My last two posts on the origin of Okinawan ʔy- (as in *ʔyaa 'you' (sg.)) got me thinking about the origin of Proto-Tai ʔy-, whose Siamese reflex [j] is still spelled อย <ʔy>: e.g., อยู่ <ʔyuu> yuu 'to stay, be at'.

Li Fang-Kuei (1977: 58) reconstructed a "glottalized" series of initials for Proto-Tai:

*ʔb- *ʔd- *ʔy- *ʔ- (*ʔw-?)

Cf. his "voiced" PT series:

*b- *d- *j- *g-

I cannot find any discussion of PT *ʔw- in Li (1977). If such a consonant ever existed, could it have merged with PT *w-, *hw-, or ʔb-?

Could PT *ʔy- be partly from an earlier *ʔj-? And could PT *ʔ- be partly from an earlier *ʔg- as well as *ʔ-? In turn, could each glottalized voiced stop be from glottal stop + consonant clusters: e.g., *ʔj- < *ʔc-, *ʔj-, *ʔɲ-?

In Okinawan, ʔy- is from *ʔi- before a glide or vowel. PT *ʔj- appears in at least 21 PT root words, so it must have a more general origin. Moreover, PT *ʔb- and *ʔd- are very common.

For a long time, I have suspected that *ʔC- is from an earlier *ʔVC-. To confirm this, I'd like to see foreign *ʔVC- corresponding to PT *ʔC-. Unfortunately, the only PT *ʔy-word with a potential external connection that I know of is PT *ʔyaŋ (tone B) 'manner' which may be a borrowing from Late Old Chinese 樣 *yɨaŋh. Did PT borrow from an LOC dialect with initial *ʔy- or is the *ʔ- of PT *ʔy- a trace of a PT prefix?

4.17.00:31: Northern Tai languages have forms pointing to earlier *y- in 'manner'. Could these forms be later borrowings from Chinese that replaced the *ʔyaŋ inherited from PT?

樣 is not attested in Early Old Chinese. Could LOC 樣 *yɨaŋh be a borrowing from PT *ʔyaŋ (whose 'tone B' might have been a segmental *-h) with simplification of the initial to *y-?

If 樣 were a borrowing, it shouldn't have a Chinese-internal etymology. However, "Chinese commentators imply cognation with 像象" (Schuessler 2007: 535). 像 and 象 'to resemble' were EOC *syaŋs. If their root were *yaŋ, then perhaps 樣 was *ʔyaŋs < *ʔV-yaŋ-s with a prefix *ʔ(V)-. See Sagart (1999: 108-109) for more potential examples of EOC words with a *ʔ-prefix. (Sagart would now reconstruct some of these EOC *ʔ-C-words with uvulars.) UCHINAAGUCHI PART 24: CUTTING 'YOU' IN TWO

Last night, I failed to mention why I cut Okinawan ʔyaa 'thou' and O ʔittaa 'you' in two, even though I disapprove of arbitrary segmentation.

O ʔittaa 'you' is part of a larger paradigm. Nearly all of the plurals contain a common suffix:

Person Singular Plural
1st wan wattaa
2nd ʔyaa ʔittaa
2nd formal naa nattaa
ʔunju ʔunjunaa (not ʔunjuttaa)
3rd < mesial 'that' ʔuri ʔuttaa
3rd < distal 'that' ʔari ʔattaa

Sakihara (2006: 74, 196) also glossed ʔittaa and ʔunjunaa as 'your house' and that got me thinking that ʔyaa 'thou' could be from the root of ʔittaa plus yaa 'house'. Cf. J otaku 'you', literally 'honorable house'.

ʔyaa could be from *ʔe-ya or *ʔi-ya, but I chose the former since the plural had -tt- instead of a -tch- that would have been palatalized by an earlier preceding *i. UCHINAAGUCHI PART 23: FROM ʔI TO ʔY

Just as Okinawan ʔw- came from *ʔu-,ʔy- came from *ʔi- before a glide or vowel. This *ʔi- may in turn come from *ʔe- as well as *ʔi-:


O ʔyaa < *ʔe-ya 'thy house'

cf. O yaa, J ya 'house'

and O ʔittaa 'you' (pl.); the lack of palatalization of -tt- points to a preceding *ʔe- 'thou'

Proto-Ryukyuan *ʔe- 'thou' is homophonous with PR (and Proto-Japonic) *ʔe- 'wh-' in question words. Could the latter be a borrowing from Koreanic into Proto-Japonic? Cf. Korean ə-interrogative pronouns. See Vovin (2010: 66-67). I don't know of any other language in which 'thou' shares a root with interrogatives.

O ʔy(-un) < Proto-Japonic *ip- 'to say' (cf. J i-u)

cf. inflections:

O ʔi-i : J i-i < ip-i 'says, and'

O ʔi-chi : J it-te < ip-i-te  'saying'

O ʔy-an : J iw-an < ip-an-u  'does not say'

J i-u 'to say' is phonetically [yuu], but is written etymologically as a root i- plus an ending -u. The change of iu to [yuu] (with compensatory lengthening of u to uu but no glottal stop) is regular in Japanese and also occurred after consonants: e.g.,  琉球 Riukiu > Ryuukyuu 'Ryukyu' corresponding to O Ruuchuu ~ Duuchuu (with ch < *ky).

ʔy- also occurs before o in a few related Okinawan words

ʔyooii 'baby'

ʔyooiiʔyooii (word used to soothe a baby)

ʔyooiigwaa 'baby' (cf. kwaa 'child')

but I don't know whether their ʔy- goes back to an earlier *ʔiy- or *ʔey-. They could even have been created after ʔy- was developed.

4.15.00:55: The male speaker in this database prefers ʔyooʔii except in ʔyooʔiiʔyooii. The female speaker consistently pronounces ʔyooii. The transcriptions in the database mix the two, but I've chosen the latter for consistency. UCHINAAGUCHI PART 22: ʔY-UST AN INTERMISSION

I don't have time to write a full follow-up to part 21, so for now I'll just mention a couple of words which are the only root words in Sakihara (2006) with initial ʔy-:

O ʔyaa 'you' (no J cognate)

O ʔyun : J iu 'to say'

Where do you think the rare Okinawan initial ʔy- comes from? What might have been the earlier form of ʔyaa? UCHINAAGUCHI PART 21: LOST IN SPACE

When working with related languages, it's really easy to fall into the trap of mindlessly converting forms from a familiar language into a less familiar language: e.g., taking the Japanese word sore 'that' and turning it into a nonexistent Okinawan suri (the actual word is ʔuri with an unexpected initial glottal stop). So I'd expect Japanese aida 'space, interval' to correspond to an Okinawan ʔeeda. But in fact, aida also corresponds to

ʔeeja with -j- < *-d- after *i

ʔweeda with ʔw- corresponding to zero in (Old) Japanese

Moreover, there are other Okinawan words with ʔee- ~ ʔwee- variation:

ʔeeka ~ ʔweeka 'relative'

ʔeeki ~ ʔweeki 'wealth'*

ʔeeku ~ ʔweeku 'oar'

ʔeema ~ ʔweema : Middle Japanese aɸi 'interval', J ma 'interval'

Sakihara (2006: 37-38) lists yet other ʔee-words without such variation:

O ʔee : J ai 'indigo'

O ʔeesachi : J aisatsu 'greeting' (挨拶)

O ʔeeti : J aite 'partner'

Conversely, Sakihara lists no ʔwee-words without ʔee-variants.

What's going on? Should one reconstruct a Proto-Japonic ʔw- that was sometimes retained in Okinawan but lost in Japanese? I wouldn't. This phenomenon seems to be unique to ʔ(w)ee-. There are no cases of O ʔaa- ~ ʔwaa- corresponding to J a-, etc. Could the following words be relevant to a solution?

O wii : J yoi < Middle Japanese yoɸ-i, Old Japanese wep-i 'intoxication'

O wiiri : J eri < yeri 'collar'

O wiiruu ~ yiiruu : J ? 'cord'; cf. O yii ~ ʔii 'cooperative labor' below

O wuu ~ ʔuu : J o < wo 'cord'

O wusamar- ~ ʔusamar- : J osamar- < wosamar- 'to be at peace'

O yii ~ ʔii : J i < wi 'boar' (of the twelve Earthly Branches)

O yii ~ ʔii : J i < wi 'rush' (plant)

O yii ~ ʔii : J e < we < Middle Chinese 繪 *ɣwajh 'picture'

O yiifee : J ihai < wipai 'mortuary tablet' < Middle Chinese 位 *wih 'position' + 牌 *bæj 'tablet'

O yiigoo- ~ ʔiigoo- : J egu- < wegu- 'acrid'

O yir- : J ir- < wir- 'to sit'

O yii ~ ʔii 'cooperative labor' : J yui 'tie'; cf. O yiiruu 'cord' above

O Yeema ~ ʔeema : J Yaeyama 'Yaeyama'

O Yeigo ~ ʔeigo : J Eigo 'English language'

O yeisaa ~ ʔeisaa 'group Bon dance'

More tomorrow.

*4.13.1:21: O ʔeeki superficially resembles J eki < Middle Chinese 益 *ʔiek 'profit', but I would expect Sino-Okinawan *ʔichi, not ʔeeki. UCHINAAGUCHI PART 20: A ʔW-EALTH OF WUN-DERS

So far, I've dealt with unexpected initial w- and unexpected -w- after k- and g- in Okinawan. The only cases of ʔw- that I've talked about were in ʔwi- corresponding to Japanese ue-: e.g.,

O ʔwii : J ue 'top'

O ʔwir- : J uer- 'to plant'

The glottal stop originally preceded a vowel *u- that became a glide -w-.

But not all O ʔw- are in syllables corresponding to J ue. Given

O ʔwiimaas- : J oimawas- 'to chase'

O ʔwaabi : J uwabe 'surface'
what would be the Okinawan words corresponding to the following Japanese words?

J oitsuk- 'to overtake'

J uwaki 'fickleness' ('affectation' in Okinawan)

Don't worry about Okinawan vowel length.

What is surprising about the following Okinawan ʔw-ords? Which forms would you have expected and why?

O ʔweeda : J aida < Middle Japanese aɸida < Old Japanese apinda 'interval'

also cf. O ʔweema 'interval' : Middle Japanese aɸi 'interval', J ma 'interval'

Compare them with

O ʔee : J ai 'indigo'

O ʔeesachi : J aisatsu 'greeting' (挨拶)

O ʔeeti : J aite 'partner'


Yesterday I asked,

Which of those borrowings are older? Newer? Why?

Assuming that none of these words were consciously Okinawanized (i.e., archaized, or as Leon Serafim might put it, 'grandfathered'):

- kayaku 'fireworks' is the newest; it looks like a loan from Japanese 火薬 kayaku after 火 kwa 'fire' became ka in Japanese.

- the remaining four words preserve a medial -w- lost in Japanese and must have been borrowed before -w-loss in Japanese

- kekkwa 'result' has a short -e- that must postdate the raising of *e to i, so it must have been borrowed from a Japanese *kekkwa when Japanese still had -w-

- kwagun 'saying too much' has a high vowel u < *o, so it must have been borrowed before the raising of mid vowels *o and *e in Okinawan and is therefore older than kekkwa, which would have become *kikkwa if it had undergone raising

- kwaashi 'confection' has a long vowel absent from modern Japanese, so it must have been borrowed from a nonstandard (and premodern?) Japanese dialect with medial -w- and a long vowel

- kwaji has no archaic features other than medial -w-, so its dating relative to the other three medial -w-words can't be determined

Unfortunately, such an assumption may be wrong because it is possible for a recent loan to be Okinawanized / archaized / grandfathered: e.g., Japanese 銀行 ginkou 'bank', coined during the Meiji period, corresponds to Okinawan jinkoo instead of ginkoo even though 'bank' long postdates the palatalization of *g to j. Was 'bank' was borrowed by eye rather than by ear? An Okinawan speaker could have seen the written word 銀行 'bank' and read 銀 'bank' and 行 'shop' using Sino-Okinawan readings jin and koo*. But do such readings exist? The Okinawan word for 'silver' is gin, not jin. Was there ever a word jin 'silver' in Okinawan? The only two Okinawan words for 'silver' that I know of are gin and nanja (cf. Middle Japanese 南鐐 namreu 'southern fine silver'**). The latter appears as 南者 in Chinese transcriptions of Old Ryukyuan in 琉球館譯語 and 使琉球録. gin may be a modern loanword from Japanese. Maybe a better question is: was there ever a Sino-Okinawan reading jin < *gin that was (1) borrowed prior to palatalization and (2) coexisting alongside nanja?

Vovin (2010: 41) lists a number of potential cases of 'grandfathering' in Yonaguni: e.g., dasai 'vegetables', possibly a recent loan from Japanese yasai (< Chn 野菜) with y shifted to d, replicating a change that occurred earlier in native words: e.g., duru < *yoru 'night' (cf. Old Japanese yoru, Okinawan yuru). (The native Yonaguni word for 'vegetables' is atakui.) The Yonaguni name for Yonaguni, Dunan, underwent this change sometime after the 15th century when it was transcribed by a Korean as 閏伊 zyuni. Vovin (2010: 44) reconstructed 15th century "Yonaguni [zuni] or [zyuni], which is apparently an intermediate stage of the fortition *y- > z- > d- in Yonaguni." (There was no character pronounced zun in Middle Korean, so 閏 zyun could represent a foreign zyun or zun.)

*Cf. how Koreans borrowed Japanese words written in Chinese characters and pronounced them in Sino-Korean: e.g., J 時計 tokei > SK 시계 shigye (rather than something like 도케이 *tokhei).

**The final -m of 南 'south' is preserved in Miyako namza 'silver'. Other languages have an assimilated nasal: O and Amami nanja, Nakijin nanza.

A match between Okinawan -ja and Middle Japanese -reu > later -ryou is not impossible given the following rare correspondences:

Okinawan d- : Japanese r-
Okinawan j- < dy- : Japanese ry-

e.g., O juri 'courtesan' : J ryouri 'cooking' (! - identified by Sakihara 2006: 77)

This site derives juri from juri nu yaa 'house of cooking', a euphemism for where courtesans were.

Okinawan -aa : Japanese -ou < -au

in O yuuwaa : J yuou < yuwau (< Chn 硫磺) 'sulfur'

Could O nanja, etc. be from a earlier Japanese form like *namrya(u)?

Tangut fonts by Mojikyo.org
Tangut radical font by Andrew West
All other content copyright © 2002-2011 Amritavision