DOCUMENT RESUME

ED 072 673 FL 003 624

AUTHOR Bedell, GeorgeTITLE Verbs and Adjectives in Literary Japanese According

to Suzuki Akira.INSTITUTION California Univ.,- Berkeley. Japanese Linguistics

Workshop.PUB DATE Jun 72NOTE 13p.; In Papers in Japanese Linguistics, v1 n1 p11 -23

June 197 2

EDRS PRICE MF -$0.65 HC-$3.29DESCRIPTORS *Adjectives; *Classification; Criteria; Descriptive

Linguistics; *Form Classes (Languages); *Japanese;Language Research; Linguistic Theory; Suffixes;*Verbs

ABSTRACTThis paper presents two views on the categorization

of Japanese nouns, verbs, and adjectives into form classes: thetraditiOnal view and a view set forth in the writing of Suzuki Akira.The fundamental issue here is the criterion for categorization. Is itthe meaning of words, or is it their grammatical behavior thatdetermines the system of categories? The traditional view establishesa criterion that is grammaticalthe presence or absence ofinflection is crucial in categorization. Suzuki (Akira's) argumentsfor classification appeal to the overall regularity or simplicity ofthe language; his fundamental assumption is that the grammaticalforms of Japanese reflect in a direct way its semantic aspects. Bothapproaches are illustrated with examples. (VM)

0

11

Verbs and Adjectives in Literary,

Japanese According to Suzuki Akira

George Bedell

UCLA

I This paperl is concerned with lexical categories,or in more traditional European grammatical terminology,the parts of speech. This notion has been central inEuropean grammar from its first elaboration by theStoics and Alexandrian h until the present time. Theearliest, categorization was a binary distinction betweennouns and verbs; but traditional grammar generallyrecognizes at least a third category, adjectives:

(1) Word

Noun Adjective Verb

The place of adjectives in the system has proved aperennial focus of controversy. In fact many classicalgrammarians2 considered those words which would now bethought of as adjectives to he a subcategory of nouns:4

(2)

Noun

Word

Verb

Substantive Ad ective

There was as well a third alternatives which consideredadjectives to be a subcategory of verbs:5

(3) Word

Noun Verb

Adjective

U.S. DEPARTMENT OF HEALTH. EDUCATION & WELFARE

OfFICE Of EDUCATION

Verb proper

'PERMISSION TO REPRODUCE THIS COPYRIGHTED MATERIAL HAS BEEN GRANTEOBY

aoarr.-Bo

THIS DOCUMENT HAS SEEN REPRODUCED EXACTLY AS RECEIVED FROM THE

PERSON OR ORGANIZATION ORIGINATING IT. POINTS Of VIEW OR OPINIONS

STATED DO NOT NECESSARILY REPRESENT OFFICIAL OfFICE OF EDUCATION

;G21/111 OR DOIICY

TO ERIC A410 ORGANIZATIONS OPERATINGUNDER AGREEMENTS WITH THE US OFFICEOF EDUCATION FURTHER REPROOUCTIONOUTSIOE THE ERIC SYSTEM REQUIRES PERMISSION OF THE COPYRIGHT OWNER-

.

1.2

One of the oldest issues involved in this contro-versy has to do with the criteria for categorisation.Is it the meanings of words, or is it their grammaticalbehavior, that is to deteraine the system of categories?One might hope, as many grammarians have, that the twokinds of criteria would converge on the same analysis.But such is seldom the result in practice: for 'typical'Indo-European languages like Latin, those who emphasizesemantic considerations tend toward position (3); thosewho emphasize grammatical considerations tend towardposition (2).°

It is interesting that the European grammaticaltradition is not unique in having developed a system oflexical categories. Similar ideas have appeared inIndia and in the Par East, though perhaps not attractingquite as much attention as in Europe. More interestingstill is the appearance of similar problems and con-troversies in largely independent grammatical tradi-tions. A case in point, and the focus of this paper,is the treatment of the question of adjectives vis-a-vis verbs in Japanese traditional grammar.7

II The dominant opinion on the adjective question inJapanese traditional grammar is clearly that adjectivesand verbs both belong to a 'predicative' category whichis itself on a level with the category of nouns. Inother words, the position is variant of (3) above:°

(39 Word

Noun Predicative'

Adjective Verb

The traditional criterion for this particular system ofcategories is grammatical: the presence versus absenceof inflection (katsuT5). Nouns (taigen) are those wordswhich lack the capability; Ipredicatives' (Ogen) arethose which possess that capability. The further sub-categorisation into adjectives (keijBgen) and verbs(sayBgen) is based on differences in the endings whichare attached and in the number of forms which exist(i. e. on the Iconjugation').9

13

Inflection in Japanese is rather different in naturefrom typical Indo-European forms. In the traditionalanalysis,five forms were'recognized for each inflectedword. These are:1°

Imperfect/Negative (ShUengen)

Adverbial/Conjunctive (Esnloglo

Predicative (Setsudangen)

Attributive/Substantival (Rentaigen)

Perfect (IZengen)

The grammatical and semantic functions of these formsare elaborately described An moat grammars of Japanese,literary or colloquial. The predicative form is thatwhich appears in sentence final position; the adverbial/conjunctive form is used in adverbial and (non-final)conjoined clauses; the attributive/substantival form isused relative clauses and in nominalizations; theimperfect/negative form is used in hypothetical clauses,and in future and negative sentences; the perfect formis used in conditional and concessive clauses. Theseare only the primary uses.

The regular verb paradigms are:11

Class I: Class IIa:

tatsu 'stand' oku 'get up'

tata- oki

tachi oki

tatsu oku

tatsu okuru

tate okure

Class IIb: Class III:

u 'be able' kiru 'wear'

e ki

e ki

4

14

uru

Ure

kiru

kiru

kire

Notice that in none of the four regular classes of verbsare all five forms' phonologically distinguished. Theimperfect/negative form, in the one class where it isphonologically unique, appears with a hyphen because itis bound, and must always be followed by some additionalsuffix.

There are in addition a few irregular verbs, ofwhich only one class is of concern here:

Class IV:

ari 'exist' wori 'exist'(= wi+ari)

ara- wora-

ari

ari

aru

are

wori

wori

wore

wore

nari 'copula' -tari 'copula'(= ni+ari) (= to+ari)nara- -tara-

ni -to

nari -tars

naru -taro

nare -tare

The only difference between this irregular class andClass I is the predicative form, which ends in i likethe adverbial/conjunctive rather than in u like theattributive/substantival. It is irregular principallybecause it consists of the single verb ari, togetherwith its compounds. Ari plays a role ale 2 in theformation of periphrastic verbal tenses:"

15

'Stative present'

taLar!, tachi+ari) 'is standing'

o!itari okite+ari) 'is up'

'Stative negative'

tatazari (so tatazu+ari) 'is not standing'

'Stative future'

tatameri (Imo tatalu+ari). 'will be standing'

'Stative past'

tachikeri (r tachiki+ari) 'was standing'

'Stative perfect'

tachitari (mg tachite+ari) 'has stood'

All of these forms are capable o inflecting in thesame general way as the verb ari (though some forms arenot to be found for other reasons). With some of thecompounds, particularly nari and -tari, the adverbial/conjunctive form is normally without ari; however thecompound may occasionally be used in that functionalso.

The adjective paradigms are:13

Class Lea: C ass Yb:

yoshi 'good' ashi 'bad'

yoku (yokara-) ashiku(ashikara-)

yoku ashiku

yoshi ashi

yoki ashiki

(yokere) (Ishikere)

The only difference in the two adjective classes is inthe predicative form; adjectives whose stem ends in millilose the ending -shi. Notice in addition the peculiar-ities in the imperfect/negative and perfect forms.They seem to be compounds of ari similar to the corres-

1$

ponding forms of the copula nari above. There is infact an alternate adjectival paradigm in -ari:14

yokara- ashikara-

yoku ashiku

yokari ashikari

yokaruu ashikaru

yokare ashikare

The forms yokiri and Baru replace, Yoshi and Loki incertain literary styles regularly, in others occasion-ally. Ashikari may also occasionally be used forashiku, just as nari is for ni. Forms from this para-digm seem to have replaced (or possibly supplied) theadjectival imperfect/negative form in some uses, andthe perfect fori entirely (though the reason for thevowel change is unclear).

Finally, there are some adjectives with a boundcopula:15

haruka nari 'distant'

haruka nara-

haruka ni

haruka nari

haruka naru

haruka nare

These have the same peculiarities as ether compounds ofari.

Japanese nouns and non-derived adverbs do notundergo inflection of the sort just described. Thus inJapanese one would perhaps expect both semantic andgrammatical criteria for lexical categorization toconverge on an alysis of type (3) -- and in fact thatis the case. There are, however, dissenters within theJapanese grammatcal tradition. It should be clearfrom the above paradigms that while Japanese verbs andadjective do indeed share the important property ofinflectability, they nevertheless also differ in some

17

ways. The present paper will investigate the ideas ofone such dissenter.

III The only book in pre-modern Japanese grammardeVoted exclusively to the elaboration of a system oflexical categories is Suzuki Akira's Gengyo Shishuron.16In this short treatise, Akira takes two dissentingpositions, the first being that there are in Japanesethree major lexical categories, nouns, adjectives andverbs (in his terms, respectively tai no kotoba, arikatano kotoba, shiwaza no kotoba), wharhe correlftessemantically with oqects, states and actions. ', Thesecond is that ari and its compounds belong in theadjective class rather than among the verbs, which isalso presented first in se antic terms: the words inquestion refer to states.l°

But Akira does not content himself with semanticarguments alone; he lists a number of grammaticalarguments in support of his stand, and against theprevailing grammatically based view. It is theselatter which are most worthy of scrutiny. Argumentsin favor of the categorization of ari And its compoundswith adjectives will be considered first, followed bythose in favor of the independence of the categoriesof adjectives and verbs.

First of all, Akira observes that, in the predi-cative form, those words which he wishes to calladjectives end in i; those which he wishes to callverbs end in u. He draws the conclusion that it is thefinal vocalism in the predicative form of inflectedwords which expresses the semantic value of stateversus action. Secondly he observes that the oppositeof ari is nashi.lnot exist', a clear adjective. Thegrammar of Japanese will be the more regular if sup-pletive positive/negative pairs belong to the samelexical categories. Thirdly, he notes the alternativeadjectival forms in ari. Stylistic variants such asYoshi and yokari also should belong to the same lexicalcategory if regularity is to be valued. Fourthly,similar cases can be adduced from other kinds ofstylistic phenomena: Akira points out the equivalenceof words in X-X-zentari in kanbun, and in Y-Y-shi innative styles. The former is of course one of thecompounds of ari; the latter is unambiguously adjectival.Again, Japanese grammar will be more regular if the twocan be included in the same pan-stylistic lexicalcategory.

18

The force of these arguments fs clearly in anappeal to the overall regularity or simplicity of thelanguage. Akira's fundamental assumption is that thegrammatical forms of the Japanese language reflect in adirect way its semantic aspects. He is claiming that,given his system of categoriesp.which can be motivatedboth semantically and grammatically, the relationshipbetween meaning and grammatical form becomes clearer.Other criteria, such as the overall similarity of theparadigm, are simply irrelevant.

On the matter of the independence of adjectivesand verbs as categories, Akira points out the occur-rence of adjectives, but not verbs, befgre the particleno, a position characteristic of nouns: 49

yoshi no kuruma 'good wagon'

yoki kuruma 'good wagon'

Tare no kuruma 'Tare's wagon'

*fuku no kaze

fuku kaze 'blowing wind'

Ari may also occur in this position, as in:

ari no mama as is'

Secondly, he observes the existence of deverbal nounswhich end in i:

miyuki 1(hon) going' from yuku 'go'

tsukai 'messenger' from tsukau 'employ'

omoi 'thought' from omou 'think'

Thirdly, he notes that generally with derivatives, allthree categories equally may serve as a base to obtainother categories:

Verb as base:

kou 'desire'

Adjective as base:

kanashi 'sad'

koishi 'dear, beloved'

kanashibu 'grieve'

19

Noun as base:

/NM*kuro 'blackness' kuroahi 'black'

kuromu 'blacken'

shruien 'tenacity' shrineshi 'tenacious'

s5zoku 'inheritance' s5zoku 'inherit'

The last two examples are particularly interesting asrare cases of Sino-Japanese words which have taken onthe inflectional properties of native Japanese words.The normal way to use them.predicLLively would be incompounds like shanen pmaelhi or sOzoku su, also citedby Akira.

The force of these arguments is that adjectives,in addition to sharing grammatical properties withverbs, also share some with nouns. They are in a sensean intermediate category between the two polar cate-gories, nouns and verbs, and not necessarily closer toone than to the other. This supports their indepen-dence, and of course the idiom with ari is a nice extrain favor of its being an adjective.

IV The present paper has been directed toward gram-matical theory, in the sense of the kinds of statementsand analyses that have been put forth to account forcertain linguistic facts, and the motivations accompany-ing them. It will not attempt to pass judgment as tothe ultimate correctness of any of the views mentioned.This is not because it would not be possible to resolvethe controversies or desirable to try, but reflects apractical limitation on the scope of the paper. Nodoubt a convincing resolution would be difficult, re-quiring the formulation of a very substantial segmentof Japanese grammar, and quite likely the considerationof other languages as well. But that is another paper..

What should perhaps be clear is the relevance ofAkira's observations to the problem. If they do notultimately lead to the conclusions he draws, they stillmust be accounted for in the context of any betteranalysis. Akira's work seems to represent a high pointof Japanese traditional grammar in'terms of the qualityof argument offered in support of grammatical proposals.However, some of his positions are rather less obvious-ly relatable to contemporary ideas. One such deservesnotice in the present context, as a more or less

20

fitting conclunion to this investigation of verbs andadjectives.

Akira observes that his system of overt expressionfor certain basic semantic relations (the notion lob.jectl by absence of inflection; 'state' by inflectionhaving the predicative fora in i; 'actions by inflection

'having the predicative form in u) does not exist inliterary Chinese, though the same relations must beconveyed nonetheless. He concludes that this is thereason why Chinese literature is difficult to interpretand requires extensive commentary; the formal grammati-cal structure simply fails to adequately express thosebasic relations, resulting in rampant.and deleteriousambiguity. He takes this us convincing proof of thesuperiority of the Japanese language. In Akirals view,then, a language is highly valued to the degree towhich its grammar is h*th regular 'and semanticallyexpressive. This is rminiscent of a view once heldvia-a-visthinese by his better known contemporaryWilhelm von Humboldt. Von Humboldt, however, laterconceded that perhaps Chinese represented an alternativeequally valid linguistic type-- obviously he neverseriously tried to study the Chinese classics."

21

Notes and References

1. This paper was originally prepared for presentationto the 1971 annual meeting of the Association for AsianStudies, as part of a panel on Japanese linguistics, andunder the general rubric 'Kokugaku grammatical traditionand generative transformational grammar'.

2. Cf. the extant work of grammarians like DionysiosThrax, M. Terentius Varro and Apollonios Dyskolos; forcommentary; cf. R. He Robins, A Short History of Lin-/Hi:sties, 1967.

3. The schemata (1), (2) and (3) are over simplifiedin several respects. In particular, the category ofadverbs and its place in the system is ignored.

4. The terminology here is traditional, and not to Lataken too seriously. For example, one could reservethe term noun for substantives, and relabel the largercategory. C ?. notes 5 and 9.

S. This system seems to be that intended in the earn-eat European work on the parts of speech.by the philo-sophers Plato and Aristotle. It is also found in suchlater works as James Harris' Hermes, 1751. As before,the terminology is somewhat arbitrary.

6. These are tendencies only, and there are numerousexceptions. Recently the question of adjectives andtheir role in a lexical category system has receivedsome attention in the context of generative transfor-mational grammar. Paul Postal and George Lakoffassembled a long list of arguments intending to showthat for English, (3) is the correct analysis. Cf.Lakoff, Irregularity in Syntax, 1970, Appendix I. Somecounterarguments may be found in Noam Chomsky, 'Remarkson nominalization', in Jacobs and Rosenbaum, eds.,Read in s in En lish Transformational Grammar, 1970,pp. 19 -9. C . also Eamon Bach, 'Nouns and nounphrases', in Bach and Harms; eds., Universals inLinguistic Theory, 1968.

7. For a general study of Japanese traditionalgrammar; cf. my unpublished MIT dissertation Koku akuGrammatical Theory, 1968. A useful presentat on ofthe diverse views on the parts of speech in Japanesetraditional grammar is Fomin, 'Ix istorii japonskogojazykoznanija: uaenije o aastjax reai u tokugava3cixfilologov', in Paikovskij, ed., Japonskif Lingvisti-aeskii Sbornik, 1959. There is of course a vast

22

Japanese literature on the subject, a partial bibljo-graPhy of which may be found in my dissertation.

8. This is the position reflected in the earliestaccount of Japanese parts of speech, the introductionto Fujitani Seisho's Amt. she, 1778, and in the varioustreatises of Gimon. It is often taken by contemporaryadvocates of (3) for English to be the obviously rightanalysis of Japanese.

9. The terminology is by and large that used by Gimonin his textbook Katsugo shinan, 1843, though it alsoappears in his earlier works, e. g. Yams chi no ahiori,1818. His terms are chosen because they ormeaWi-basis of the usual modern school grammar. Seisho'sterminology was completely different, and had littleeffect on subsequent work. As suggested in note 4, thelabels are hare 1..Iss important than the brackets.

10. The Japanese terms are again those of Gimon; theEnglish ones are those of George Sansom in his AnHistorical Grammar of Japanese, 1928. Other transla-tions are in use, but these are as good as any.

11. This is by and large the classification systemelaborated in Motoori Haruniwa's Kotoba no achimata,1803, selected as before because it became t e rounna-tion of modern school grammar. Ignored in the paradigms

is the imperative (meireigen).

12. These are not normally grouped together in Japanesetraditional grammar: much less considered 'tenses'. TheEnglish terms are my own, and are not to be taken asnecessarily having any descriptive semantic significance.

13. These paradigms are mentioned by Haruniwa, thoughhe includes no detailed treatment.

14. Like the -ari tenses, these forms are not usuallyaccorded any special status in early Japanese tradition-

al grammar.

15. Words like this correspond to the 'traditional'

category keiyeashi. Most early grammarians don'tmention them as a separate entity. Seisho does in theintroduction to the Ani sh3; the examples are his.

16. The book was first published in 1824, though there

is some reason to think that it was written ratherearlier, perhsips at the same time as Akirals other two

23

short grammatical works, Gago oni5 k5) 1801) and Xatsu-:odanzoku fu, 1803. A full English translation appears inray dissertation, and translations of some passagesrelevant to this paper in the notes to Ishigaki's papercited below in note 18.

17. There have been a number of later Japanese gram-marians who treated adjectives as an independent cate-gory--but usually for no more reason than that Europeantraditional grammar so treated them.

18. The only subsequent Japanese gram=ariaa tnat I knouof who has followed Akira in this in Ishigaki Kenji) ina paper titled 1Say3sei yogen hanpatsu no h5sokul) 1942)which first appeared in,_Kokugo to kokubungaku) and isreprinted in Ishigaki's Joshi no rekiahiteki kenkya,1955 and in Yamagiwa, ed., Readings in javanene Lanp:nap7eand Linguistics, 1965. The latter edition contains notesand an English glossary. Though Ishigaki's claims areinteresting, they extend Akira's po:.i.ion somewhat) andare not really relevant here.

19. In the following, the underlined examples are thoseactually cited by Akira.

20. Cf. the discussion of Chinese in Uumboldtzs Lettrca M. =Abel-Remusat sur la nature des formes n.rat'eales

geen aTql, et sur le de la languc chi noise enparticulier, 1827, and :Ober die Vera cniedenheit desmenschlichen Sprachbaues und ihren EinfIusEl auf diegeistige Entwicklu2g des Menschengeschlechts, 1936.

Deoartment of Lin5,10:.-ticsUCLALos Anpr.1Ps, Cplifornin 90N6

...C..FILMED FROM BEST AVAILABLE COPY