11.3.26.23:36: UCHINAAGUCHI PART 6: WHY CONVERSION DOESN'T ENTAIL DESCENT

From last night: Why doesn't Japanese ki correspond to Okinawan chi in these words?

J ki : O kii (not chii!) 'tree'

(Okinawan monosyllabic words cannot end in short vowels.)

J oki- : O uki- (not uchi-!) 'to rise'

J kizu : O kiji (not chiji!) 'wound'

I gave a hint: Is Japanese really the ancestor of Okinawan? The answer is no. Although I've been treating Japanese as if it were the ancestor of Okinawan, the two languages are cousins rather than sisters. They share a common ancestor (Proto-Japonic), but one is not descended from the other:

Since Okinawan is not descended from Japanese, Okinawan can preserve contrasts lost in Japanese:

The mystery vowel *X must have been less palatal than *i because it didn't condition the palatalization of k to ch in Okinawan.

It turns out that there was more than one nonpalatalizing *X. (For our purposes, we can regard Old Japanese as if it were the ancestor of Japanese*.)

The Old Japanese forms corresponding to the three words I started with are

OJ kɨ 'tree'

OJ əkɨ- 'to rise'

OJ kinzu 'wound'

How would these words be reconstructed in Proto-Japonic?

What else has Okinawan preserved that Japanese has lost? Find more conservative features in the following data. Two should be easy, one is medium, and one is hard.

O kan : J kan 'sense' (感)

O Kwannun : J Kannon 'Avalokiteśvara' (観音)

O gakumun : J gakumon 'learning' (学問)

O gwansu : J ganso 'ancestor' (元祖)

O yin : J en 'relation' (縁)

O fiijii : J heizei 'usual' (平生)

Answers tomorrow.

*Old Japanese is a western dialect whereas standard Japanese is an eastern dialect, so the former is not the ancestor of the latter. OJ is an 'aunt' to standard Japanese.

11.3.25.23:45: UCHINAAGUCHI PART 5: TREE RISING WOUNDS

Here are the answers to last night's questions:

1. Short e and o are rare in Okinawan. An example of a short e word is penki 'paint' (cf. J penki 'id.'). I've already mentioned a short o word. Why don't these words have i and u? (Bonus: Did you notice another seemingly 'violated' rule in one of these words?)

These exceptional words with short e and o (penki 'paint', pasokon 'personal computer') were borrowed after the shift of e and o to i and u.

Penki ends in -ki rather than -chi because it was borrowed after the shift of ki to chi.

(Penki is said to be from Dutch pek, but that word now means 'pitch' in Dutch. Was there a Dutch pek 'paint'? Why does the Japanese word have an -n- absent from the Dutch original?)

2. What is the Okinawan equivalent of J are (wa) 'that' (distant from both speaker and listener)?

J are > O ari

J are wa > O aree

Although one might think -e wa has fused into long -ee, the Okinawan long vowel is actually a fusion of -e and the topic marker -ya (not wa from earlier *pa). The original form of this topic marker is preserved after nouns ending in long vowels and after all nouns in literary language. Examples from Sakihara (2006: xii) and Karimata:

ʔyaa 'you' > ʔyaa-ya (note the ʔy- cluster absent from Japanese)

shinshii 'teacher' (先生) > shinshii-ya

chuu 'today' > chuu-ya

hisa 'leg' > hisaa (lit. hisa-ya)

kami 'god' > kamee (lit. kami-ya)

chinu 'horn' > chinoo (lit. chinu-ya)

jin 'money' (銭) > jinoo (lit. jin-ya)

In Japanese, the topic marker wa is invariable.

3. Given

O mee : J mae 'front'

O eeji : J aizu 'signal'

the Okinawan equivalents of the following words are

J sao 'pole' > O soo (cf. ae > ee)

J nau 'to twist' > O noo-in (-in is a verb ending; cf. ai > ee)

J koori 'ice' > O kuuri 'rock candy' (o > u by the Rule of Three)

Okinawan also has koori 'ice', but I assume that's a loan from Japanese postdating the Rule of Three.

The first rule I taught after the Rule of Three was the rule of velar palatalization:

J kin : O chin 'gold' (金)

J kiyo- 'pure', kiyo-me- 'to purify' : O chu-ra- 'pretty', O chiyu-mi- 'to purify'

(note how J -iyo- corresponds to O -u- as well as -iyu-)

J kugi 'nail' : O kuji

J ki : O kii (not chii!) 'tree'

(Okinawan monosyllabic words cannot end in short vowels.)

J oki- : O uki- (not uchi-!) 'to rise'

J kizu : O kiji (not chiji!) 'wound'

Unlike penki and pasokon, all three of these exceptional Okinawan words are native. Moreover, these three words are not isolated archaisms. Why didn't their k become ch? Answer tomorrow.

Hint: Is Japanese really the ancestor of Okinawan?

11.3.24.22:44: UCHINAAGUCHI PART 4: PC V-O-W-E-LS

Here are the answers to last night's questions:

1. Can you figure out the reason for these apparent exceptions?

O hita (not hicha) : J heta 'unskilled'

O hidati (not hijati) : J hedate 'barrier'

t and d become ch and j before original i, not secondary i corresponding to Japanese e. These exceptions indicate that consonantal shifts predated vocalic shifts:

First: ity > ich, idy > ij

Then: e (or some intermediate vowel like ɪ) > i

2. The following words

O shina : J suna 'sand'

O shiwa : J sewa 'care' (世話)

O chira : J tsura 'face'

O maji : J mazu 'first'

O jihi : J zehi 'certainly' (是非)

reflect these new rules:

1. su, tsu, zu > shi, chi, ji (via si, tsi, zi which were still in Okinawan in the early 20th century)

2. se, ze > shi, ji (via si, zi)

3. Can you predict the Okinawan equivalents of the following Japanese words?

J sensei 'teacher' (先生) > O shinshii

J netsu 'fever' (熱) > O nichi

J suzu 'tin' > O shiji

J kaze 'wind' > O kaji

J kazu 'number' > O kaji

J kaji 'rudder' > O kaji

Bonus question: Why are the rules in these Okinawan posts unlikely to apply to languages like Thai or Vietnamese with monosyllabic native roots?

The changes that occurred in Okinawan create a lot of homophones: e.g., the three kaji above. (There is a fourth kaji meaning 'fiber; muscle' and a fifth meaning 'nape'.) Such homophony would be even more extreme in a Chinese-like language with monosyllabic roots unless new tones were developed to compensate for lost segments and/or roots combined to form polysyllabic compounds replacing monosyllabic words.

I have never heard of a Chinese-like language with only three vowels. As the title of this post implies, even Okinawan has more than three vowels. "PC" stands for personal computer - pasokon (< 'perso-com'), which appears in lesson three of Karimata's textbook:

O kuree pasokon yaibiin.

J kore wa pasokon desu.

lit. 'this (topic marker) perso-com is.'

'This is a personal computer.'

(Obviously yaibiin cannot be derived from desu, which would theoretically correspond to an Okinawan dishi.)

Tonight's questions:

1. Short e and o are rare in Okinawan. An example of a short e word is penki 'paint' (cf. J penki 'id.'). I've already mentioned a short o word. Why don't these words have i and u? (Bonus: Did you notice another seemingly 'violated' rule in one of these words?)

2. What is the Okinawan equivalent of J are (wa) 'that' (distant from both speaker and listener)?

3. Given

O mee : J mae 'front'

O eeji : J aizu 'signal'

what do you think the Okinawan equivalents of the following words would be?

J sao 'pole'

J nau 'to twist'

J koori 'ice'

Answers tomorrow.

11.3.23.20:23: UCHINAAGUCHI PART 3: FURTHER AD-J-USTMENTS

Here are the answers to last night's questions:

Can you predict the Okinawan equivalents of the following Japanese words?

J kinjo 'neighborhood' (近所) > O chinju

J kyoomon 'Buddhist scriptures' (経文) > O chuumun

J ginmi 'scrutiny' (吟味) > O jinmi

J chiri 'dust' > O chiri

J moji 'letter, character' (文字) > O muji

J juu 'ten' (十) > O juu

Three new rules applied to the first three words:

ki > chi

ky > ch

gi > ji

O icha : J ita 'board'

cf. O kata (not kacha) : J kata 'shoulder'

cf. O futa (not fucha) : J futa 'cover'

O hijai : J hidari 'left' (Don't worry about the lack of an -r- in the Okinawan word.)

cf. O hada (not haja) : J hada 'skin'

cf. O fuda (not fuja) : J fuda 'card'

1. Can you figure out the reason for these apparent exceptions?

O hita (not hicha) : J heta 'unskilled'

O hidati (not hijati) : J hedate 'barrier'

Sh as well as ch and j turn up in unexpected words in Okinawan from a Japanese perspective:

O shina : J suna 'sand'

O shiwa : J sewa 'care' (世話)

O chira : J tsura 'face'

O maji : J mazu 'first'

O jihi : J zehi 'certainly' (是非)

2. What new rules are needed to account for the above words?

3. Can you predict the Okinawan equivalents of the following Japanese words?

sensei 'teacher' (先生)

netsu 'fever' (熱)

suzu 'tin'

kaze 'wind'

kazu 'number'

kaji 'rudder'

Bonus question: Why are the rules in these Okinawan posts unlikely to apply to languages like Thai or Vietnamese with monosyllabic native roots?

Answers tomorrow.

11.3.22.23:59: UCHINAAGUCHI PART 2: AD-J-USTMENTS

Here are the answers to last night's questions:

1. What is the Rule of Three?

The correspondence of Japanese e, o to Okinawan i, u.

2. Can you figure out what the 'Three' in the Rule of Three refers to?

The three vowels (a, i, u) in the Okinawan words presented so far.

Japanese has five vowels (a, e, i, o, u).

3. Can you predict the Okinawan equivalents of the following Japanese words?

J ashi 'leg' > O ashi (no phonetic change) 'leg as food'

J kane 'metal' > O kani

J tora 'tiger' > O tura

J uta 'song' > O uta (no change)

O chimu : J kimo 'liver'

O chuuji : J kyuuji 'waiter' (給仕)

O jiri : J giri 'duty' (義理)

(Parentheses indicate Chinese characters for borrowings with Chinese roots.)

But some Okinawan ch and j match Japanese ch and j:

O chikara : J chikara 'strength'

O chuumun : J chuumon 'order' (注文)

O jiyuu : J jiyuu 'freedom' (自由)

kinjo 'neighborhood' (近所)

kyoomon 'Buddhist scriptures' (経文)

ginmi 'scrutiny' (吟味)

chiri 'dust' (the only native word in this problem!)

moji 'letter, character' (文字)

juu 'ten' (十)

Answers tomorrow.

11.3.21.23:59: UCHINAAGUCHI PART 1: THE RULE OF THREE

Uchinaaguchi means 'Okinawan language' in Okinawan, the language of Okinawa which is not mutually intelligible with Japanese (and hence counts as a 'language' in my book).

A lot of Okinawan sounds like Japanese: e.g.,

Okinawan asa : Jpn asa 'morning'

Okinawan mimi : Jpn mimi 'ear'

Okinawan fuyu : Jpn fuyu 'winter'

Many Okinawan words can be guessed by applying the Rule of Three to Japanese words.

(The following exercise is loosely based on homework I assigned to my History of the Japanese Language class 9 years ago. It requires no knowledge of Okinawan, Japanese, or linguistics.)

1. Look at the following data. Can you figure out what the Rule of Three is?

Okinawan tin : Jpn ten 'heaven'

Okinawan saki : Jpn sake 'alcoholic drink'

Okinawan kumi : Jpn kome 'rice'

Okinawan suruban : Jpn soroban 'abacus'

Okinawan yuru : Jpn yoru 'night'

Okinawan sun : Jpn son 'loss'

2. Can you figure out what the 'Three' in the Rule of Three refers to?

3. Can you predict the Okinawan equivalents of the following Japanese words?

ashi 'leg'

kane 'metal'

tora 'tiger'

uta 'song'

Answers tomorrow.

11.3.20.15:50: A ธน THON OF SIMILARITY

(Originaly written 3.7.)

Thai ธน thon 'money' sounds like Korean 돈 ton 'money'. This is not 'proof' of a genetic relationship between the two. 'Money' is a concept that is likely to be borrowed. Thai ธน thon is from Sanskrit or Pali dhana 'wealth' (via Khmer ធន *dhɔn via some Middle Indic *dhɔn?). Korean ton is native* unlike Japanese zeni 'money' and Vietnamese tiền 'money' which are both from Middle Chinese 錢 *dzien 'money'**.

*Middle Korean tǒn had a rising tone implying a disyllabic pre-Middle Korean *tònv́. Although the final vowel was lost, the low-high pitch melody of the word remained. The final vowel and the earliest (i.e., nonmonetary) meaning of the word are unknown.

**Vietnamese tiền is a direct borrowing from a southern Chinese language whereas Japanese zeni may have been borrowed through a peninsular Korean language (most likely Paekche) rather than directly from Chinese.