Vowel alternance in Japanese

I love this apparently archaic phenomenon of Japanese (it only
happens with native, yamato-kotoba words). Here are
some examples.

Martin proposes eleven sets of vowel alternation, many (but not
all) with sparse data. They are listed below with his examples (I’m
preserving some of Martin’s phonetic spelling, which include
pitch-accent marks; for wider reception, I changed his consonants to
Hebpurn-style, dispensing with marking the standard/Eastern
realization of /g/ as [ŋ]).

a:o  ‘tooth’ ho.né ‘bone’
a:u futa.tsú ‘two’ futsu.ka ‘two days’
háta.chi ‘age 20’ hatsu.ka ‘20 days’
káta ‘shoulder’ katsú.g-u ‘carries on shoulder’
e:a fúne ‘ship’ funa.bito ‘passengers; boatman’
kaze ‘wind’ kaza-muk.i ‘wind direction’
kaza-kami ‘windward’
 ‘eye’ má.buta ‘eyelid’
ma.no-átar.i ‘before one’s eyes’
má.yu ‘brow’
sake ‘wine’ [i.e. liquor] saka.ba ‘bar’
saka.ya ‘liquor store’
 ‘hand’ ta.gúr-u ‘pulls with hands’
ta.motó ‘sleeve’
ta.yasú-i ‘is easy’
ta.yór-u ‘depends, leans on’
ue//uwe ‘above’ uwa.gi ‘outer garment’
uwa.te ‘upper part’
e:o se ‘back’ so.múk-u ‘turns the back’
i:o hi ‘fire’ ho.kàge ‘flicker’
ho.nò.o ‘flame’
ho.gá.ra.ka ‘clear, bright’
ho.shi ‘star’
hó.ya ‘lamp chimney’
hí.ru ‘dries out’ hó.s-u ‘dries something’
 ‘tree’ ko-bik.i ‘sawyer’
ko.kage ‘shade’
kò.dachi ‘thicket, grove’
okí.ru ‘rises’ okó.s-u ‘raises’
i:u kuchi ‘mouth’ kutsu.wa ‘(bridle) bit’
mi ‘body’ mu.né ‘chest, breast’
o:a shiró.i ‘is white’ shira.ha ‘white teeth’
shirá.m-u ‘lightens’
iso ‘beach’ isago ‘sand’
no ‘of’ ma.na-ko ‘pupil of the eye’
u:a atsú.i ‘is hot’ ata.ta.mé-ru ‘heats up’
ata.ta.ká-i ‘is warm’
yasú.i ‘is easy’ yasa.shí-i ‘is gentle’
u:a:i asú ‘tomorrow’ asá.t.te ‘day after tomorrow’
ashi.tá ‘tomorrow’
u:o karu.i ‘is light’ karo.garo.shí-i ‘is easy-going’
karo.n-zú.ru ‘takes lightly’
kázu ‘number’ kazo.é-ru ‘counts’

For i:u, he also gives chikará ‘force’
vs. mí.zukara/ono.zukara ‘by
oneself’. For e:a he gives in a footnote the delicious case of the
“specificative or diminutive” suffix, which has the two allomorphs
-ne/-na:

-ne ha ‘wing’ ha.ne ‘feather’
ho- ‘tooth’ ho.ne ‘bone’
 ‘tree’ kí.ne ‘pestle’
( ‘field’?) tá.ne ‘seed’
mu- ‘body’ mu.né:mu.na- ‘chest, breast’
ya ‘house’ ya.ne ‘roof’
-na  ‘edge’ há.na ‘brink, edge’
ha ‘leaf’ ha.ná ‘flower’
 ‘powder’ ko.ná ‘powerder’
se ‘back’ sé.na ‘back’
su ‘shoal’ su.na ‘sand’
 ‘circle’ wá.na ‘trap, snare’

Of Martin’s 11 sets, Miller highlights three as important and
frequent enough to be used in internal reconstruction of Old
Japanese: e:a, i:o, and o:a (he’s using criteria set by Marchand).
Here are Miller’s examples (a capital F in Miller’s notation
represents Old Japanese *[p], Early Middle Japanese [ɸ], modern ‘h’):

e:a sake ‘rice wine’ sakaya ‘sales place for sake’
sakazuki ‘cup for sake’
ue ← *uwe ← *uFe ‘topside’ uwagi ‘overcoat’
uwagusuri ‘ceramic glaze’
ame ‘rain’ amado ‘rain doors’
amakaze ‘wind that warns of coming rainstorm’
tsume ‘nails’ tsumasaki ‘tip-toe, a toe’
kane ‘metal’ kanabashi ‘metal chopsticks’
yone ‘rice’ yonamushi ‘rice-insect pest’
kaze ‘wind’ kazamatsuri ‘wind festival’
kazashimo ‘downwind’
koe ← *kowe ‘voice’ kowadaka ni ‘in a loud voice’
ie ‘house’ ya ‘shop’
yanushi ‘master of the house’
i:o ki ‘tree’ kokage ‘shade’
kodachi ‘thicket’
hi ‘fire’ hotaru ‘firefly’
hono’o ‘flame’
ine ‘rice’ yone ‘rice’
niki ‘gentle, soft’ nikoge ‘pinfeathers’
ki’iro-i ‘yellow’ kogane ‘gold’
o:a shiro-i ‘white’ shiraga ‘grey, white hair’
shiraha ‘white teeth’
shiraha ‘naked sword’
shiraha ‘white feather’
shirahada ‘leucopathy’
shirahagi ‘a
white hagi
flower
(Lespedeza tomentosa)’
iso ‘beach, surf’ isago ‘sand’
hiro-i ‘broad’ hirak-u ‘open’
kuroi-i ‘black’ kura-i ‘dark’

In the text, Miller claims the uma-
in umabito “noble person” must be the same
as uwe/uwa “above”, and not uma
“horse” as in popular etymology. The -w-:-m- variation appears
in other places, such as mi-zu “water”
and wi-do (→ido) “well”. He thinks the story that
Prince Shōtoku was born in a stable was probably created to
rationalize in this way his childhood name, Umayado.

The extra y- in ine/yone is easy to accept when we
remember that yi- has long since been realized as i-, as
in yoi / ii (←yii) “good”. In the ki/ko group, neither author mentions
kodama,
those fascinating spirits (tama) of the trees
(ki).

Many of those alternations have even more interesting cases if we compare
dialects, but I don’t have enough energy to get into all the data of
dialect studies right now.

References: Miller, The Japanese Language,
pp. 185–187;
Martin, Morphophonemics of Standard Colloquial
Japanese
, pp. 85–86.

2 thoughts on “Vowel alternance in Japanese

  1. I don’t know much about Japanese, but in some e~a alternations, e seems to be definitely from a-i (-i being some kind of noun ending).

    Interesting, because the same phenomenon can be observed in Korean (*kag-i > kahi > kai > kɛ ‘dog’, *kag-ak-i > kaŋadʑi ‘puppy’).

Leave a Reply

Your email address will not be published. Required fields are marked *