zompist bboard

At some point, Amqoli or its protolang shifted /k/ to /q/ before back vowels, and possibly to /tS/ before front vowels. If protolang, it also did it to g > R / dZ (or dz? or Z? or j?). This is because /q/ is common and /k/ is rare.

Presumably, it got new /k/ from /kr/ or /kl/ or somesuch. Medial clusters in Amqoli tend to have the liquid first and then the plosive, so maybe it got rid of some others that way too.

_________________
Siöö jandeng raiglin zåbei tandiüłåd;
nää džunnfin kukuch vklaivei sivei tåd.
Chei. Chei. Chei. Chei. Chei. Chei. Chei.

Nort, I ran across an interesting source on a dialect of the Naxi language you might want to read. The language has a byzantine vowel inventory, including retroflex consonants. It reminded me of a couple of your phonologies.

https://en.wikipedia.org/wiki/Naxi_language

http://www.persee.fr/web/revues/home/pr ... _35_1_1745

Yeah, Naxi phonology is a mess -- consonants that are only phonemic before one vowel, etc. As a general principle, weird shit happens in the Sichuan/Yunnan/Tibet region. Polysynthetic Sino-Tibetan languages! Ergativity! Epiglottals! Eight stop POAs! Syllabic bilabial trills! Retroflex secondary articulation on peripheral plosives!

_________________
Siöö jandeng raiglin zåbei tandiüłåd;
nää džunnfin kukuch vklaivei sivei tåd.
Chei. Chei. Chei. Chei. Chei. Chei. Chei.

not to mention the tibetan disease of having five evidentiality distinctions plus a conjunct/disjunct oppositions and finding this perfectly reasonable

wh

_________________
Siöö jandeng raiglin zåbei tandiüłåd;
nää džunnfin kukuch vklaivei sivei tåd.
Chei. Chei. Chei. Chei. Chei. Chei. Chei.

I was just adding to the list of weird shit around the tibetan plateau.

Tibetan proper is kinda the epicenter of a zone of silly evidential/egophoricity systems, which are totally worth ripping off.

_________________
Siöö jandeng raiglin zåbei tandiüłåd;
nää džunnfin kukuch vklaivei sivei tåd.
Chei. Chei. Chei. Chei. Chei. Chei. Chei.

_________________
Blog:
Conlangs: | | |

I'd been going back and forth on whether to include this in Kannow, but it turns out that Kiranti and Rgyalrongic do it, so:

Nominalizations of transitive verbs can take possessive affixes coreferent with the patient.

Athpare:
a-ka-pik

"the one who speaks to me"

Japhug:
a-kw-fstwn

"the one who serves me"

Kannow:
idyqʼymbuw
i-daqʼ-mb-ow

"the one who speaks to me"

cf. idyqʼuw "the one who speaks", halaatʼámbow "my tablet(s) [DEF]" / halaatʼow "the tablet(s)" -- Kannow only distinguishes number in the indefinite

_________________
Siöö jandeng raiglin zåbei tandiüłåd;
nää džunnfin kukuch vklaivei sivei tåd.
Chei. Chei. Chei. Chei. Chei. Chei. Chei.

PK had initial fricative+nasal clusters. For the root *xɲapeɓɨ:

Amqoli and Rengni shift implosives to nasals and denasalize backer-than-alveolar nasals, so xdʒapem-. Amqoli simplifies the cluster by devoicing and dropping the fricative, so tʃapem- > tʃpem > tʃhem > chxem. Rengni runs its voice assimilation in the opposite direction and doesn't simplify: xdʒapemu > ɣdʒâpmu > ɣdʒâuʔm. Deghuri is boring: naapiɓa.

Also, nasal+approximant clusters: ml mr mlʲ mrʲ mj.

In AR, mlʲ mrʲ > mj. In D, ml mr mlʲ mrʲ mj > n n ɲ ɲ ɲ. Then, in A and R separately:

A: ml mr mj > mn mn mj, then mj > ɲw > dʒw > dʒb > bdʒ ( > bʒ).
R: ml mr mj > mbl mbr ɲw > bʙl bʙr dʒw > v v dzv.

So, cognates:
tʃʰem - naapiiɓa - ɣdzɔ̂ʔm (xɲapéɓɨ)
bʒaq - naaga - dzvɔg (mrʲágɨ)
mnelʲa - naɗaa - vuɲa (mlɨɗá)

Also Rengni simplifies diphthongs, fuck diphthongs

_________________
Siöö jandeng raiglin zåbei tandiüłåd;
nää džunnfin kukuch vklaivei sivei tåd.
Chei. Chei. Chei. Chei. Chei. Chei. Chei.

So, we know PK has:
ml mr mlʲ mrʲ mj
xm xn xɲ

And the lexicon has:
bj mb
But bj is bʲ.

Presumably there were sN clusters, since xN:
ml mr mʲlʲ mʲrʲ mj
xm xmʲ xn xɲ
sm s(ʲ?)mʲ sn sʲɲ

mb is unnecessary; better to make that nw. Let's say PK had a sonority hierarchy. It's needed to get A initial mb- in clusters (mbaɗé > mbrʲe), but if nw, A has w > b, so nw > nb > mb. nwaɗé > nwarʲé > nbarʲé > mbarʲé > mbrʲé.

That would imply lw rw, but no need for lʲw rʲw. Those would give lb- rb- in A, and those clusters exist.

Could maybe analyze palatalized labials as clusters then, as has already been done for mj. But palatals can occur word-finally, so can't be. But what about palatalized liquids? They can't occur syllable-finally in Amqoli, but there's no /j/ and ly ry contrast with lzh rzh so they have to be analyzed as units. Could be neutralization as in Polish.

So, initial clusters:
pj bj mj (lj rj)
sm s(ʲ?)mʲ sn sʲɲ
xm xmʲ xn xɲ
ml mr (mlʲ mrʲ) mj
nw lw rw
And then that last row implies more Cw clusters.
tw tsw tʃw kw qw
dw dzw dʒw gw ʁw
sw ʃw xw (hw)
zw ʒw (ɣw)

What happens to these?

A:
pʃ bʒ bʒ lʲ rʲ
m? bʒ? l? tʃ
m? bʒ? l? tʃ
mn mn (bʒ bʒ) bʒ
mb lb rb
th tsh tʃh kh qh
bd b(d)z b(d)ʒ bg b
sp? ʃp? xp? ?
zb? ʒb? b?

Well, go back to AR. We know the fricative clusters have to be there.
AR:
pj bj mj lj rj
sm [s/ʃ]mj sn [s/ʃ]dʒ
ml mr (mj mj) mj
nw lw rw
tw etc.

But: xw > f, f > h in A.

_________________
Siöö jandeng raiglin zåbei tandiüłåd;
nää džunnfin kukuch vklaivei sivei tåd.
Chei. Chei. Chei. Chei. Chei. Chei. Chei.

Two competing strategies for marking repeated action:

PK *-rwe = an aspect marker of some sort
reduplication of onset and vowel of first syllable (with accent transfer leftward)

so: *geli > *gélirwe / *gégeli > gelirbi / gegli

in Amqoli, these converge on the same meaning, with *-rwe used for some classes of verb (verbs with no onset, recent loans, verbs derived from nouns with directionals) and reduplication used for most native verbs

_________________
Siöö jandeng raiglin zåbei tandiüłåd;
nää džunnfin kukuch vklaivei sivei tåd.
Chei. Chei. Chei. Chei. Chei. Chei. Chei.

_________________
Siöö jandeng raiglin zåbei tandiüłåd;
nää džunnfin kukuch vklaivei sivei tåd.
Chei. Chei. Chei. Chei. Chei. Chei. Chei.

in which the gloss for the vengic language contains no person marking and the gloss for the hluic language does

_________________
Siöö jandeng raiglin zåbei tandiüłåd;
nää džunnfin kukuch vklaivei sivei tåd.
Chei. Chei. Chei. Chei. Chei. Chei. Chei.

related what vowel does <z> raise to where Zzyxwqnp verbs have it in the root
or does it behave like <y v>

There are a few words with invariant /ə/ (from words ending in V'eng -aʔ -ar), but it usually raises to /i/.

_________________
Siöö jandeng raiglin zåbei tandiüłåd;
nää džunnfin kukuch vklaivei sivei tåd.
Chei. Chei. Chei. Chei. Chei. Chei. Chei.

Some Hlu cognates:

*ʔbæɲ > *ɓein > ɓoin
*mɣa > *mbɣa > mbu
*tteŋ > *nteŋ > ntʰiŋ
*ŋgas > ŋgaɬ
*kke > *ŋke > ŋkʰi
*sæ̃ > *ʃã > sa

Note that *æ *e become e i, but *æ̃ becomes ja.

More: (beginning with Proto-Yia)

*karaŋa > *kraŋ > kaŋ
*muŋa > *mŋa > ŋɔ
*reɣa > *rɣa > rɯ
*tuNo > *tro > to
*tezi (> *tzi) > *ti > ʄi
*tego > *tɣo > ɗo

Note that /ɣ/ was lost before prevocalic voiceless plosives became implosives.

V'eng cognates are k-rang, mung, reg, t-myö, të, teg; Vian descendants are ná, mur (syllabic m), lê, njù, chi, tè.

nb: *N is a symbol for something that becomes /mj/ in Vengic after a rounded vowel and /dr/ otherwise; the cognate in V'eng is t-myö. In Hluic, it merges with /r/. It was probably /ɲ/.

_________________
Siöö jandeng raiglin zåbei tandiüłåd;
nää džunnfin kukuch vklaivei sivei tåd.
Chei. Chei. Chei. Chei. Chei. Chei. Chei.

Hlu phoneme inventory:

Initials:
/pʰ pˡʰ tʰ cʰ kʰ/ <p pl t ch k>
/p pˡ t c k/ <b bl d j g>
/ɓ ɗ ʄ/ <ḅ ḍ x̣>
/ⁿpʰ ⁿpˡʰ ⁿtʰ ⁿcʰ ⁿkʰ/ <mp mpl nt nch nk>
/ⁿb ⁿbˡ ⁿd ⁿɟ ⁿg/ <mb mbl nd nj ngh>
/f s ɬ ç h/ <f s hl x h>
/l r j v/ <l r y v>

Finals:
/a ɔ e ɤ o i ɯ u/ <æ a e œ o i ᵫ u>
/æw aw ew ɤw iw/ <æu au eu œu iu>
/æj ɔj ɤj oj ɯj uj/ <æy ay œy oy ᵫy uy>
/æʔ aʔ ɔʔ eʔ ɤʔ oʔ iʔ ɯʔ uʔ/ <æk æp ak ek œk ok ik ᵫk uk>
/æs as ɔs es eç ɤs os iç ɯs us/ <æl æs as el es œs os is ᵫs us>
/aɬ ɔɬ eɬ ɤɬ oɬ ɯɬ uɬ/ <æhl ahl ehl œhl ohl ihl ᵫhl uhl>
/ɔm em om im um/ <am em om im um>
/æn on ɯn un/ <an on in un>
/ɔjn ujn/ <ain uin>
/ɔŋ eŋ ɤŋ/ <ang eng œng>


e lowered before -n -l, a raised before -k, then -s lateralized and -l merged into -ʂ, which became -s. buh who cares

_________________
Siöö jandeng raiglin zåbei tandiüłåd;
nää džunnfin kukuch vklaivei sivei tåd.
Chei. Chei. Chei. Chei. Chei. Chei. Chei.

What happened to initial nasals?

right yeah, /m n ɲ ŋ/ <m n ny ng>

_________________
Siöö jandeng raiglin zåbei tandiüłåd;
nää džunnfin kukuch vklaivei sivei tåd.
Chei. Chei. Chei. Chei. Chei. Chei. Chei.

I've seen a lot of Sino-Tibetan (and SEAsian) languages organized in terms of finals v. initials, which you are using on most of your languages. Do you find that there is an advantage with defining the possible nucleus and codas as units v. using phonotactic rules? Taking Hlu, is it just because of the variable permissible combinations of vowels like /æ/ with coda?

Yeah, the advantage is that not all vowels can occur with all codas. The inspiration there is actually English, but I'm analyzing them as initial + rhyme instead of VC sequences (which makes it difficult to specify all the restrictions, which I want to do so I don't have to think about it when I write the diachronics) or the Balto-Slavic 'liquid diphthong' method, which would be useful for cases (like English!) where it's mostly resonants that affect the vowels, not also stops and fricatives as in Hlu.

_________________
Siöö jandeng raiglin zåbei tandiüłåd;
nää džunnfin kukuch vklaivei sivei tåd.
Chei. Chei. Chei. Chei. Chei. Chei. Chei.

yəbe yəbi
yibende yibendzi

yəbezə yəbendzə
yəgebə yəgembə
yəbeşə yəbenşə
yəgemen yəgenmen
yəbene yəgenme
yəbeñi yəbeñi

sen yəbe juto

_________________
Siöö jandeng raiglin zåbei tandiüłåd;
nää džunnfin kukuch vklaivei sivei tåd.
Chei. Chei. Chei. Chei. Chei. Chei. Chei.

idea: impersonal conjugation develops into being used for hypothetical clauses and statements of possibility and so forth

_________________
Siöö jandeng raiglin zåbei tandiüłåd;
nää džunnfin kukuch vklaivei sivei tåd.
Chei. Chei. Chei. Chei. Chei. Chei. Chei.