AAAAARRRGH]) fricated vowels, written yr ur, but I'm not sure where retraction comes from, or even if the contrast in y u vs. yr ur has the same origin as in the non-fricated vowels, where +/-AAAAARRRGH maps to low/high. The +AAAAARRRGH fricated vowels can only occur in the mid tone (or the falling tone, a variant of the mid tone), whereas the low vowels can mostly occur with any tone.

Coronally and labially fricated vowels are written /ɿ ʮ/ in the Sinological tradition. These symbols are not technically in IPA, but probably should be. (The problem is that they're more useful as *phonemic* symbols than as *phonetic* ones, but everyone uses IPA for phonemics now, so...)

Developing /ɿ/ isn't hard -- you could even develop /ɿ/ without /ʮ/. IIRC, there's a Mandarin dialect that had i ie > ɿ i. Developing a +/-AAAAARRRGH contrast on /ɿ/ is a little harder, and I'm not sure what the precedent would be for it (AFAIK the historical phonology of Nisoish is not well understood), but if you get that, /ɿ ɿ̠/ would affricate preceding coronal plosives, and wouldn't necessarily affricate other ones. You could then shift them to /ɨ ɑ/, but retain the coronal frication.

(It's been suggested that Japanese tu du > tsɯ dzɯ happened as a side effect of u > ɿ > ɯ. This would also explain the occasional reduction of mu > ɴ, which is really a placeless moraic nasal. The IPA symbol for a uvular nasal can be used for it because phonemic uvular nasals are vanishingly rare.)

There would be a few other side effects of this. If your language has consonant clusters, you might get reduction of sɿ > s in some environments. You'd also have to figure out what would happen to resonants preceding ɿ -- they'd probably become syllabic, but there are plenty of ways to get rid of syllabic resonants later. (See PIE.)

Oof! I really need to read stuff more carefully before posting cause I missed the part about about shifting the vowels.

While I like the idea of having paired rounded and unrounded fricative vowels—not necessarily of the same height, I want the lang to sound decent without having triplicates of fricative and non-fricative vowels. Which two among /ʅ → ɪ/, /ʯ → ə̹/, and /ɿ → ɑ/ are most likely?

[Zaarin]

(I think there's a tendency for *ejectives* to affricate, so maybe this is evidence for the glottalic theory...)

I only know of this occurring to ejective fricatives (see Semitic). I've seen people cite this for implosives, but I'm not familiar enough with any language family with implosives to say myself.

[Nortaneous]

While I like the idea of having paired rounded and unrounded fricative vowels—not necessarily of the same height, I want the lang to sound decent without having triplicates of fricative and non-fricative vowels. Which two among /ʅ → ɪ/, /ʯ → ə̹/, and /ɿ → ɑ/ are most likely?

In languages with /ɿ/ as the only fricated vowel, it's usually analyzed as a high central vowel (Iau), but in Mandarin, where it can only appear after fricatives and affricates, it's often analyzed as the absence of a vowel altogether. In languages with both, frication can be seen as a *height* -- so /ɿ ʮ/ are higher than /i u/. But, given the phonetic realities, it's (probably) common for these vowels to become schwa, or some schwa-like central vowel. I don't know of any language where they just 'lower' to high vowels.

What I'm saying is that you could get *a vowel pharyngealization contrast* (or something along those lines) and develop /ɿ/ *and /ɿˤ/*, and then have them affricate some stuff and become /ə a/. (Where /a/ would pattern as a pharyngealized vowel.) Retroflexion doesn't have anything to do with it -- there are probably languages somewhere that contrast /ɿ ʅ/, but I don't think I've seen them.

If you want more solid precedent, you could see how Nuosu developed the vowels yr ur.

[Nerulent]

I only know of this occurring to ejective fricatives (see Semitic). I've seen people cite this for implosives, but I'm not familiar enough with any language family with implosives to say myself.

Osage reflects Siouan *t’ as ts’, and Wintu has qx’ but no q’ or qx, suggesting it affricated. Apparently Hadza has kx’ and kxʷ’ but no k’ (or t’), but these can vary from plain ejectives to ejective lateral affricates to ejective fricatives. Also apparently qx’ occurs allophonically in Caucasian languages.

@Nort, thanks for the interesting info on fricated vowels. Why are they more likely to become schwas?