Morsiθ, an extinct language of the Korean peninsula

[Ketsuban]

I've come into possession of a document in Classical Chinese dating from the end of the seventh century. It's mostly not that interesting - a short but flattering history of Silla and its conquests of Baekje and Goguryeo by a scribe hoping to get into the good books of a new ruler - but it gets interesting when he spends a little time talking about the people he's encountered in the south. He complains about how poorly the sounds of the language of one village in particular (which he claims is named mormur, and its language morsiθ) map onto Chinese sounds and includes a shortish word list to prove the point. I've done the tedious work of determining what phonetic distinctions he was trying to convey and include the list here for posterity.

Lexeme	Gloss
aᵐb	horsefly
af	millet
am	rain
as	hemp
d͡ʑam	mountain
d͡ʑøn	dog
kaᵐb	lichen
katix	oak
kax	persimmon
kɛt͡ɕ	crow
kɛn	crab
kɛr	grass
kitys	wolf
kɔm	tortoise
koⁿd	whale
kom	spider
komɛs	bear
kokøf	swan
kyr	walnut
mør	seaweed
møx	wheat
pad͡ʑɛm	snake
pœθ	wasp
pɔθs	pigeon
pøx	pufferfish
puxmɛr	owl
pyr	pigweed
saᵐb	mackerel
tar	chicken
tɛx	bamboo
tyx	moon
usaᵑg	rabbit

It looks like this language had fifteen consonants /m n p t t͡ɕ k d͡ʑ ᵐb ⁿd ᵑg f θ s x r/ and ten vowels /i y u e ø o ɛ œ ɔ a/. Unfortunately all I have is this short list of words for animals, plants and natural features.

More: show

This was supposed to be an exercise in diachronics more than anything else... usaᵑg is a bit of a giveaway, see if you can find any other cognates.

[fusijui]

It looks like a pretty straight Japonic deal, though I do suspect I'm missing something. I'd like more, regardless!!

BTW, how did you define/discover the dental fricative, palaeographically speaking?

[Travis B.]

Are you sure that d͡ʑ was not ⁿd͡ʑ ?

[Rounin Ryuuji]

I think I've figured out most of the correspondences:

More: show

*aᵐb "horsefly" : 虻 abu (id.)
*am "rain" : 雨 ame, -zame (id.)、天 ama-, ame- "sky, heavens"
*as "hemp" : 麻 asa (id.), probably a Koreanic loanword, cf. OJ also *wo
*d͡ʑam "mountain" : 山 yama (id.)
*d͡ʑøn "dog" : 犬 inu (id.; PJ *yenu, dialectal Japanese enu)
*ka^mb "lichen" : 蕪 kabu "turnip", 株 kabu "stump, clump of roots" (?)
*katix "oak" : 柏 kashwa "Mikado oak", 樫 kashi "evergreen oak", assuming the *x is a reflex of PJ *ki₂ - "tree"
*kax "persimmon" : 柿 kaki (id., but thought to be derived from an Old Japanese verb kakayaku "to shine", modern kagayaku, so maybe a false positive)
*kɛn "crab" : 蟹 kani (id.)
*kɛr "grass" : 刈 kari "cut, clipping" (?)
*kitys "wolf" : the initial element may be identical to the kitsu in 狐 kitsune, with a secondary element *si attached
*kɛt͡ɕ "crow" : 褐 kachi "dark indigo" (?) (if so, a Sinitic borrowing initially)
*kɔm "tortoise" : 亀 kame (id.)
*koⁿd "whale" : probably from the initial element in 鯨 kujira, Proto-Japonic (probably) *kwondi(ra), though the form not being **køⁿd is a little surprising
*kom "spider" : 雲 kumo, OJ *kwomo (id.)
*komɛs "bear" : 熊 kuma (id.)
*kokøf "swan" : 鵠 kugui (id.), OJ presumably *kukupi or *ku^ŋgupi (the form is probably onomatopoeic, from the look of it)
*kyr "chestnut" : 栗 kuri (id.)
*mør "seaweed" : 森 mori "forest" (?), the word originally meant "piling up, heap" apparently; alternately 盛り mori "serving of food"
*møx "wheat" : possibly akin to 蒸し mushi "steaming", or 麦 mugi "barley, wheat"; possibly also connected to 餅 mochi — sticky rice, but I would've expected another development from that;
*pad͡ʑɛm "snake" : bears some resemblance to 蛇 hebi (id.), which has a dialect form habu; probably suffixed with something that looks like a cognate to 闇 yami "dark, darkness";
*pœθ "wasp" : 蜂 hashi "bee, wasp, hornet" (?)
*pɔθs "pigeon" : 鳩 hato (id.), but probably with a suffix?
*pøx "pufferfish" : possibly akin to 河豚 fugu, but the form is unexpected, unless I misunderstand the other correspondences
*puxmɛr "owl" : the form recalls the fuku of 梟 fukurou, with some uncertain element whose corresponding Japanese word would presumably be *mari; 鞠 mari — "ball" — does exist, and may be connected, given that owls are round, but this begins to feel far-fetched...
*pyr "pigweed" — 振り furi "shake, swing"; (note that Japanese yuri - "lily" - seems to be from a word meaning "sway"; I suspect a similar semantic development).
*saᵐb "mackerel" — 鯖 saba (id.)
*tar "chicken" — 鳥 tori "bird, chicken"
*tex "bamboo" — 竹 taka "bamboo", which may be connected with 高 taka "high"
*tyx "moon" — 月 tsuki, tsuku- (id.)
*usaᵑg "rabbit" — 兎 usagi (id.), suprisingly no umlaut; **(w)øsɛᵑg would have been what I expected, so perhaps an interdialectal borrowing

It's nice to see another Japonic thing. I'm also curious as to how many I got right.

[Ketsuban]

More: show

BTW, how did you define/discover the dental fricative, palaeographically speaking?

*Jedi mind trick* You don't need to know that. (I didn't think too hard about the framing device.)

Are you sure that d͡ʑ was not ⁿd͡ʑ ?

Yes. Technically there should be a ⁿd͡ʑ from -Nc- clusters, but unless I've made a mistake (which is possible) those don't appear in the corpus - the d͡ʑ which does appear is fortition of /j/ (which is actually just me getting lazy after fighting SCA2 getting confused between /y/ and /j/ <y>).

I think I've figured out most of the correspondences:
You do reasonably well, though your starting point isn't quite right - I used Dr Francis-Ratte's reconstructed Proto-Korean-Japanese as a starting point, so not all Japanese sound changes are going to show up here. Every term is of course going to have a Japanese reflex due to the nature of the source, of course.

*aᵐb "horsefly" : 虻 abu (id.)
Correct; pKJ *əŋpur. I was a bit lazy with the gloss here, the original might just have meant "fly".
*am "rain" : 雨 ame, -zame (id.)、天 ama-, ame- "sky, heavens"
Correct; pKJ *əmaŋ. There is no proto-Japonic /z/ - the linking -s- in words like harusame is a reduced adjectiviser *-si.
*as "hemp" : 麻 asa (id.), probably a Koreanic loanword, cf. OJ also *wo
Correct; pKJ *əsa - it's not a loan.
*d͡ʑam "mountain" : 山 yama (id.)
Correct; pKJ *jəma.
*d͡ʑøn "dog" : 犬 inu (id.; PJ *yenu, dialectal Japanese enu)
Correct; pKJ *jenu.
*ka^mb "lichen" : 蕪 kabu "turnip", 株 kabu "stump, clump of roots" (?)
Incorrect - pKJ *kənpom gives Japanese 黴 kabi. The vowel is deceiving here - Morsiθ doesn't have oj > e > i like Japonic.
*katix "oak" : 柏 kashwa "Mikado oak", 樫 kashi "evergreen oak", assuming the *x is a reflex of PJ *ki₂ - "tree"
Correct; pKJ *kati-kɨr.
*kax "persimmon" : 柿 kaki (id., but thought to be derived from an Old Japanese verb kakayaku "to shine", modern kagayaku, so maybe a false positive)
Correct; pKJ *kakəm. Count me as skeptical of associations with kagayaku - it looks like a compound to me, possibly involving kage < pKJ *kanxər.
*kɛn "crab" : 蟹 kani (id.)
Correct; pKJ *kane.
*kɛr "grass" : 刈 kari "cut, clipping" (?)
Incorrect; pKJ *kərra "long grass" gives Japanese 茅 kaya "thatch".
*kitys "wolf" : the initial element may be identical to the kitsu in 狐 kitsune, with a secondary element *si attached
Correct; pKJ *kituj. The -s is the -si found in many Japanese words for creatures - ushi, odoshishi, inoshishi, mushi. Francis-Ratte thinks it's a creature suffix; I thought kitys was more interesting than **kyθ.
*kɛt͡ɕ "crow" : 褐 kachi "dark indigo" (?) (if so, a Sinitic borrowing initially)
Incorrect; pKJ *kacɨ-saŋi gives Japanese 鵲 kasasagi.
*kɔm "tortoise" : 亀 kame (id.)
Correct; pKJ *kamoŋ.
*koⁿd "whale" : probably from the initial element in 鯨 kujira, Proto-Japonic (probably) *kwondi(ra), though the form not being **køⁿd is a little surprising
Correct; pKJ *kontɨj. Francis-Ratte thinks 鯨 is a lexicalised plural. I'm not sure why the vowel isn't umlauted, come to think of it... let's call it an irregular form.
*kom "spider" : 雲 kumo, OJ *kwomo (id.)
Correct; pKJ *komo.
*komɛs "bear" : 熊 kuma (id.)
Correct; pKJ *koma-si (there's that "creature" suffix again).
*kokøf "swan" : 鵠 kugui (id.), OJ presumably *kukupi or *ku^ŋgupi (the form is probably onomatopoeic, from the look of it)
Correct; pKJ *kokopi.
*kyr "chestnut" : 栗 kuri (id.)
Correct; pKJ *kɨroj.
*mør "seaweed" : 森 mori "forest" (?), the word originally meant "piling up, heap" apparently; alternately 盛り mori "serving of food"
Incorrect; pKJ *mej-ru (from *mej "water") gives Japanese 海松 miru.
*møx "wheat" : possibly akin to 蒸し mushi "steaming", or 麦 mugi "barley, wheat"; possibly also connected to 餅 mochi — sticky rice, but I would've expected another development from that;
Correct; pKJ *morki gives Japanese 麦.
*pad͡ʑɛm "snake" : bears some resemblance to 蛇 hebi (id.), which has a dialect form habu; probably suffixed with something that looks like a cognate to 闇 yami "dark, darkness";
Correct; pKJ *pəjami gives proto-Japanese *pemi.
*pœθ "wasp" : 蜂 hachi "bee, wasp, hornet" (?)
Correct; pKJ *pator.
*pɔθs "pigeon" : 鳩 hato (id.), but probably with a suffix?
Correct; pKJ *pa:to. I forget what I suffixed to get -s; if it's *-si again then that's a mistake. It might have been *saŋi, come to think of it.
*pøx "pufferfish" : possibly akin to 河豚 fugu, but the form is unexpected, unless I misunderstand the other correspondences
Correct; pKJ *pokɨ. It's not certain whether this is even reconstructible to pKJ; if it's from puku "blow" then the vowel is wrong for the connection with Korean 복(어).
*puxmɛr "owl" : the form recalls the fuku of 梟 fukurou, with some uncertain element whose corresponding Japanese word would presumably be *mari; 鞠 mari — "ball" — does exist, and may be connected, given that owls are round, but this begins to feel far-fetched...
Correct; pKJ *puku-mari. Francis-Ratte thinks *-mari "raptor" > Japanese -me, and assumes the *-ropu also found in 蜉蝣 "mayfly" is an unknown suffix.
*pyr "pigweed" — 振り furi "shake, swing"; (note that Japanese yuri - "lily" - seems to be from a word meaning "sway"; I suspect a similar semantic development).
Incorrect; pKJ *pirrum gives Korean 비름. Apparently it also gives Early Middle Japanese piyu "pigweed", but I can't find that to give kanji for it. Let's call it a Korean loanword.
*saᵐb "mackerel" — 鯖 saba (id.)
Correct; pKJ *sampə.
*tar "chicken" — 鳥 tori "bird, chicken"
Correct; pKJ *tərəŋ.
*tex "bamboo" — 竹 take "bamboo", which may be connected with 高 taka "high"
Correct; pKJ *takəj. The connection with "high" might actually be an illusion like 上 and 神, since the pKJ for that is *takar.
*tyx "moon" — 月 tsuki, tsuku- (id.)
Correct; pKJ *tɨkor.
*usaᵑg "rabbit" — 兎 usagi (id.), suprisingly no umlaut; **(w)øsɛᵑg would have been what I expected, so perhaps an interdialectal borrowing
Correct; pKJ *usənki. It's not **usɛᵑg because the middle vowel is /ə/ rather than /a/.

[Rounin Ryuuji]

I think I've figured out most of the correspondences:
You do reasonably well, though your starting point isn't quite right - I used Dr Francis-Ratte's reconstructed Proto-Korean-Japanese as a starting point, so not all Japanese sound changes are going to show up here. Every term is of course going to have a Japanese reflex due to the nature of the source, of course.

Ah, so it's closer to a Para-Japonic-Koreanic than a Japonic language (I started from a Proto-Japonic cobbled together from several sources, and I'm still not sold on Proto-Korean-Japanese ever having existed; this, however, doesn't change anything about this project being interesting. Also, it seems like terminal stops regularly lenite, so I wouldn't worry too much about the dental fricatives).

I don't suppose you've produced any textual matter yet?

[Ketsuban]

More: show

I don't suppose you've produced any textual matter yet?

I have not - the main reason I stopped at a wordlist is there's basically nothing syntactic in Francis-Ratte's dissertation, just some nominal and verbal morphology. I need to read my copy of Frellesvig a few hundred more times before I'll feel confident trying to come up with actual grammar.

Incidentally, there's a couple words you didn't have a go at - the names of the village and language!

[Rounin Ryuuji]

My guess based on what you've said is that *sith is connected with Japanese 舌 shita ("tongue"), and the element *-mur is probably connected with Japanese 村 mura ("village"); as far as the *mor- element goes, it probably isn't cognate to "mori" ("forest") or "more" ("leak, filter"); my guess, since some of the -r- in Morsith (if you won't mind my giving it a Romanisation without a random IPA symbol in otherwise ASCII text) appear to correspond to Japonic -y-, I might guess it's cognate to Japanese 靄 moya ("mist, fog"), making Mormur the "Mist Village".