Rendaku in the Kahoku Dialect of Yamagata Japanirwin/site/Home_files/Rendaku in...N b 3 d 2012...

N b 3 d 2012

Rendaku in the

November 3rd, 2012

Rendaku in the Kahoku Dialect of Yamagata JapanYamagata JapanMizuki Miyashita - U MontanaTimothy J. Vance - NINJALM k I i Y t U iMark Irwin -Yamagata Univ.

ACOL 2012, Lethbridge, ABACOL 20121 ACOL 20121

Miyashita, Vance, and Irwin2

RendakuRendaku Morphophonemic phenomenon found

mainly in compounds in Japanesemainly in compounds in Japanese. Rendaku in TJ (Tokyo Japanese)

/ori/ ‘fold’/kami/ ‘paper’/ori + kami/

[origami]ACOL 2012


QUESTIONQUESTION How does rendaku surface when a

dialect exhibits different voicing contrasts dialect exhibits different voicing contrasts than Tokyo Japanese?

Tokyo Yamagata Gloss

[mato] [mado] ‘target’

[mado] [mando] ‘window’

[origami] or [oringami]?ACOL 2012


OutlineOutline Rendaku Background (TJ)

Y t (K h k ) di l t Yamagata (Kahoku) dialect Yamagata Rendaku

Co-existing variations in Yamagata rendaku Co existing variations in Yamagata rendaku Dialect Contact Problems toward synchronic analysis

ACOL 2012


Rendaku in TJRendaku in TJInitial voiceless obstruent is voiced in a compound word (preceded by a vowel)compound word (preceded by a vowel)

/ori/ ‘fold’/kami/ ‘paper’/ / p p/ori + kami/

[origami]

However, rendaku is highly irregular.ACOL 2012


Rendaku irregularityRendaku irregularity Many morphemes do not behave this way,

although some cases are systematically constrained for rendakue.g. Lyman’s Law (Layman 1894).

/tori + kago/ [torikago] ‘birdcage’ (bird + cage)

* [t i ]* [torigago]

Many morphemes show unpredictable alternationfor no apparent reason.e.g. /kami/ ‘hair’

/mae + kami/ [maegami] ‘front hair (= bangs)’ /kuro + kami/ [kurokami] ‘black hair’[ ]

ACOL 2012


Rendaku irregularity (cont.)Rendaku irregularity (cont.) Historically speaking, some alternations in TJ

involve more than just a voicing difference.j gThat is, /b/ alternates with /h/, not with /p/.

(note: /h/ is realized as [ɸ] before [u])e g /se + hone/ [sebone] ‘backbone’e.g. /se + hone/ [sebone] backbone

/gomi + hukuro/ [gomibukuro] ‘trash bag’

TJ speakers see all the rendaku alternations as i t f th l hinstances of the same general phenomenon./h/~/b/ are just like /t/~/d/ & /k/~/g/

ACOL 2012


Kahoku DialectKahoku Dialect It is a Tohoku dialect.

Many Tōhoku dialects have word-medial pre-nasalizedvoiced obstruents voiced obstruents.

e.g. [kaᵐbe] ‘wall’ ([kabe] in TJ)

These dialects have word-medial [ᵐb], [ⁿd], [ⁿdz], and [ᵑg] (or [ŋ]) corresponding to TJ word-medial [b d z g].

ACOL 2012[b d z g].


Kahoku Dialect (cont )Kahoku Dialect (cont.) One of the Yamagata dialects. Yamagata: Yamagata:

Population: 1,223,739 (2012)(www.geocities.jp)

There is considerable diversity among the dialects in the area.

We focus on this traditional dialect (Kahoku) of Yamagata Prefecture.

ACOL 2012


Kahokuin Yamagatain Yamagata A Yamagata dialect spoken in

Kahoku-cho (town).

A co-author (Miyashita) grew up in Kahoku-cho, and finds the dialect spoken in the pShōnai region (yellow) almost incomprehensible.

The Kahoku dialect is the Kahoku Population The Kahoku dialect is the representative Yamagata dialect and source for our presentation. YJ (Yamagata Japanese)

Kahoku Population20,990 (2012)(www.geocities.jp)

YJ (Yamagata Japanese)ACOL 2012


Prenasalization & VoicingPrenasalization & Voicing In Yamagata Japanese (YJ),

prenasalized voiced obstruents correspond prenasalized voiced obstruents correspond to plain voiced obstruents in TJ.

e.g. YJ TJ[mado] [mato] ‘target’[mãⁿdo] [mado] ‘window’ [ ] [ ]

ACOL 2012


Yamagata RendakuYamagata Rendaku Prediction:

Instances of rendaku in YJ are realized as Instances of rendaku in YJ are realized as prenasalized voiced obstruents

In deed, prenasalized voiced obstruentsare found in many cases of YJ rendaku.

ACOL 2012


Rendaku in YamagataRendaku in YamagataYamagata (YJ) Tokyo (TJ) Gloss Morphology (TJ)

[hi b d ɨ] [hib ʧi] /hi + hachi/[himbadzɨ] [hibaʧi] ‘hibachi’ /hi hachi/(fire – pot)

[sundaɾe] [sudaɾe] ‘bamboo blind’ /su + taɾe/(rattan – drape) ( p )

[tondana] [todana] ‘cabinet’ /to + tana/ (door – shelf)

[hamãᵑgɨɾɨ] [hamaguɾi] ‘clam’ /hama + kuɾi/ (beach – chestnut)

There are some compounds with ACOL 2012

pno prenasalized voiced obstruents.


No Prenasalization: [beɲɨbɑnɑ]No Prenasalization: [beɲɨbɑnɑ] YJ – TJ correspondence of /mb/ and /b/ is

particularly interesting.

In TJ, rendaku pairs /h/ with /b/ (/h/ < OJ [p]).

e g /beɲi/ ‘rouge’e.g. /beɲi/ rouge/hana/ ‘flower’ /beɲi+hana/ [beɲibɑnɑ]‘safflower’

In YJ, it is also /beɲɨ+hana/[beɲɨbɑnɑ].Without prenasalization

ACOL 2012


Contrasts in YJContrasts in YJ Word-initially

Three way contrast

[hɑsɨ] ‘chopsticks’ /haʃɨ/

[pɑpɑ] ‘daddy’ /papa/Three-way contrast[h]≠[p]≠[b]

[pɑpɑ] y p p

[bɑgɑ] ‘fool’ /baka/

Word-mediallyFour-way contrast[h]≠[p]≠[b]≠[ᵐb]

[sɨohɑmˑɑ] ‘salt flat’ /sɨo+hama/

[pɨɾɨpɨɾɨ] ‘spicy’ /pɨrɨ+pɨrɨ/

[h]≠[p]≠[b]≠[ᵐb][beɲɨbɑnɑ] ‘echo’ /beɲɨ+hana/

[dẽᵐbeso] ‘protruding navel’ /de+heso/

ACOL 2012


No Prenas: Possible AccountNo Prenas: Possible Account[beɲɨbɑnɑ] ‘safflower’

Possibile account: this Yamagata word for ‘safflower’ used gto have prenasalization and lost it.

M d h li d d Many compounds have prenasalized and non-prenasalized variations.

ACOL 2012


Co-existing AlternativesCo existing AlternativesFree Variations Gloss

[ i d ] [ ɨ d ] ‘ d f l’ /mɨ + koto/ [migodo] [mɨᵑgodo] ‘wonderful’ /mɨ koto/ (see – thing)

[tegamɨ] [tengamɨ] ‘letter’ /te + kamɨ/ (hand + paper)(hand paper)

[jɑmɑɡɑdɑ] [jɑmɑᵑgɑdɑ] ‘Yamagata’ /yama + kata/ (mountain – shape)

ACOL 2012Old Fashioned


Dialect ContactDialect Contact The situation in YJ is complicated by the strong

influence of TJ.

Government’s role on “Standardization” in education and academia since late 19th contributed to the idea of “standard = proper” among communities.

This also contributed to the fact that non-TJ dialect speakers becoming bidialectals: some in younger generations prefer using TJ over traditional dialects. g p g

Examples of non-prenasalized rendaku could be the case of assimilation to the “standard”.

ACOL 2012


Toward Synchronic AnalysisToward Synchronic Analysis We need a synchronic analysis of the

phonological system of YJ before we can p g ydecide what counts as rendaku.

The presupposition that the “underlying The presupposition that the underlying system is just like TJ” must be abandoned.

Also it is necessary to resist the temptation to Also it is necessary to resist the temptation to think that all Tōhoku dialects are pretty much the same.

ACOL 2012


Some phonetic detailsSome phonetic details There are phonetic details that require

careful considerationcareful consideration. Prenasalization seems to have developed

into a moraic nasal in some words:YJ TJ[kɑŋːge] /kage/ ‘behind’

[kɑᵑge] /kage/ ‘shadow’

ACOL 2012


Some phonetic details (cont ) Some nasals corresponding to TJ syllable-initial

nasals are longer than others, making the di l l

Some phonetic details (cont.)

preceding vowel nasal[bɑmːbɑ] /baba/ ‘grandma’[kɑmɑᵐbogo] /kamaboko/ ‘fish paste’

It could be that the dialect has developed a phonemic distinction between nasal and oral vowelsvowels

[ɕɨme] ?/hɨme/ ‘princess’[ɕ mˑo] ?/h mo/ ‘string’

ACOL 2012


Co-phonology problemCo phonology problem Problem of how to model the phonological system

of a bi-dialectal speaker. The languages of bilingual speakers influence each

other to some degree even when the two languages are not related or typologically similar.Thi ki d f t l i fl i f d This kind of mutual influence is even more profound in the case of two dialects, especially when one is a “standard” variety of a national language.

“Co-existent phonemic systems” and “Co- Co-existent phonemic systems and Co-phonologies” have been around for more than half a century. (Fries & Pike 1949)

ACOL 2012


Co-phonology problem (cont )(cont.) A bi-dialectal speaker does not have just

two discrete varietiestwo discrete varieties. S/he controls a continuum that ranges

from near-“standard” to something much closer to a “pure” traditional variety.

N t d d T diti lNear standard Traditional

ACOL 2012


Bidialectal ContinuumBidialectal Continuum When an non-prenasalized, word-medial

voiced stop in YJ (e.g. [beɲɨbɑnɑ]) voiced stop in YJ (e.g. [beɲɨbɑnɑ]) corresponds to a voiced stop in TJ, it is tempting to handle it as a kind of unassimilated or only partially assimilated unassimilated or only partially assimilated loanword from TJ.

H th i fl f TJ i t fi d However, the influence of TJ is not confined to cases of borrowing.

ACOL 2012


ExampleYamagata Free Variations Gloss (Tokyo)Yamagata Free Variations Gloss (Tokyo)

[mɨᵑgodo] [migodo] ‘wonderful’ /mɨ + koto/ (see – thing)

[tengamɨ] [tegamɨ] ‘letter’ /te + kamɨ/ ( )[te gamɨ] [tegamɨ] letter (hand + paper)

[tẽᵐbɨgɨɾo] [tebɨgɨɾo] ‘glove’ /te + hɨgɨɾo/(hand – bag)

[ mbɨd ] [ bɨd ] i /me + ɸɨtɑ/[mambɨdɑ] [mabɨdɑ] ‘eyelid’ /me + ɸɨtɑ/(eye – lid)

[jɑmɑᵑgɑdɑ] [jɑmɑɡɑdɑ] ‘Yamagata’ /yama + kata/ (mountain – shape)

Native TJ-InfluencedACOL 2012

Native TJ Influenced


Over Time (generational)Over Time (generational) A TJ word can influence the

pronunciation of a YJ cognate as in the pronunciation of a YJ cognate, as in the case of innovative pronunciation pushing out traditional one.

[jɑmɑᵑgɑdɑ]Traditional

[j d ][jɑmɑgɑdɑ]TJ-Influenced

ACOL 2012


ConclusionConclusion Rendaku in Yamagata Japanese: both

prenasalized and non-prenasalized voiced p pobstruents are found.

Some are always non-prenasalized. Rendaku investigation is complicated by the Rendaku investigation is complicated by the

problem of co-phonology and bidialectalism. Until we determine what counts as rendaku in

Yamagata Japanese (and other Tohoku Yamagata Japanese (and other Tohoku dialects), we cannot advance in rendakuinvestigation.

ACOL 2012


ConclusionConclusion Raises issues

Importance of research in Importance of research in bilingualism/bidialectalism.

Methodology in determining a grammar of one language in the mind of bilinguals.

Linguistic study of a language which is in the process of rapid linguistic change due p p g gto bilingualism/bidialectalism.

ACOL 2012

29 Miyashita, Vance, and Irwin

ACOL 2012


AcknowledgementsAcknowledgements National Institute for Japanese Language and

Linguistics (NINJAL)g ( ) The Research Institute for Humanity and

Nature (RIHN) Yamaguchi Fund, Mansfield Center, University Yamaguchi Fund, Mansfield Center, University

of Montana. Mr. Haruki Yahagi, Mr. Toshihiro Saito, and Ms.

Asako Saito.Asako Saito. Field Research Participants in Kahoku-cho,

Yamagata, Japan.

ACOL 2012


Appendix - kanaAppendix kana In kana spelling, however, rendaku is

transparent: the difference between the transparent: the difference between the voiced alternant and the voiceless alternant of a morpheme is always just the

b f th d k tpresence versus absence of the dakutendiacritic on the first letter.

ACOL 2012


Appendix - kanaAppendix kana/fune//hako/

ふね

はこ

/kawa+bune//haši+bako/

かわ＋ぶね

はし＋ばこ/hako//tama//kami/

はこ

たま

かみ

/haši+bako//me+dama//kabe+gami/

はし＋ばこ

め＋だま

かべ＋がみ/kami//cuka//sora/

かみ

つか

そら

/kabe+gami//ari+zuka//hoši+zora/

かべ＋がみ

あり＋づか

ほし＋ぞら

/či//šika/

ち

しか

/hana+ǰi//ko+ǰika/

はな＋ぢ

こ＋じかACOL 2012

Rendaku in the Kahoku Dialect of Yamagata Japanirwin/site/Home_files/Rendaku in...N b 3 d 2012...

Documents

Transcript of Rendaku in the Kahoku Dialect of Yamagata Japanirwin/site/Home_files/Rendaku in...N b 3 d 2012...