The interactional origin of nominal predicate structure in Japanese: A comparative and historical pragmatic perspective

The interactional origin of nominal predicate structure in Japanese: A comparative and historical pragmatic perspective

Journal of Pragmatics 44 (2012) 663–679 Contents lists available at SciVerse ScienceDirect Journal of Pragmatics journal homepage: www.elsevier.com/...

287KB Sizes 5 Downloads 70 Views

Journal of Pragmatics 44 (2012) 663–679

Contents lists available at SciVerse ScienceDirect

Journal of Pragmatics journal homepage: www.elsevier.com/locate/pragma

The interactional origin of nominal predicate structure in Japanese: A comparative and historical pragmatic perspective Kaoru Horie * Graduate School of Languages and Cultures, Nagoya University, Furo-cho, Chikusa-ku, Nagoya 464-0601, Japan

A R T I C L E I N F O

A B S T R A C T

Article history: Received 5 January 2011 Received in revised form 11 September 2011 Accepted 26 September 2011

Nominalization plays an important interactional role in East Asian languages. Informed and guided by the collective insight from Typological and Emancipatory Pragmatic approaches to grammar, this paper presents a comparative and historical pragmatic analysis of the preference for sentence-final nominal predicates in Japanese and addresses its communicative/interactional motivation. ß 2011 Elsevier B.V. All rights reserved.

Keywords: Nominal predicate Japanese Korean Attributive forms vs. final forms Backgrounding vs. assertion Noun-oriented structure vs. verb-oriented structure

1. Introduction It is a well-established areal linguistic fact that nominalization is a frequently employed rhetorical strategy in many East and South East Asian languages (Noonan, 1997). Japanese and Korean are no exceptions, where nominalization is extensively used as the predicate structure in sentence-final position. It remains to be seen where such preference for nominalized structure in East Asian languages comes from, in the absence of analytical frameworks where such issues can be appropriately addressed. However, the time has now become ripe for such an inquiry, particularly with the development of relevant disciplines that address cross-linguistically differing trade-off relations between Pragmatics and Grammar. Recent developments in the ‘‘Discourse and Grammar’’ and ‘‘Grammar in Interaction’’ frameworks (e.g. Ford, 1993; DuBois, 2003; Ariel, 2008) and the Pragmatic Inferencing Approach to Grammaticalization (e.g. Bybee et al., 1994; Bybee, 2003; Traugott, 1989, 2003, papers in Onodera and Suzuki, 2007) have enabled one to analyze the role of discourse and pragmatic inference in shaping grammatical structures/constructions and compensating for the potentially underspecified grammatical structure. Underlying these approaches is the insight from functional typology (most explicitly advocated in Myhill’s Typological Discourse Analysis (Myhill, 1992)) that the manifestation of pragmatics–grammar interfaces can vary considerably across languages. Emancipatory Pragmatics (henceforth abbreviated as EP; see Hanks et al., 2009), has now come along, which pays close attention to native speakers’ interactional practice in a given speech community as it relates to the cultural values and

* Tel.: +81 52 789 4799; fax: +81 52 789 4799. E-mail address: [email protected]. 0378-2166/$ – see front matter ß 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.pragma.2011.09.020

664

K. Horie / Journal of Pragmatics 44 (2012) 663–679

communicative priorities of that community. The EP perspective is crucial in shedding new light into the study of pragmatics–grammar interface. Informed by the findings in Discourse-Functional and Typological approaches to pragmatics–grammar interface and inspired by the EP proposal for the need to address socio-cultural foundations of linguistic practices, this paper probes the preferential tendency to nominal predicate structure in Japanese from a comparative perspective by comparing it with other East Asian languages. Specifically, this paper makes a comparison between Japanese and Korean in terms of the extent of nominal predicate structure as these languages exhibit subtle semantic–pragmatic differences in spite of overall structural similarities in their morpho-syntactic structures. It must be noted that the incidence of noun-oriented predicate structures beyond East Asian languages is not entirely clear at this stage, though it is clearly absent from major European languages. A full-scale typological study of the nominal predicate structure is thus next in our agenda. The tendency toward nominal predicate structure in Japanese was noted in previous studies, and some comparisons were made with other languages including English (Hinds, 1986; Maynard, 1997). However, these observations were not embedded into the broader theoretical context that this paper presents. To give the necessary background information, an introduction to the view of cross-linguistically variable pragmatic/ discursive foundations of grammar as conceived of in this paper is given in section 2, with some reviews of relevant literature. This is taken a step further in section 3 with a comparative pragmatic analysis of nominal predicate constructions in Japanese, Korean, and Mandarin Chinese that focuses particularly on the former two languages. Section 4 then presents a historical pragmatic analysis of the communicative/interactional factors motivating and leading to the preference for sentence-final nominal predication in Japanese to tie it all together. Finally, the conclusions I drew from my findings are presented in section 5. 2. Pragmatics and Grammar: cross-linguistic variable pragmatic foundations of grammars in individual languages The division between Pragmatics and Grammar has received focused attention from functionally oriented linguists (e.g. Givo´n, 1979; Matsumoto, 1997; DuBois, 2003; Ariel, 2008) as well as from pragmaticists of more formal linguistic persuasions (e.g. Sperber and Wilson, 1996). However, relatively little attention has been paid to the issue of whether or not, and how languages differ in terms of the division of labor between Pragmatics and Grammar. One of the few exceptions was a programmatic analytical framework ‘‘Typological Discourse Analysis’’ proposed by Myhill (1992:1–2) who defined it as below: ‘‘. . .typological discourse analysis is the cross-linguistic study of the factors affecting the choice of one construction or another in a given language, taking the surrounding discourse context into consideration as having a crucial effect on this choice. The goal of this research is to develop a language-universal framework for characterizing linguistic functions which will make it possible to describe objectively the function of a given alternation in a given language, concretely relate it to alternations in other languages, and ultimately use these descriptions and comparisons to develop a cognitively based language-universal theory of the relationships between form and function.’’ This study does not conflict with the basic assumptions of Typological Discourse Analysis and pays close attention to the cross-linguistically differential pragmatic/interactional foundations of grammatical structure. This study thus views a grammatical construction as an interactional resource employed in a specific socio-cultural situation with particular discursive-pragmatic functions and communicative/rhetorical effect. Crucially, this study postulates that the pragmatic/ interactional functions of a grammatical construction arguably vary considerably between different speech communities as a reflection of differing interactional practices. Among various factors that potentially influence the cross-linguistically differential divisions of labor between Pragmatics and Grammar, this study recognizes the prominent role played by pragmatic inference/enrichment in interpreting a grammatical form. Ariel (2008:159–160) argues this in her statement on pragmatic enrichment’s role in grammaticalization: ‘‘Finally, we come to pragmatic enrichments as a source of linguistic change. A significant factor in the grammaticalization of pragmatics is the fact that although we use the existing grammar to produce our utterances, grammar always leaves out (i.e. does not provide an obligatory code for) some of our intended meanings. We routinely rely on our addressees’ ability to draw inferences, which they perform based on the semantic meanings we do provide them with. Some of these inferences are one-offs. Many are not. Often, the message conveyed by the code augmented by the inference is useful to other speakers as well. Moreover, even if linguistic utterances are not underdetermined, the states of affairs about which utterances are predicated are far richer in detail, so addressees may associate an uncoded but concomitant interpretation with the meaning of the form (. . .).’’ This study argues that pragmatic enrichments are not only operative in language change and grammaticalization, but also account for the availability of conventional interpretations of a grammatical construction. Noun-modifying clauses in Japanese (e.g. (1) and (2)), which have been extensively discussed by Teramura (1992) and Matsumoto (1997), are a prime

K. Horie / Journal of Pragmatics 44 (2012) 663–679

665

example of a grammatical construction whose uncoded yet conventional interpretations are made possible by pragmatic enrichments. (1)

[atama-ga

yoku naru]

kusuri

head-NOM

well become medicine

(a) ‘the medicine which one become intelligent’ (literal meaning) (b) ‘the medicine which causes one to become intelligent’ (enriched interpretation) (2)

[toire-ni ike-nai]

komaasyaru

toilett-to go:POTENT-NEG

commercial

(a) ‘the TV commercial which (one) cannot go to a restroom’ (literal meaning) (b) ‘the TV commercial which (is so interesting by itself that one) cannot (may fail to) go to a restroom’ during the break’ (enriched interpretation) These noun-modifying constructions are known to require pragmatically enriched interpretations, i.e. (1b) and (2b). These interpretations contrast with the literal meaning (1a) and (2a), which suggest that there is a potential semantic gap between the head noun and the modifying clause. For instance, one needs to figure out a causal relationship between the TV commercial and the (consequential) state of affairs in which one cannot go to a restroom. The relatively greater degree of availability of uncoded yet conventional interpretations to Japanese noun-modifying constructions is demonstrated by comparing these examples with their direct English counterparts (10 a, 20 a). (10 )

(20 )

(a)

?* the medicine which one becomes intelligent

(b)

the medicine which causes one to become intelligent

(a)

?*the TV commercial which one cannot go to a restroom

(b)

the TV commercial because of which one may fail to go to a restroom during the break

The causal/logical relationship between the head noun and the modifying clause need to be elaborated in English. It seems that Japanese belongs to a minority of those languages in which such ‘short-circuited’ (or ‘‘tanraku’’ as originally coined by Hideo Teramura (e.g. Teramura, 1992)) noun-modifying clauses are licensed, with Korean as a possible companion. My preliminary investigation reveals that pragmatic enrichment observed with Japanese noun-modifying constructions is largely available in Korean, while Mandarin Chinese and Marathi exhibit patterns like English.1 Pragmatic enrichment, which was observed with noun-modifying constructions that regularly occur in non-final (intrasentential) position, is also observed in nominal predicate constructions occurring in sentence-final position. As discussed extensively in Suzuki (1998, 1999, 2007), sentence-final position in SOV languages like Japanese is a hatching ground for pragmatically motivated grammaticalization phenomena such as the development of the intersubjective meaning of the quotative construction –tte in example (3) from Suzuki (2007:230): (3)

Moo

ikkai it-ta

kara

already once go:PST so

doraibu ni

wa

driving with TOP

zisin

mo tuiteru si

confidence also have

and

tte? TTE

‘Since (I) drove (all the way to San Francisco) once, I have confidence in my driving, also (just kidding).’ Suzuki proposes that –tte in (3) is a recent use of ‘‘the speaker presenting her thoughts to the addressee as if it is uttered by someone else’’ (p. 230). Among the rich inventory of sentence-final expressions in Japanese are a group of ‘nominal’ predicates that index various pragmatic meanings, as shown in (4). These nominal predicates are composed of nominals with rather abstract or virtually no lexical meaning, such as no (genitive/nominalizer, virtually without lexical meaning), mono (‘‘concrete thing’’), koto (‘‘abstract matter’’), wake (‘‘reason, cause’’), and the copula da, the latter of which may be absent. (4)

Sentence-final nominal predicates in Japanese no-(da) (providing an explanation based on some contextual evidence), mono-(da) (moral obligation, past habit), koto-(da) (advisement, directive), wake (da) (providing an explanation) etc.

Sentence-final nominal predicates play an important role in Japanese conversational interaction, as illustrated by example (5) from Hinds (1986:62–63, glosses and emphasis added, translations partially modified):

1 As pointed out by an anonymous reviewer, the on-going research by Comrie may suggest that the type of noun modification exemplified by (1) and (2) may not be so rare cross-linguistically (see, Comrie, 1996, 2002).

666

(5)

K. Horie / Journal of Pragmatics 44 (2012) 663–679

K:

A, honto.

Soide

sokode

umare-ta

wake.

oh really

well

there

be.born-PST

reason

‘Oh really. Well, (is it that) you were born there?’ A:

Sokode

umare-ta

wake.

there

be.born-PST

reason

‘(It’s that) I was born there.’ K:

Soo.

Sokode

umare-te

sokode

sodat-ta

wake.

oh

there

be.born-CONT

there

grow up-PST

reason

‘(Is it that) you were born there and grew up there?’ A:

sodat-ta

wake.

grow.up-PST

reason

‘(It’s that) I grew up there.’ Each utterance in this short conversational interaction ends with a sentence-final nominal predicate wake meaning ‘reason’, which serves to indicate the background information queried by the interlocutor K and presented by the interlocutor A. Prominent among these sentence-final nominal predicate constructions is the so-called noda construction whose primary functions are to signal relevance of some existing context (linguistic or non-linguistic) to the current situation and thereby to provide explanation as to why/how the situation is as it is, as in (6): (6)

Tyotto komat-ta a bit

(60 )

naa. Ginkoo-ga

become annoyed-PST SFP

zenbu simat-te iru

bank-NOM

all

nda.

be closed-CONT NODA

I’ve got a bit of a problem. It is that all the banks are closed. (Otake, 2010:60, glossed added)

In (6), nda (a variant of noda) serves to explain why the speaker is annoyed by proving its relevance to the fact that all the banks are closed. In this instance, noda corresponds to an English construction it is that (Otake, 2010), though such correspondence is not always guaranteed, as noda is functionally more versatile and is more extensively employed than it is that. See the following pair of examples from Maynard (1997:113, glosses partially modified). In this instance, noda does not translate directly into English. (7)

kono ame-ga

agare-ba

this

let up-if

rain-NOM

moo

sugu natsu

already soon summer

na be

no

daroo.

NODA

‘(lit.) It would be that, when the rains let up, it will be summer soon.’ (Abe, 1968:6) (70 )

Soon, when the rains let up, it would be summer. (Saunders, 1966:4)

Based on a quantitative parallel comparison between Japanese novels and their English translations, Maynard (1997:113) argues that ‘‘(t)here is a distinct preference for nominalized expressions, at least in some Japanese discourse. When original Japanese and English translations are compared, there are many cases where the Japanese writer uses nominal clauses, but the English translator does not’’ (emphasis added). The occurrence of nominal predicates such as noda in sentence-final position in Japanese is noteworthy in respect to the prototypical correlations of lexical categories (verbs, adjectives, nouns) (8) proposed by Croft (1991). (8)

Prototypical Correlations of Syntactic Categories (Croft, 1991:55) Syntactic Category Noun

Adjective

Verb

Semantic class

Object

Property

Action

Pragmatic function

Reference

Modification

Predication

In light of the prototypical pragmatic functions of lexical categories observed cross-linguistically, sentence-final position in SOV languages is typically a locus where verbal expressions are expected to occur to serve predicating functions. The fact that ‘nominal’ predicates like (4) occur in sentence-final position is thus a pragmatically marked phenomenon. This leads us to the following research questions: (I) To what extent is the preference for nominal predicate structure unique to Japanese? (II) Where does such preference in Japanese come from?

K. Horie / Journal of Pragmatics 44 (2012) 663–679

667

The following section looks into question (I) through a comparative pragmatic analysis of nominal predicate constructions that serve predicating functions in three East Asian languages: Japanese, Korean, and Mandarin Chinese. 3. Nominal predicates in Japanese, Korean, and Mandarin Chinese: Functional similarities and differences Japanese, Korean, and Mandarin Chinese have formally similar formulaic nominal predicate constructions which literally translate into English as ‘‘it is that’’. These constructions are similar in that they formally consist of the copulas and the nominalizers, as shown in (9): (9)

Japanese:  noda no = nominalizer, genitive, da = copula Korean:  kes-ita kes = nominalizer, lexical noun ‘‘thing’’, ita = copula Mandarin Chinese: shı`de de = genitive, nominalizer, shı` = copula

Interestingly, the three nominal predicate constructions, which show close structural correspondence, also exhibit functional parallelism in that they serve ‘‘to characterize or explain a situation by affirming or denying some presupposition, as opposed to simply reporting an event’’ (Li and Thompson, 1981:589). Specifically these constructions are employed in written and spoken discourse and serve to increase textual cohesiveness. This functional characterization, which was intended for the shı`de construction, applies to both noda and kes-ita. Examples (10)–(12) illustrate the typical use of three nominal predicate constructions: (10)

Boku

asita-wa

I

tomorrow-TOP come-NEG-SFP

ko-nai-yo.

Yoozi-ga

aru-nda.

business-NOM

exist-NODA

‘I will not come tomorrow. It’s that (=because) I have something to do.’ (11)

Ku-ka

wa-ss-ta.

Ku-ka

cengmallo

nay

he-NOM

come-PST-DECL

he-NOM

really

my

ap-ey

nathana-n

kes-ita.

front-at

appear-ATTR:PST

KES-ITA

(Noda, 1997:67)

‘He came. (It’s that) he really presented himself before my very eyes, as I’m reporting in disbelief.’ (Kim and Horie, 2009:282–283) (12)

wo (shı`)

gēn

n

kāiwa´nxia`o

de.

I

with

you

joke

GEN/NOML

COP

‘(The situation is that) I am joking with you.’ (as an answer to the addressee who failed to take the speaker’s joke as such) (Li and Thompson, 1981:591) In each instance, the writer/speaker employs the nominal predicate construction to explain why/how the current situation is as it is by relating the situation to some previously existing linguistic or extra-linguistic context. In order to examine the extent to which these nominal predicate constructions correspond to each other in actual use, I did a preliminary parallel text comparison of tokens of these constructions by using Japanese novels and their Korean and Mandarin Chinese translations. The tokens of the Japanese noda construction that I used were collected from a Japanese popular autobiographical novel Tottochan by Tesuko Kuroyanagi (1984) and I examined the extent to which they did or did not correspond with their Korean and Mandarin Chinese counterparts kes-ita and shı`de. For comparison I used a Korean translation (2000, Prometheus Publishing) and a Mandarin Chinese translation (2002, Nanhai Publishing Corporation). The results of our (one-sided) parallel corpus comparison are presented in (13): (13)

Correspondence between noda, kes-ita and shı`de in a parallel corpus comparison Japanese

Korean

Mandarin Chinese

465

107

61

Tokens common to Japanese, Korean, and Chinese: 10 Tokens common to Japanese and Korean but not to Chinese: 97 Tokens common to Japanese and Chinese but not to Korean: 51

668

K. Horie / Journal of Pragmatics 44 (2012) 663–679

Tokens common to Korean and Chinese but not to Japanese: 0 Tokens found only in Korean: 0 Tokens found only in Chinese: 0 Uniquely Japanese tokens: 307 The results in (13) demonstrate that, in spite of formal and functional similarities, there are fundamental quantitative differences between the three nominal predicate constructions in terms of their frequency of use in written texts. That is, the Japanese noda tends to be employed more frequently than the Korean kes-ita or the Mandarin Chinese shı`de. I must note that there is a greater degree of correspondence between noda and kes-ita than that between noda and shı`de in terms of frequency. Reflecting the overall structural similarities between the two languages, Japanese noda and Korean kes-ita manifest close functional parallelism in many instances (examples are due to Kim and Horie, 2009:283). In (14a, b), Japanese noda and Korean kes-ita are employed by the speaker/writer, not to simply report a situation, but to explain its relevance to some previously existing circumstances (or presupposition). (14)

(a)

Watasi-wa-ne,

anata-to

kekkonsi-yoo-to

zutto

I-TOP-SFP

you-with

marry-will-QT

long time

omot-te

ita -no.

think-CONT:PST-NODA ‘You know, I have been dreaming of marrying you for a long time’. (b)

Mancey

pelye-ss-e.

Kulayse

kunynag

sa-n

ke-ya.

touch

end-PST-SFP

so

just

buy-ATTR:PST

KES-ITA

‘(I) touched (it). So (I) just bought (it), you see.’ No (a variant form of noda) in (14a) is used when the speaker makes a confession that she has wanted to marry him (= the addressee), which was not known to him until that point in speech. In a similar fashion, ke-ya (a variant form of kes-ita) in (14b) is used to express the reason that the speaker bought it (= the clothes) to her friend (= the addressee), which was not revealed until that point in speech. In contrast, there are cases where noda, but not kes-ita, is employed. These include contexts like (15) where noda serves to bridge a possible inferential gap between the current situation and the relevant linguistic or non-linguistic context that serves as evidence to make a bridging inference. (15)

A:

Zimen-ga

nure-te

iru.

ground-NOM

be wet- CONT:PRES

‘The ground is wet.’ B:

Ame-ga

hut-ta

*(nda).

rain-NOM fall-PST

NODA

‘It’s that (=This situation compels me to infer/conclude that) it rained.’ Such inferential evidential meaning is not manifest with kes-ita. Kes-ita is thus not felicitous in (16 (i)). Instead the bare verbal predication is preferred (16 (ii)). (16)

A:

Kil-i

cec

-eiss

-ney.

road-NOM be wet-CONT -SFP ‘The ground is wet.’ B:

Pi-ka

{(i) ?o-n

rain-NOM

keya/

(ii) wa-ss-e}.

come-ATTR:PST KES ITA come-PST-SFP

‘(i) It’s that (=This situation compels me to infer/conclude that) it rained.’ ‘(ii) It rained.’ Another noticeable difference between noda and kes-ita, of a more pragmatic nature, concerns interpersonal politeness strategies. Noda can be used to preempt a potentially face-threatening act (Brown and Levinson, 1978) aimed at an addressee, as a negative politeness strategy, as in (17a). The Korean counterpart kes-ita is not used in this context and the bare conclusive verb final ending form is used, as in (17b). (17)

(a)

Onegai-ga

aru-n

favor-NOM

exist-NO DA

desu.

K. Horie / Journal of Pragmatics 44 (2012) 663–679

(b)

Pwuthak-i

{(a)

favor-NOM

exist-POL/

iss-supnita./(b)*iss-nun

669

kes-ita.}

exist-ATTR:PRES

KES-ITA

‘I have a favor to ask of you.’ (Horie et al., 2007:5) In fact, the preferential tendency toward nominal predicate structure in Japanese is not limited to the use of noda. Kim (2003) presents a rather comprehensive contrastive study of preferential tendencies in Japanese and Korean toward ‘‘nounoriented’’ versus ‘‘verb-oriented’’ structures based on a (a one-sided) parallel corpus comparison of Japanese original texts matched with Korean translations. Kim found that there is a consistent tendency for a noun-oriented structure (nominal structure) to be employed in Japanese regardless of the grammatical function and the intra-sentential position of a structure, i.e. whether it is a subject NP (non-final position), an object NP,2 (non-final position), a modifier (non-final position), or a predicate (sentence-final position). In contrast, Korean leans toward ‘verb-oriented’ structure compared to Japanese, particularly in sentence-final predicate position. In order to understand precisely what are meant by ‘noun-oriented’ and ‘verb-oriented’ structures, consider the following pairs of Japanese and Korean sentences (18)–(21) from Kim (2003) (Romanization, glosses, and translations by the current author unless otherwise noted): (I) Subject/Object NPs (Subject NP) (18) Japanese original (a)

Iki-wa

ii

go:CONJ-TOP

good but

kedo, kaeri

-ga

sinpaida

return:CONJ-NOM

nee.

worrisome SFP

‘I’m not worried about going (there), but I’m more concerned about returning (from there).’ Korean translation (a’)

Ka-l

ttay-nun kwaynchanh-ciman,

tolao-l ttay-ka

go-FUT:ATTR

time-TOP good-but

return-FUT:ATTR-NOM worrisome-SFP

kekcengi-ya.

‘I’m not worried about the time when (I) will go (there), but I’m more concerned about the time when (I) will return (from there).’ (p.24) (II)

Modifier

Japanese original (19) (a)

Basyo osie-te

itadake

reba, watasi-no hoo-kara

place teach-GER

humbly receive

if

I-GEN

sensei-no

tokoro-ni oukagaisi

direction-from teacher-GEN place-to

masu.

humbly visit POL

‘If (you) let me know the place, I will pay a humble visit to your residence (lit. the place of the teacher).’

Korean translation (a’)

Cangso-lul malssumhay cwu-si-myen cey-ka

sensayngnim kyeysi

place-ACC tell:HON give-HON-if I(HUM)-NOM teacher chacapoyp-keyss-supnita. humbly visit-FUT-POL

-n

tey-lo

stay:HON-ATTR:PRES place-to

‘If (you) tell me the location, (I) will humbly visit where you live (lit. the place where the teacher resides.).’ (p. 40) (III) Predicate (20) Japanese original (a)

Kyoo-wa

sirabe

mono desu

ka?

today-TOP

investigate:CONJ

thing COP:POL

Q

2 In order to avoid potential terminological confusion over the term ‘object’, which can be used to refer to an ‘entity’ (semantically defined), as in (8), or to an ‘object of the verb’ (syntactically defined), I use the term ‘object NP’ to refer to the latter.

K. Horie / Journal of Pragmatics 44 (2012) 663–679

670

Table 1 The extent to which noun-oriented structure is used in Japanese and Korean (based on Kim 2003). Intra-sentential function/position

Japanese

Korean

(I) Subject NP (non-final position) Object NP (non-final position) (II) Modifier (non-final position) (III) Predicate (final position)

100% 100% 100% 100%

90.4% 92.8% 81.6% 60.4%

‘(lit.) Today is (there) something (for me) to look up?’ Korean translation (a’)

Onul-un

mwe

chacu-si-nun

ke3-ey

today-TOP

what

search-HON-ATTR:PRES

KES-ITA SFP:POL

yo?

‘Today what (is it that you) are looking for something?’ (p. 56) (21) Japanese original (a)

Sono sei de

okot-te ite

mo

warat-te iru

yooni mieru nda.

for that reason

get angry-CONT

even

laugh-CONT

like

look NODA

‘For that reason, even if (he) is angry, it is that (he) looks as if he were laughing.’

Korean translation (a’)

Kulayse

hwa-ka

na-ss-eto

for that reason

anger-NOM

come out-PST-even if

wus-ko iss-nun

keschelem

laugh-CONT-ATTR:PRES like

poye. look:SFP

‘For that reason, even if (his) anger comes out, (he) looks as if her were laughing.’ (p. 57)

In (18), while Japanese converts the verbs iku (‘to go’) and kaeru (‘to return’) into their derived nominal forms iki (‘going’) and kaeri (‘a return’), Korean employs the future/irrealis attributive forms of the corresponding verbs ka-l (go-ATTR:FUT; ‘who will go’) and tolao-l (return-ATTR: FUT; ‘who will return’). In (19), while Japanese fails to employ any verbal element and uses a possessive noun phrase sensei-no (‘teacher-GEN’), Korean elaborates by employing the past attributive form of a lexical subject-honorific verb kyeysi-n (stay:HON-ATTR:PRES; ‘who respectfully stays’). In (20), Japanese converts the verb siraberu (‘to search’) into its derived nominal form sirabe (‘a search’) and forming a noun compound sirabe mono (‘something to look up’) by combining it with a noun mono (‘thing’). Korean, in contrast, employs the present attributive form of a honorific verb chacu-si-nun (search-HON-ATTR:PRES; ‘who respectfully searches’) and combining it with an interrogative mwe (‘what’). In (21), while Japanese embeds the verb mieru (‘to be visible’) into a nominal predicate noda, Korean ends the sentence with the final verb form poye (to be visible: SFP). Thus, in each syntactic function and position (I)–(III), recurrent cross-linguistic contrasts are observed where Japanese noun-oriented structures are rendered into verb-oriented structure in Korean translations. Kim’s quantitative findings are presented in Table 1 (based on Table 12 in Kim (2003), p. 67, translation by the current author). Though Kim (2003) failed to mention this point, the declining degrees of preference toward noun-oriented structure in Korean (from (I) to (II) to (III)) is suggestive of the prototypical correlations of syntactic categories (6) mentioned earlier (repeated): (8)

Prototypical Correlations of Syntactic Categories (Croft, 1991:55) Syntactic category Noun

Adjective

Verb

Semantic class

Object

Property

Action

Pragmatic function

Reference

Modification

Predication

3 Please note that ke (> kes ‘thing’) here doesn’t directly correspond to mono (‘thing’) in Japanese. Instead, it is a variant of the nominal predicate kes-ita (see 14b, 16). What corresponds to mono in Korean is an interrogative mwe (‘what’).

K. Horie / Journal of Pragmatics 44 (2012) 663–679

671

Table 2 Attributive verb forms vs. Final verb forms in Classical Japanese and Modern Japanese. Classical Japanese Attributive form (i) oturu tori fall:ATTR bird ‘a falling bird’

vs.

Modern Japanese (ii) otiru tori fall:ATTR bird ‘a falling bird’

vs.

Final form Tori otu. bird fall:FIN ‘A bird falls.’

Tori-ga otiru. bird-NOM fall:FIN ‘A bird falls.’

The Subject/Object NP Position (I), the Modifier Position (II), and the Predicate Position (III) respectively are a locus where ‘object’-referring nouns, ‘property’-referring adjectives, and ‘action’-referring verbs are most prototypically expected to occur and serve their prototypical functions, i.e. ‘reference’, ‘modification’, and ‘predication’. The declining degrees of preference for noun-oriented structure in Korean, i.e. (I) (approx. 10%) to (II) (approx. 20%) to (III) (approx. 40%), can be interpreted as a reflection of the stativity-dynamicity (‘time-stability’) cline from the most static (or most time-stable) ‘noun’ to the most dynamic (or least time-stable) ‘verb’, with the category of ‘adjective’ situated in between (Givo´n, 1979). In other words, noun-oriented structure is most expected in the Subject/Object NP positions where nouns are the unmarked choice. It is less preferred in the Modifier position where adjectives are the unmarked choice and is least preferred in the Predicate position where verbs are the unmarked choice. Based on the contrastive observations summarized in Table 1, Kim presents the following generalization: ‘‘When representing approximately the same meaning in the same linguistic setting, the Japanese language exhibits higher preference for a noun-oriented structure, while the Korean language prefers a verb-oriented structure.’’ (p. 3, English translation by the original author, emphasis by the current author). Kim’s Japanese-Korean contrastive findings revealed that the preference for nominal structure in final predicate position in Japanese is quantitatively higher than that it is in Korean. The following section addresses the second research question ‘‘Where does such preference in Japanese come from?’’ from a historical pragmatic perspective. 4. The interactional origin of the nominal predicate structure in Japanese An inquiry into the preferential tendency toward sentence-final nominal predicate structure in Japanese requires historical linguistic consideration. Particularly relevant in this connection is a morpho-syntactic change in the history of Japanese referred to as the ‘‘merger of attributive and final verb forms’’ which happened in Middle Japanese (arguably completed by the 14th century), which is schematically illustrated in Table 2. In Classical Japanese, a morphological distinction was made between attributive and final forms of adjectives as well as of some (five out of seven) inflectional classes of verbs (see Takeuch, 1999:90). However, this distinction had ‘‘gradually disappeared, and by the 14th century, the old Final form of some verbs had been replaced by the Attributive form. In other words, the same (Attributive) form now functions both as the Final and as the Attributive’’ (Iwasaki, 2000:243; emphasis and the information in the parentheses added by the current author). As a result of this historical process, the morphological distinction between attributive and final verb forms, which existed with some inflectional classes of verbs, was virtually lost, as in Table 2. How and why this morphological change took place is a controversial research question in the field of Japanese linguistics. In what follows, I will present a brief historical overview of the factors leading to this change by drawing upon a description and examples presented in Iwasaki (2000). It is generally agreed that the extension of attributive forms to the sentenceconcluding function was mediated by a special rhetorical strategy, traditionally referred to as ‘‘kakari musubi’’ (focusconcord), of ending a sentence with attributive forms when some focus/emphasis particles are present in the same sentence, such as zo, namu, and ya, as in (22): (22)

Ware nomi

zo

kimi

ni

wa

kouru.

I

FP

you

DAT

TOP

yearn:ATTR

only

‘It is I alone who yearn for thee.’ In (22), the presence of a focus particle zo triggered the choice of the attributive form kouru instead of the regularly expected sentence-final form kou. This special rhetorical strategy came to facilitate the use of attributive forms in sentence-final position even when no focus particles were present. Such ‘‘attributive-ending’’ sentences originally served to convey the writer/speaker’s emotive stance toward the proposition expressed, e.g. a sense of joy, sorrow, and recollection, as illustrated in a poem in example (23):

K. Horie / Journal of Pragmatics 44 (2012) 663–679

672

(23)

Tiru

to

scatter QT

mite/

aru beki

see:CONJ

exist-should thing-PRT

mono wo/

ume

no hana //

utate nioi no/

sode ni

plum

GEN flower

very scent GEN sleeve LOC

tomare RU. lay-PERF:ATTR

‘It would have been best simply to watch them scatter – now, alas, the scent from the blossoms of the plum still lingers upon my sleeve.’ (Iwasaki, 2000:246–247, glosses partially modified) In (23), the attributive form ru (perfective auxiliary) occurs in sentence-final position where the final form ri is expected. Iwasaki (2000) proposes that ending sentences with attributive forms was quite distinct from ending them with final forms. Specifically he argues that ‘‘assertion is a feature that is present in the Conclusive finite sentence, while ending with the Attributive form is a way to suppress or background the force of the assertion’’ (Iwasaki, 2000:239, emphasis added). Attributive-final sentences such as (23) became more and more common and eventually led to the replacement of the final (conclusive) forms by the attributive forms. One of the pragmatic consequences of this morphological change in the history of Japanese is the loss of ‘privileged’ verb forms that serve to make assertion in sentence-final position, as illustrated in (iii) in Table 2, repeated as (24): (24)

Attributive form otiru

tori

fall:ATTR

bird

Final form vs.

‘a falling bird’

Tori-ga

otiru.

bird-NOM fall:FIN ‘A bird falls.’

This sharply contrasts with Korean, which distinguishes attributive and final verb forms systematically, as in (25): (25)

Attributive form ka-nun

salam

Final form vs.

Ku salam-i

ka-n-ta

go-ATTR:PRES person

that person-NOM go-PRES-FIN:DECL

‘a person who goes/is going’

‘That person goes.’

Table 3 illustrated the non-distinctiveness versus distinctiveness of attributive and final verb endings in Modern Japanese and Modern Korean. The fact that Japanese has merged these forms by expanding the functional domain of attributive forms is not insignificant in view of the pragmatic function of conclusive verb forms characterized by Iwasaki (2000), mentioned earlier, which I summarize in (26): (26)

Pragmatic Functions of Final and Attributive Forms (Iwasaki, 2000) Attributive forms: Suppressing or Backgrounding the Force of Assertion Final forms: Assertion

The functional expansion of attributive forms to sentence-final position means that the formal-functional opposition between attributive and final forms was lost. It is important to note in this connection that nominal predicates (e.g. noda), which are more prevalent in sentence-final position in Modern Japanese than in Korean (see Table 1), also typically serve to index some background or presupposed information, as seen in examples (15) and (17a) (repeated below):

Table 3 Attributive and final verb endings in modern Japanese and modern Korean. Tense

Attributive

Final

Past Present (non-past)

Modern Japanese -ta -(r)u

-ta -(r)u

Past Present (non-past) Future (intention)

Modern Korean -n -nun -l

(-a/e)ss-ta -n-ta -keyss-ta

K. Horie / Journal of Pragmatics 44 (2012) 663–679

673

Table 4 Frequency of sentence-final forms in three-minute segments of conversation among twenty Japanese pairs (Maynard, 1997:119, emphasis added).a Type

Number

Final particles Nominals Taglike Auxiliary Forms Verb (simple nonpast) Verb (gerund) Conjunctions Adverbial phrases Grammatical particles Verb (simple past) Fillers Nominalization Total

436 204 121 103 93 70 69 47 46 42 13 1244

% 35.05 16.40 9.73 8.28 7.48 5.63 5.55 3.78 3.70 3.38 1.05 100.03 (sic)

a

‘Nominals’ refer to nominal predicates such as no da, while ‘nominalization’ refers to instances of nominalizers such as koto ending a sentence by themselves.

(15) A:

Zimen-ga

nure

ground-NOM

be wet-CONT:PRES

-te iru.

‘The ground is wet.’ B:

Ame-ga

hut-ta

rain-NOM fall-PST

*(nda). NODA

‘It’s that (=This situation compels me to infer/conclude that) it rained.’ (17) (a)

Onegai-ga

aru-n

favor-NOM

exist-NO DA

desu.

‘I have a favor to ask of you.’ Nodas in (15) and (17a) do not serve to make assertion. Instead, they serve to index some background information, e.g. the preexisting ground condition (15) and the preexisting need for the speaker to make a request (17a). This suggests that there is arguably a preferential tendency to ‘‘background/presuppose’’ in sentence-final position in Modern Japanese and the function is typically carried out by nominal predicates such as noda. This preferential tendency in Modern Japanese is not unexpected in view of the morphological change discussed earlier (see Table 2).4 Through time, independent final verb forms, which typically serve to make assertion, were obliterated. Instead, attributive verb forms, which typically function to background/presuppose, were preserved and extended to function in a sentence-final position. It is an intriguing empirical question where this preference for final nominal predication in Japanese comes from. The collective wisdom of research in interactional and historical pragmatics, enriched by the emergent Emancipatory pragmatic perspective, informs us that the grammatical resources of a language have an interactional basis and are tailored to the communicative needs of the language community. In this respect, we draw particular attention to the observation made by Maynard (1997) regarding the frequency of sentence-final forms in the face-to-face conversational interaction of Japanese pairs, shown in Table 4. Table 4 points to the relatively high frequency of nominal predicate structure, i.e. ‘nominals’ (16.40%) such as noda. The frequency of nominal predicate usage accords with the preferential tendency toward nominal predicate usage mentioned earlier. No less noteworthy is the relative infrequency of ‘final’ conjugated verbs, i.e. ‘verb (‘simple past)’ (8.28%, e.g. kiku ‘(I, you, etc.) hear, (one) hears’) and ‘verb (‘simple non-past’) (3.70%, e.g. kii-ta ‘(I, you, one etc.) heard’), which are marked in bold in Table 4. Final conjugated verbs amount to only 11.98% of all the sentence-final forms. Maynard (1997:119) offers the following analysis of the tendency to avoid final conjugated verbs in Japanese conversational interaction: ‘‘This strategy of ending sentences with elements other than verbs helps make the sentence less final, implying that it remains open for interpersonal negotiation.’’ (emphasis added) In this connection, it is not insignificant to point out that simple final verbs were rather infrequent in Classical Japanese, as observed in Frellesvig (2010:53): 4 Iwasaki (1993) does argue that the pragmatic function of attributive-final sentences in Classical Japanese was later taken over by the nominal predicate noda in Modern Japanese. As pointed out by an anonymous reviewer, it is debatable whether such a historical process was really to be considered a loss of the original distinction followed by introduction of another way of making the distinction, or an overlap of two expressions with gradual transfer from the one to the other.

K. Horie / Journal of Pragmatics 44 (2012) 663–679

674

‘‘It should be noted, however, that it is quite rare to find simple verbs (without auxiliaries attached) in the conclusive forms concluding a main clause within the OJ (Old Japanese: KH) corpus; most occurrences of simple conclusive verb forms are found with extensions or particles.’’ (emphasis added) This suggests the possibility that the tendency to avoid bare final verbs has long been present throughout the history of Japanese. Table 4 convincingly demonstrates the variety of grammatical resources, including nominal predicates, that are tailored to the communicative need to avoid making a statement sound final or assertive. The majority of sentence-final forms are indeed elements other than final conjugated verbs. Particularly noticeable among these resources are highly frequent (35.05%) final interactional particles such as ne (monitoring the addressee’s response), yone (presenting the speaker’s opinion while monitoring the addressee’s response), kana (wondering about one’s thought/assessment by monitoring the addressee’s response).5 These interactional particles are ubiquitous in Japanese conversational interaction, as shown in a conversational segment (27) from ‘‘Mr. O. Corpus’’6: (27) J1 Task (‘‘Mr. O Corpus’’) L:

Kore-wa

demo

kokorahen ni hairi

masu yo ne, zettai,

this-TOP

but

hereabout to fit in

POL

SFP

tabun

absolutely probably

‘But this one will fit in around here, certainly, (or) probably, (as I think and I hope you agree with me).’ R:

Soo. Okot te iru

ndesu

yo ne.

right be angry:CONT

NODA

SFP

‘Right. (There is contextual evidence suggesting) that (it) is getting angry (as I think and I hope you agree with me).’ Nande okotta why

kao

get angry:PST face

site ru

ndesyoo

ne.

do:CONT

NODA:POL

SFP

‘Why is it that (he) is getting angry, (I wonder and hope you’re with me).’ Doo

si masyoo.

Saigo koko.

what do POL:PROP

at last here

‘What shall we do?

A final piece here.’

Mittu-wa

ii

no

ka na.

three pieces-TOP

good

NODA

SFP

‘These three pieces are OK, (I wonder, what do you think?)’ L:

Soo desu

yo

ne. Tabun.

right COP:POL SFP

probably

‘(I think) that is the case, probably (and hope you’re with me),’ R:

Soo

desu

yo ne.

right COP:POL SFP ‘(I think) that is the case (and hope you’re with me, too)’ Crucially these final particles serve to monitor the addressee’s attention and/or solicits her/his response, as illustrated in the italicized portions of English translations in (27). Based on his analysis of Mr. O. Corpus in Japanese, English, and Arabic, Katagiri (2009) proposes that, in Japanese conversational interaction, ‘‘participants seem to suppose that ideas are shared from the beginning, even before they are actually worked out’’ (emphasis added). The communicative strategy of making a sentence sound less final or assertive is arguably facilitated by such interactional style. Even when an assertion is made, it may be presented as if it were a piece of backgrounded/shared information. This naturally accounts for the preference for nominal predicate structure in Japanese. It must be noted, however, that the communicative need to make ‘‘the sentence sound less final’’ is not unique to Japanese conversational interaction. Maynard (1997:122) finds that the following array of ‘attitudinal phrases (extras)’ can occur in sentence-final position in English conversational interaction: 5 As pointed out by an anonymous reviewer, Japanese does have more ‘‘assertive’’- sounding interactional particles like yo and na, though these tend to be less frequent than collaboratively-oriented particles, particularly ne. 6 ‘‘Mr. O. Corpus’’ is a spontaneous language corpus consisting of video and audio-taped two-party conversations over a collaborative picture card arrangement task and one-person narratives. The languages covered include Japanese, English, Korean, Mandarin Chinese, and Libyan Arabic.

K. Horie / Journal of Pragmatics 44 (2012) 663–679

675

(28) (a)

Insertion phrases including ‘‘you know,’’ ‘‘right,’’ ‘‘OK.’’

(b)

Tag questions

(c)

Calling the listener by his or her first name

(d)

Adding phrases to express hesitation and ambiguity such as ‘‘or something,’’ ‘‘like’’

(e)

Ending the utterance with conjunctions like ‘‘though’’ and ‘‘but’’

Maynard’s (1997:123) study of sentence-final forms in three minute conversations among twenty American pairs revealed a remarkable cross-linguistic contrasts between English and Japanese. That is, the frequency of such ‘extras’ was found to be very low (10.07%), which sharply contrast with the high frequency of various sentence-final forms other than final conjugated verbs (88.02%). A contrast between Japanese sentence-final forms and their Korean counterparts would be expected to more subtle than a comparison of Japanese and English. Indeed, Korean conversational interaction abounds with a similar array of sentencefinal expressions to those listed in Table 4 in Japanese, e.g. tag-like auxiliaries, conjunctions (e.g. Kim, 2008). Their token frequency is also held to be higher than that it is in English. Even so, as we can infer from the differential degrees of preference for final nominal predicates (cf. Table 1), final conjugated verbs (e.g. (25)) are expected to occur more frequently in Korean than in Japanese. Conversely, nominal predicates (e.g. (14b)) are predicted to occur less frequently in final position in Korean than in Japanese. Furthermore, final particles (or suffixes) in Korean appear to convey different types of discursive meanings than their Japanese counterparts. Consider the following Korean conversational segment from the Mr. O. Corpus (29): (29) K2 Task (‘‘Mr. O Corpus’’) L:

Huyn sayk iss-ci.

Nolan sayk

white color exist-SFP

yellow color

iss-ci.

Kkaman sayk iss-canha.

exist-SFP

black

color exist-SFP

‘There’s a white one (and I’m committed to its truthfulness), a yellow one (and I’m committed to its truthfulness), and a black one.’ R:

Ung. A, kuleh-ney.

E, kuleh-kwuna.

right

be so-SFP

be so-SFP

‘Oh, I see (now I realize the fact based on some evidence). Yeah, I can see that (and I perceive it as new information).’ The Korean segment (29), similar to the Japanese segment (27), has various particles/suffixes occurring sentence-finally, e.g. –ci, -ney, -kwuna. However, unlike their Japanese counterparts such as ne and yo ne in (27), these particles are more directly tailored to the expressive needs of the speaker, i.e. the speaker’s degree of commitment to the proposition expressed (-ci), her/his factual realization based on some evidence (-ney), then newly perceived information (-kwuna), as illustrated in the italicized portions of English translations in (29) (see Horie and Taira, 2002).7 The relatively higher degree of preference for final verbal structure (see Table 1) in Korean is of particular interest from an interactional perspective. It suggests that the communicative need to make the sentence sound less ‘‘final/assertive’’ is not as imperative in Korean as it is in Japanese. The communicative strategy of making the sentence sound less final/assertive, is arguably related to the characterization of Japanese conversational interactions proposed by Katagiri (2009), based on his analysis of parallel conversational corpora (Mr. O. Corpus) in Japanese, English, and Arabic. As typically observed in a Japanese conversational excerpt (25), ‘‘participants seem to suppose that ideas are shared from the beginning, even before they are actually worked out.’’ (Katagiri, 2009, slide presentation). Korean conversational interactions (e.g. (29)), do not appear to lend themselves naturally to such characterization. Recruitment of more nominal (less verbal) forms to serve as sentence-final predicates is thus a recurrent process in the history of Japanese, suggesting that the communicative need to avoid making the sentence final has been consistently a high socio-cultural priority throughout the history of Japanese. A prime example of the recurrent historical process is the emergence of ‘attributive-final’ forms in Modern Japanese. Consider the attributive nominal adjective form mitaina in (30) and the attributively used verb form iu in (31), which serve to modify nouns:

7 Some Korean evidential markers, e.g. inferential -na pota (‘it seems that’) and -kes kathta (‘it seems that’), can be used to serve some interactional function of mitigating the force of assertion, as in Nwu-ka wa-ss-na pota (I-NOM come-PST-NA POTA; ‘I think someone is here.’ (Martin, 1993:705). However, these evidential markers are not usually considered to be sentence-final particles/suffixes unlike –ci, -ney, and -kwuna.

676

(30)

K. Horie / Journal of Pragmatics 44 (2012) 663–679

tyookoku

mitaina

nuigurumi

sculpture

be like: ATTR

stuffed doll

‘a sculpture-like stuffed doll’ (http://blogs.dion.ne.jp/jacoboo/ 2007.5.27) (31)

yuurei-ga

deru

ghost-NOM

appear QT

to

iu

yakata

say:ATTR

mansion

‘a mansion (where (people) say) ghosts appear’ Mitaina in (30) is the attributive form of a nominal predicate mitaida (‘be like’). Iu in (31) is the attributively used form of the verb iu (‘say’) which coalesces with the quotative marker to and forms a complementizer. To iu modifies a head noun (yakata ‘mansion’ in (31)) and serves to spell out the content of the head noun by introducing a quotative complement clause, as in (31). Recently, these attributive predicate forms were recruited to serve as final predicates with newly acquired discursive/ pragmatic meanings, as in (300 ) and (310 ): (300 )

Sensee-ga

tyoo

ozisan

na-no-ne.

teacher-NOM

extremely

old man

be:ATTR-NODA-SFP

Huransugo

nannka syaberen

no-ka-yo

mitaina.

French

such as

NODA-Q- SFP

MITAINA

speak

(. . .)

‘The teacher is extremely old (. . .) I was like, ‘‘Hey, can you speak French at all ?’’’ (Fujii, 2006:71) (310 )

Dare-mo

tasuke ya

who-even help

FP

si-nai

to iu.

do-NEG

TOIU

‘Nobody helped (me), allegedly.’ (Ohori, 2002:131; glosses and translation added) Mitaina in (300 ) serves to index the speaker’s ‘‘distancing him/herself from the authenticity of the quoted speech or thoughts’’ (Fujii, 2006:91). To iu in (310 ), in turn, serves to mystify the information source (‘‘anti-evidential’’; Ohori, 2002:131). The newly acquired pragmatic meanings of these ‘attributive-final’ forms are reminiscent of the ‘‘attributive-final sentences’’ (23) in Classical Japanese (repeated), (23)

Tiru

to

mite/

aru beki

mono wo/

scatter QT

see:CONJ exist-should thing-PRT

ume

no

hana //

plum

GEN flower

utate nioi no/

sode ni

very scent GEN sleeve LOC

tomare RU. lay-PERF:ATTR.

‘It would have been best simply to watch them scatter – now, alas, the scent from the blossoms of the plum still lingers upon my sleeve.’ (Iwasaki, 2000:246–247) Similar to Modern Japanese attributive-final predicates, Classical Japanese attributive-final sentences such as (23) were originally employed with special pragmatic meaning such as emotive stance. Such attributive-final forms are predicted not to occur in Korean, e.g. (32), in view of the rigid morphological distinction between attributive and final forms (23) (repeated): (32)

Ku salam-i

(25)

Attributive form ka-nun

{(a) *ka-nun./ (b) ka-n-ta.}

salam

Final form vs.

Ku salam-i

ka-n-ta.

go-ATTR:PRES person

that person-NOM go-PRES-FIN:DECL

‘a person who goes/is going’

‘That person goes.’

Recently, contrary to such prediction, an emerging sentence-final use of a quotative construction –ta nun in Korean can be observed, whose original attributive (noun-modifying) use is illustrated in (33):

K. Horie / Journal of Pragmatics 44 (2012) 663–679

(33)

salam-kwa

konglyong-i

hamkkey

human-with

dinosaur-NOM

together

sal-aa-ta-nun

677

cungke-tul

live-PST-DECL-PRES:ATTR evidence-plural ‘The (pieces of) evidence (saying) that dinosaurs lived with human beings’ –Ta nun consists of the plain speech level declarative ending -ta and the present attributive verb ending -nun-, a truncated form of an utterance verb -ha-nun. –Ta nun- in (33) modifies a head noun and serves to spell out the content of the head noun by introducing a quotative complement clause, similar to to iu in Japanese (31). As analyzed by Kim and Horie (2008), the final use of -ta nun is frequently observed in Korean Internet-based written communication, particularly in personal blog pages, as in (34) and (35) (the examples and analysis are due to Kim and Horie, 2008, see also Park, 2010): (34)

Hakchangsicel pihn-imye

kwikeli-mye

ilehkey

kulim-ul

school days

earrings-and

like this

picture-ACC

pin-and

kulye-se

nocem-ul

hay-ss-ta-nun.

draw-CONJ

street stall-ACC

do-PST- DECL-PRES:ATTR

(http//:blog.naver.com/dlqlsqk 2007.5.4) ‘I opened a street stall, drawing pictures of hair pins and earrings like this during my school days (saying. . .).’ (35)

Yang-i

manh-ase

twul-i

mek-taka

volume-NOM

much-CONJ

two-NOM

eat-CONJ

namkye-ss-ta-nun. leave-PST- DECL-PRES:ATTR (http://blog.naver.com/sthe2002 2007.5.4) ‘The volume (of the food we had) was too much for two of us, so we left (some), (saying. . .)’. The attributive-final –ta nun serves two distinctive pragmatic functions: (i) to quote or describe a writer’s own speech/ thought (including experience, feelings) as if it were someone else’s, as in (34), and (ii) to signal the information judged by the writer to be ‘unfamiliar/unexpected’ to him/herself and to readers, as in (35). The use of –ta nun is arguably motivated by the need to convey a sense of detachment in presenting personal information to the ‘anonymous’ cyberspace readership (function (i)) as well as by the need to highlight the newsworthiness of the information presented (function (ii)). The emerging -ta-nun construction in Korean suggests that the communicative need (e.g. to make a sentence sound less final/assertive) can even override the formal constraint imposed by the existing grammatical system. 5. Conclusions Informed by the insight of Emancipatory Pragmatics and Functional Typology, this paper has presented a comparative and historical pragmatic analysis of the historically persistent tendency for noun-oriented final structure in Japanese as compared to Korean and Mandarin Chinese. It addresses the following questions: (I) To what extent is the preference for nominal predicate structure unique to Japanese among these three East Asian languages? and (II) Where does such preference in Japanese come from? By adopting a new multi-dimensional pragmatic approach to grammar emancipated from the traditional disciplinary boundaries, this paper has suggested that the preferential tendency toward final nominal predication is more prominent in Japanese than it is in other East Asian languages. The preferential tendency toward nominal predicate structure, consistently observed in the history of Japanese, was contrasted with the preferential tendency toward verbal predicate structure in Korean. It was argued that the preferential tendency toward nominal predicate structure is motivated by the communicative need to make a sentence sound less final/assertive, arguably a high socio-cultural and interactional priority in Japanese interaction. Such communicative need is responsible for the grammaticalization of a host of sentence-final expressions in Japanese, including interactional final particles. A recent emerging use of an apparently anomalous ‘attributive-final’ construction in Korean suggests that such communicative need, even though comparatively less imperative in the linguistic community, can lead to some innovation in the existing grammatical system of the language.

K. Horie / Journal of Pragmatics 44 (2012) 663–679

678

Acknowledgements Thanks are due to Sachiko Ide, Yasuhiro Katagiri, Bill Hanks, Nick Enfield, Gunter Senft, Spring Ryan, Chi-chen Wei, Nathan Hamlitsch, the participants of the Emancipatory Pragmatics panel at the IPrA 2009, and, above all, to two anonymous reviewers whose extensive comments helped revise this manuscript immensely.

Appendix A. Abbreviations used in glossing

ACC

Accusative

ATTR

Attributive

CONJ

Conjunctive

CONT

Continuous State

COP

Copula

DAT

Dative

DECL

Declarative

FIN

Final

FP

Focus Particle

FUT

Future

GEN

Genitive

HON

Honorific

HUM

Humble

LOC

Locative

NEG

Negative

NOM

Nominative

NOML

Nominalizer

PERF

Perfective

POL

Polite

POTENT

Potential

PROPOS

Propositive

PRT

Particle

PRES

Present

PST

Past

Q

Question

QT

Quotative

SFP

Sentence Final Particle

TOP

Topic

References Abe, Kooboo, 1968. Tanin no Kao. (Face of Another).Sinchoosha, Tokyo. Ariel, Mira, 2008. Pragmatics and Grammar. Cambridge University Press, Cambridge. Brown, Penelope, Levinson, Stephen C., 1978. Politeness. Cambridge University Press, Cambridge. Bybee, Joan L., 2003. Cognitive processes in grammaticalization. In: Tomasello, Michael (Ed.), The New Psychology of Language, vol. 2. Lawrence Erlbaum Associates, New Jersey, pp. 145–167. Bybee, Joan, Revere Perkins, William, Pagliuca, 1994. The Evolution of Grammar: Tense, Aspect and Modality in the Languages of the World. University of Chicago Press, Chicago. Comrie, Bernard, 1996. The unity of noun-modifying constructions in Asian languages. In: Proceedings of the 4th International Symposium on Pan-Asiatic Linguistics, Institute of Language Culture for Rural Development Mahidol University at Salaya, Salaya, Thailand, pp. 1077–1088. Comrie, Bernard, 2002. Typology and language acquisition: the case of relative clauses. In: Ramat, Giacalone A. (Ed.), Typology and Second Language Acquisition. Mouton, Berlin, pp. 19–37. Croft, William, 1991. Syntactic Categories and Grammatical Relations. The Cognitive Organization of Information. University of Chicago Press, Chicago. DuBois, John, 2003. Discourse and grammar. In: Tomasello, Michael (Ed.), The New Psychology of Language, vol. 2. Lawrence Erlbaum Associates, New Jersey, pp. 47–87.

K. Horie / Journal of Pragmatics 44 (2012) 663–679

679

Ford, Cecilia, 1993. Grammar in Interaction. Cambridge University Press, Cambridge. Frellesvig, Bjarke, 2010. A History of the Japanese Language. Cambridge University Press, Cambridge. Fujii, Seiko, 2006. Quoted thought and speech using the mitai-na (‘be like’) noun-modifying construction. In: Suzuki, Satoko (Ed.), Emotive Communication in Japanese. John Benjamins, Amsterdam, pp. 53–96. Givo´n, Talmy, 1979. On Understanding Grammar. Academic Press, New York. Hanks, William, Sachiko Ide, Yasuhiro, Katagiri, 2009. Toward an emancipatory pragmatics. Journal of Pragmatics 41, 1–9. Hinds, John., 1986. Situation vs. Person Focus. Kurosio Publishers, Tokyo. Horie, Kaoru, Taira, Kaori, 2002. Where Korean and Japanese differ: modality vs. discourse modality. In: Akatsuka, Noriko, Susan, Strauss (Eds.), Japanese/ Korean Linguistics, vol. 10. CSLI, Stanford, pp. 178–191. Horie, Kaoru, Joungmin Kim, Mizuho, Tamaji, 2007. Where Japanese contrasts with Korean and Mandarin Chinese: intersubjectivity, modality, and the differential pragmatic-semantic foundations across languages. Studies in Pragmatics 9 (The Pragmatics Society of Japan) 1–16. Iwasaki, Shoichi, 1993. Functional transfer in the history of Japanese language. In: Clancy, Patricia (Ed.), Japanese/Korean Linguistics, vol. 2. CSLI, Stanford, pp. 20–32. Iwasaki, Shoichi, 2000. Suppressed assertion and the functions of the final-attributive in prose and poetry of Heian Japanese. In: Herring, Susan, et al. (Eds.), Textual Parameters in Older Languages. John Benjamins, Amsterdam, pp. 237–272. Katagiri, Yasuhiro, 2009. Finding parameters in interaction—a method in emancipator pragmatics Plenary lecture. In: 11th IPrA, Melbourne, Australia. July. Kim, Eunae, 2003. Nihongo no meisi koozoo to kankokugo no doosi koozoo.(The nominal-oriented structure of Japanese and the verbal-oriented structure of Korean). Choesn Gakuhoo 188, 1–83. Kim, Minju, 2008. A corpus-based study of the grammaticalization of the Korean connectives mye and mysense to sentence-final particles. In: Hudson, Endo M., et al. (Eds.), Japanese/Korean Linguistics, vol. 13. CSLI, Stanford, pp. 335–345. Kim, Joungmin, Kaoru Horie. 2008. The emerging sentence-final ‘attributive’ quotative construction in Korean web-based communication: a contrastive pragmatic study with Japanese. Paper presented at New Reflections on Grammaticalization 4, Leuven, Belgium, July 2008. Kim, Joungmin, Kaoru, Horie, 2009. Intersubjectification and textual functions of Japanese noda and Korean kes-ita. In: Takubo, Yukinori, et al. (Eds.), Japanese/Korean Linguistics, vol. 16. CSLI, Stanford, pp. 279–288. Kuroyanagi, Tetsuko, 1984. Tottochan. The Little Girl at the Window. Kodansha International USA, New York. Li, Charles, Thompson, Sandra A., 1981. Mandarin Chinese: A Functional Reference Grammar. University of California Press, Berkeley. Martin, Samuel E., 1993. A Reference Grammar of Korean. Tuttle, Tokyo. Matsumoto, Yoshiko, 1997. Noun-Modifying Constructions in Japanese: A Frame-Semantic Approach. John Benjamins, Amsterdam. Maynard, Senko K., 1997. Japanese Communication. University of Hawaii Press, Honolulu. Myhill, John., 1993. Typological Discourse Analysis: Quantitative Approaches to the Study of Linguistic Function. Blackwell, London. Noda, Harumi, 1997. No (da) no Kinoo. (The Function of No(da)).Kurosio Publishers, Tokyo. Noonan, Michalel, 1997. Versatile nominalizations. In: Bybee, Joan, John Haiman, Sandra, A. Thompson (Eds.), Essays on Language Function and Language Type. John Benjamins, Amsterdam, pp. 373–394. Ohori, Toshio, 2002. Ninti Gengo Gaku. (Cognitive Linguistics).University of Tokyo Press, Tokyo. Onodera, Noriko, Ryoko, Suzuki (Eds.), 2007. Journal of Historical Pragmatics 8 (2 Special Issue: Historical Changes in Japanese) . Otake, Yoshio, 2010. No (da)ni taioo suru Eigo no Koobun. (The Japanese no da-construction and the corresponding English constructions).Kurosio Publishers, Tokyo. Park, Junghee, 2010. Beheaded relativization for multimodal stance marking in reported speech in Korean internet communication. In: Iwasaki, Hoji, Clancy, Sohn, Sung-Ock (Eds.), Japanese/Korean Linguistics, vol. 17. CSLI, Stanford, pp. 455–469. Saunders, Dale E., 1966. The Face of Another. G.P. Putnam’s Sons, New York. Sperber, Dan, Wilson, Deirdre, 1996. Relevance. Communication and Cognition, 2nd edition. Blackwell, Oxford. Suzuki, Ryoko, 1998. From a lexical noun to an utterance-final pragmatic particle: wake. In: Ohori, Toshio, (Eds.), Studies in Japanese Gramamticalization. Cognitive and Discourse Perspectives. Kurosio Publisher, Tokyo, pp. 67–92. Suzuki, Ryoko, 1999. Grammaticalization in Japanese. A study of pragmatic particle-ization. Doctoral Dissertation. University of California, Santa Barbara. Suzuki, Ryoko, 2007. (Inter)subjectification in the 1uotative tte in Japanese conversation: local Change, utterance-ness, and verb-ness. Journal of Historical Pragmatics 8, 207–237. Takeuchi, Lone., 1999. The Structure and History of Japanese. From Yamatokotoba to Nihongo. Longman, London. Teramura, Hideo, 1992. Rentai syuusyokusetu to imi (The syntax and semantics of noun-modifying clauses). Teramura Hideo Ronbunsyuu: Nihongo Bunpoo hen. (Collected Works of Hideo Teramura: Japanese Grammar).Kuroshio Publisher, Tokyo, pp. 157–320. Traugott, Elizabeth Closs, 1989. On the rise of epistemic meaning in English. Language 65, 31–55. Traugott, Elizabeth Closs, 2003. From subjectification to intersubjectification. In: Hicky, Raymond (Ed.), Motives for Language Change. Cambridge University Press, Cambridge, pp. 124–139.