a neo-gricean pragmatic analysis of chinese pragmatic markers

18

Click here to load reader

Upload: guangwu-feng

Post on 09-Sep-2016

226 views

Category:

Documents


5 download

TRANSCRIPT

Page 1: A neo-Gricean pragmatic analysis of Chinese pragmatic markers

Language Sciences 33 (2011) 417–434

Contents lists available at ScienceDirect

Language Sciences

journal homepage: www.elsevier .com/locate / langsci

A neo-Gricean pragmatic analysis of Chinese pragmatic markers

Guangwu Feng ⇑Faculty of English Language and Culture, Guangdong University of Foreign Studies, 2 Baiyun Road North, Guangzhou 510420, China

a r t i c l e i n f o a b s t r a c t

Article history:Received 7 October 2010Accepted 11 October 2010Available online 20 November 2010

Keywords:ChineseConventional implicaturePragmatic markers

0388-0001/$ - see front matter � 2010 Elsevier Ltddoi:10.1016/j.langsci.2010.10.005

Abbreviation: AdvM, adverbial marker (de); BA,complement marker (de); CMP, comparative; CSC, coemphatics; EXP, experiential aspect marker (guo);nominalizer; Q, question marker; PFV, perfective as2SG, second person singular pronoun; 3SG, third pe⇑ Tel.: +86 20 36207329; fax: +86 20 36209276.

E-mail address: [email protected] Grice did not use any technical term throughout

This article presents a pragmatic analysis of Chinese pragmatic markers in Chinese. Theanalysis is Gricean in spirit but differs from Grice (1989) in that it analyses pragmaticmarkers as conventional implicature potential bearers rather than conventional implicat-ures per se. A conventional implicature potential, attributable to the conventional meaningof a pragmatic marker, is stable and invariant across contexts. However, it cannot befleshed out into an actual conventional implicature until it interacts with a specific context.Unlike other analyses (e.g., Schiffrin, 1987, Bach, 1999; Blakemore, 2002), which detachthemselves from Grice’s (1989) classical account, this analysis retains Grice’s spirit thatthe meaning of pragmatic markers is part of what the speaker implicates. It is also notedin the analysis that conventional implicatures are subjective in nature.

� 2010 Elsevier Ltd. All rights reserved.

1. Introduction

Over the past decades, quite a number of analyses have been done on roughly the same class of linguistic expressionsgenerally called pragmatic markers. Different approaches have been taken, and predictably different outcomes have beenproduced as to the theoretical status of pragmatic markers (e.g., Schiffrin, 1987; Blakemore, 2002; Potts, 2005). This articlepresents a neo-Gricean analysis of Chinese pragmatic markers, which have been largely ignored in the literature.

In his enterprise to make a conceptual analysis of meaning, Grice (1989) analysed the meaning of a pragmatic marker(e.g., but, therefore, on the other hand) as conventional implicature.1 He suggested that the meaning of such a linguistic expres-sion solely determines the conventional implicature of the utterance in which it occurs. Thus conventional implicature is stableand context-insensitive. Unlike Grice, we analyse a pragmatic marker as just a bearer of a conventional implicature potential.This potential, attributable solely to its conventional meaning, is invariant over contextual changes. However, it cannot befleshed out into an actual conventional implicature until it occurs in a specific context, linguistic and/or non-linguistic. It followsthat the actual conventional implicature associated with the same pragmatic marker varies from occurrence to occurrence. Thisanalysis is, on the one hand, Gricean in spirit, and on the other hand, removes what Bach (1999) calls ‘‘a monkey wrench’’thrown into Grice’s endeavour to distinguish what is said from what is implicated. It also helps to place the category ofconventional implicature on the pragmatics side of the semantics/pragmatics divide.

. All rights reserved.

the ba marker in the ba construction; BEI, the bei marker in the bei construction; CL, classifier; CM,mplex stative construction (de); DEM, demonstrative; DUR, durative aspect marker (zhe and zai); EMP,FUT, future aspect marker (jiang); MM, modifier marker (de); MOD, modality; NEG, negative; NOM,pect marker (le); PRT, particle; 3PP, third person plural pronoun; 1SG, first person singular pronoun;rson singular pronoun.

his discussion of such expressions.

Page 2: A neo-Gricean pragmatic analysis of Chinese pragmatic markers

418 G. Feng / Language Sciences 33 (2011) 417–434

In what follows, I will first present a rough description of Chinese pragmatic markers. The description follows basicallyFeng (2008), where I offered a much more detailed description. I will then trace back to Grice’s (1989) provisional accountof similar linguistic expressions in English. Finally, in the bulk of the article I will demonstrate how conventional implicaturepotentials of various types of Chinese pragmatic markers develop into actual conventional implicatures in specific occur-rences of use.

For convenience of exposition, I use RPM for a sentence containing a pragmatic marker, p for a proposition, ‘+>’ to signify‘‘potentially conventionally implicate’’ and ‘+�’ to signify ‘‘actually conventionally implicate’’.

2. Chinese pragmatic markers: general properties and typology

A Chinese pragmatic marker, as I characterized in Feng (2008), is a lexical expression which (i) comments on the prop-ositional content of the matrix clause but does not affect its truth conditions, (ii) operates on the proposition(s) of the matrixclause rather than its constituents, (iii) is syntactically dispensable from the matrix clause but parasitic on its propositionalcontent.2

As I pointed out in Feng (2008), Chinese pragmatic markers can be classified into two broad types: conceptual and non-conceptual. A conceptual pragmatic marker encodes certain conceptual information. For example, xingyundeshi ‘fortunately’invokes the concept [FORTUNATE]. Pragmatic markers of this type provide the speaker with linguistic means to express hispersonal opinions about what he is saying.3 Though the conceptual message is not the key message of the utterance of which itis a part, it underpins the speaker’s attitude to the key message. A non-conceptual pragmatic marker encodes no conceptualmessage. For example, danshi ‘but’ does not encode the concept [BUT]. It is just an indication that a certain contrast holds be-tween the adjoining propositions.

Conceptual pragmatic markers can be grouped into two subtypes: epistemic and evaluative. An epistemic pragmatic mar-ker is essentially an indication of the speaker’s commitment to the degree of certitude (possibility, probability, certainty,actuality, reliability or limitations) of the propositional content.4

Structurally, epistemic pragmatic markers fall into two subcategories: epistemic phrasal pragmatic markers and epistemicadverbial ones. Prototypical members of epistemic phrasal pragmatic markers include: wo xiang ‘I think’, wo cai(xiang) ‘Iguess’, wo kan ‘I reckon’, wo xiangxing ‘I believe’, wo huaiyi ‘I suspect’ and a spectrum of others. Epistemic adverbial pragmaticmarkers mark affirmation of or emphasis on the factuality of the proposition. They can be further divided into two typesaccording to the degree of commitment: those which indicate certainty, and those that mark reservation or doubt. Prototyp-ical members of the first type include qishi ‘actually’, dangran ‘certainly’, dique ‘indeed’, kending ‘definitely’, queshi ‘indeed’,xianran ‘obviously’ and so on. Examples of the other type include dagai ‘probably’, and yexu/huoxu ‘perhaps’.

An evaluative pragmatic marker is essentially an indication of attitude, feeling, value judgement, or expectation with re-gard to the proposition expressed according to its effect on the speaker. A noticeable formal property of evaluative pragmaticmarkers is that they are composed of an adjective stem and a combination of an adjective marker de with the copula shi, asillustrated by xingyundeshi ‘fortunately’ and shangxindeshi ‘sadly’.

Evaluative pragmatic markers can be subcategorized into event-oriented and agent-oriented types. The two types can beformally distinguished. When the evaluation carried by the marker is oriented to the event narrated, it can be prefixed withlingren/rangren ‘making people’ or lingwo/rangwo (making me).

jingyadeshi ? lingren jingyadeshi ? lingwo jingyadeshishangxindeshi ? lingren shangxindeshi ? lingwo shangxindeshi

By contrast, when the evaluation is oriented to the agent with respect to what he/she has done, the marker indicating theevaluation cannot be prefixed with lingren/rangren or lingwo/rangwo, as shown in the following:

xingyundeshi ? �lingren xingyundeshi ? �lingwo xingyundeshibuxingdeshi ? �lingren buxing deshi ? �lingwo buxingdeshi

A contrastive pragmatic marker indicates a judgment that there is a sort of contrast between two propositions. The mosttypical members include dan/danshi ‘but’ (which has as its variants ke/keshi), raner ‘however’, que ‘but’, zhishi ‘but’, buguo‘but’, jiushi ‘but’, nalizhidao ‘however’, buliao ‘unexpectedly’, fanzhi ‘otherwise’, sui/suiran ‘although’ and its variant jinguan.

An elaborative pragmatic marker indicates that the proposition to which it is attached is elaborative/conclusive of ormore important than the previous proposition. Members of this type include bingqie ‘moreover’, zaishuo ‘moreover’, jiazhi‘furthermore’, tuierguangzhi ‘accordingly’, gengyoushenzhe ‘what’s worse’, chucizhiwai ‘in addition’, haiyou ‘besides that’,budan . . . erqie ‘not only . . . but also’, biru ‘for example’, yejiushishuo ‘namely’, zongeryanzhi ‘in a word’, (zongzhi)yijuhua ‘ina word’, tongyang ‘similarly’, tongli ‘analogously’, youqi ‘particularly’ and so on and so forth.

2 For the delimitation and a more detailed characterization of Chinese pragmatic markers, see Feng (2008).3 I henceforth use the masculine 3rd-person singular for the speaker and feminine 3rd-person singular for the hearer.4 This group of pragmatic markers is traditionally introduced as a type of hedging whose job is to make things more or less fuzzy (e.g., Lakoff, 1975) or

studied within the scope of epistemic modality in semantics (e.g., Lyons, 1977; Palmer, 1986/2001).

Page 3: A neo-Gricean pragmatic analysis of Chinese pragmatic markers

G. Feng / Language Sciences 33 (2011) 417–434 419

An inferential pragmatic marker is an indication of an inferential relation between the propositions involved.5 The mosttypical member of this type is suoyi ‘so’. It can be used either on its own or in collaboration with yinwei ‘because’ or the single-syllable variant yin, which introduces the premise.

3. Grice (1989)

Grice (1989) offered a provisional account of similar expressions in English. In his attempt to distinguish what is said fromwhat is implicated, he identified a category of implication, as illustrated in (1):

5

res6

co

(1)

Hallidayultive catFor mor

ntroversie

a.

and Haegories.e propes, see Fe

Smith is poor, but he is honest

b. There is some contrast between poverty or Smith’s poverty and his honesty

According to Grice, (1a) implies (1b). He labelled this category of implication ‘conventional implicature’ in order to dis-tinguish it from the category of conversational implicature. A conventional implicature enjoys the following definingproperties.6

Firstly, it is context-insensitive. In other words, the inference from (1a) to (1b) is determined by the conventional mean-ing of the word but rather than the general pragmatic principle and attendant maxims, and thus do not vary with context.

Secondly, it is non-truth-conditional. Namely, it does not make any contribution to the truth/falsity of the utterance. Inthe case of (1), even if (1b) is false (there is no reason in the world to contrast poverty with honesty, (1a) could still be true ifSmith is poor and he is honest.

Thirdly, it is non-cancellable. Therefore, it sounds anomalous to say ‘Smith is poor, but he is honest, but I do not mean tosuggest there is any contrast between poverty and honesty’.

4. Conventional implicature potential bearers: preliminary remarks

Grice was right to characterize the meaning of a pragmatic marker as non-truth-conditional. He was also right to claimthat conventional implicature cannot be cancelled by the speaker in normal circumstances. However, his claim that conven-tional implicature is context-insensitive is disputable. A pragmatic marker, on my view, cannot solely determine the conven-tional implicature of the utterance to which it is attached. It only bears a conventional implicature potential – what speakersof the language know it to conventionally mean. This potential, attributable solely to its conventional meaning, is stable andinvariant across contexts. However, it does not acquire an actual conventional implicature until it occurs in a specific con-text, linguistic and/or non-linguistic. In other words, the same pragmatic marker may give rise to different conventionalimplicatures in different contexts. For example, danshi ‘but’ bears a contrastive potential. This contrastive potential – theconventional meaning of the word type – though present with every occurrence of danshi ‘but’, needs to be contextually sub-stantiated into specific contrasts – what is conventionally implicated by the speaker on a given occasion of use. For an initialidea, look at the two utterances in (2):

(2)

a. Wo linju you si ge haizi, danshi jiali feichang anjing 1PS neighbour have four CL child but home very quiet My neighbour has four children, but the house is very quiet

b.

Zhang Hua mai le xin fangzhi, danshi qizi hen bu gaoxing Zhang Hua buy PFV new house but wife very NEG happy Zhang Hua bought a new house, but his wife was very unhappy

The conventional implicature potential danshi ‘but’ bears is the same in both (2a) and (2b). However, the actual contrasts

differ. The contrast in the first case is between the speaker’s assumption that the house should not be quiet and the fact thatthe house is very quiet, and in the second case between his assumption that Zhang Hua’s wife should be happy and the factthat Zhang Hua’s wife was unhappy. To figure out each contrast, one needs contextual information in addition to the con-ventional meaning of danshi ‘but’. Also illustrative are the utterances in (3)–(5):

(3)

Wang Hong1 shi zhongguoren, danshi ta1 buhui yong kuaizi

san (1976) refer to this group as causal covering result, reason and purpose. Quirk et al.

rties which are characterized in contrast to those of conversational implicature, seeng (2010).

Wang Hong be Chinese but 3PS unable use chopsticks

Wang Hong is a Chinese man, but he is unable to use chopsticks

(continued on next page)

(1985) posit a distinction between inferential and

Horn (2001), Huang (2007). For a discussion and

Page 4: A neo-Gricean pragmatic analysis of Chinese pragmatic markers

420 G. Feng / Language Sciences 33 (2011) 417–434

(4)

A: Jiaqi dai zai jiali? vocation stay at home You are staying at home during thevacation?

B:

Chunjie qian dai zai jiali, danshichunjie hou qu Hainan spring-festival before stay at homebut spring-festival after go Hainan I’m staying at home before the SpringFestival, but after that I’ll go toHainan

(5)

Yihandeshi, women 3 bi 2 ying, yaoshi 4 bi 2, women jinru juesai le unfortunately 1PP 3 versus 2 win if 4 versus 2 1PP enter final PFV Unfortunately, we won the game 3–2. If we had won 4–2, we would havegone up into the final

In (3), the contrast is between the speaker’s assumption that Wang Hong should be able to use chopsticks and the factthat he is unable to do so. For the hearer to discover this contrast requires encyclopaedic knowledge. Namely, sheshould know that Chinese people use chopsticks rather than spoons and folks. In (4), the contrast that B envisages isnot between the two propositions, but between the one with which the marker occurs and A’s assumption that Bmay stay at home during the vocation. In (5), what the speaker thinks unfortunate is that the team which he supportshas been knocked out. This is not directly stated in the segment that immediately follows the pragmatic marker. It isimplicated in the discourse context. These examples have demonstrated that though the conventional meaning of apragmatic marker is impervious to contextual influence, conventional implicatures are dependent of contextualinformation.

5. Chinese pragmatic markers: an in-depth analysis

In Section 2, I presented a typology of Chinese pragmatic markers. In Section 4, I proposed that pragmatic markers be ana-lysed as conventional implicature potential bearers. In what follows I will demonstrate how the conventional implicaturepotentials borne by various types of Chinese pragmatic markers develop into conventional implicatures in specific occur-rences of use.

5.1. Conceptual pragmatic markers

A conceptual pragmatic marker bears a conventional implicature potential indicative of a comment of some sort upon aproposition. This potential is invariant to contexts. Any utterance of a sentence containing a conceptual pragmatic markerpotentially implicates that the speaker has a comment on p. For instance, any utterance of a sentence containing haowuyiwen‘undoubtedly’ potentially implicates the speaker’s certainty of p, and any utterance of a sentence containingdaduoshuqingkuangxia ‘in most cases’ potentially implicates the speaker’s reservation on p. Actual conventional implicaturesdo not emerge until the pragmatic marker interacts with a specific context. For the hear to figure out the conventional impli-cature requires knowledge of not only the conventional meaning of the marker but also the proposition(s) involved, and inmany cases, the non-linguistic contextual information. Let me analyse in turn the two types of conceptual pragmatic markersas identified in Section 2.

5.1.1. Epistemic pragmatic markersAn epistemic conceptual pragmatic marker bears a conventional implicature potential suggesting an epistemic judge-

ment upon a proposition. Any utterance of a sentence with an epistemic pragmatic marker potentially implicates the speak-er’s judgement as to the degree of confidence in the likelihood of the proposition to which it is attached. This can bepresented schematically as in (6):

(6)

RPM(epistemic) +> the speaker’s thought as to the degree of likelihood of p An actual conventional implicature will not come up until p is fixed, and to figure out the conventional implicature the hearer needs to rely on the content of p (and/or other contextual information) in addition to the conventionalmeaning of the epistemic pragmatic marker. This can be illustrated by a consideration of (7) and (8):

(7)

a. RPM(haowuyiwen)

+> the speaker’s thought that the degree of likelihood of p is very high

Page 5: A neo-Gricean pragmatic analysis of Chinese pragmatic markers

7

G. Feng / Language Sciences 33 (2011) 417–434 421

I do

b.

not in

Haowuyiwen, xin jichang de jiancheng jiang jiakuai huanan de

undoubtedly new airport MM complete FUT accelerate South China MM jingji fangzhan economy development Undoubtedly, the completion of the new airport will accelerate the economic development of southern China. +� The speaker thinks it very likely that the completion of the new airport will accelerate the economicdevelopment of Southern China

(8)

a. RPM(yexu)

+> the speaker’s thought that the degree of likelihood of p is not very high

b.

Yexu, Li yijing shiqu le canjia aoyunhui de jihui perhaps Li already lose PFV participate Olympic Games MM opportunity Perhaps, Li has lost the opportunity to participate in the Olympic Games +� The speaker doesn’t think it very likely that Li has lost the opportunity to participate in the Olympic Games

RPM(haowuyiwen) bears a conventional implicature potential which suggests a very high degree of likelihood of p – its utter-ance type meaning. This potential is fleshed out, in the case of (7b), into an actual conventional implicature: the speakerthinks it very likely that the completion of the new airport will accelerate the economic development of Southern China –its utterance occasion meaning. Yexu ‘perhaps’ bears a conventional implicature potential which suggests reservation aboutthe likelihood of p – its utterance type meaning. This potential is enriched, in the case of (8b), into an actual conventionalimplicature: the speaker thinks that it is not very likely that Li has lost the opportunity to participate in the OlympicGames.

At this point, it is worth noting that Chinese epistemic pragmatic markers display a scale of speaker commitment to thefactuality of the situations described by the propositions to which they are attached. For example, the epistemic phrases con-sisting of the first person pronoun wo ‘I’ (which indicates the source of commitment) and the cognitive verbs form a scalefrom the strongest to the weakest commitment to the factuality of the situation described (see Fig. 1).

The epistemic gap between these markers is more noticeable when they occur with the same proposition. Consider theutterances in (9):

(9)

a. Wo jianxing tamen zai xiangai 1PS firmly believe 3PP DUR love They are in love, I firmly believe

b. Wo tuiduan tamen zai xiangai

1PS conclude 3PP DUR love They are in love, I conclude

c. Wo xiangxing tamen zai xiangai

1PS believe 3PP DUR love They are in love, I believe

d.

Wo xiang tamen zai xiangai 1PS think 3PP DUR love They are in love, I think

e.

Wo juede tamen zai xiangai 1PS reckon 3PP DUR love They are in love, I reckon

f.

Wo caixiang tamen zai xiangai 1PS guess 3PP DUR love They are in love, I guess

g.

Wo huaiyi tamen zai xiangai 1PS suspect 3PP DUR love They are in love, I suspect

The epistemic phrases above reflect different degrees of speaker commitment to the actuality of the proposition that they arein love.

Epistemic adverbs also display a commitment scale as shown in Fig. 2.7

tend both lists to be exhaustive.

Page 6: A neo-Gricean pragmatic analysis of Chinese pragmatic markers

Fig. 1. Commitment scale represented by epistemic phrasal pragmatic markers.

422 G. Feng / Language Sciences 33 (2011) 417–434

Consider the utterances in (10):

(10)

a. Haowuyiwen tamen zai xiangai undoubtedly 3PP DUR love Undoubtedly, they are in love

b.

Queshi tamen zai xiangai indeed 3PP DUR love Indeed, they are in love

c.

Xianran tamen zai xiangai obviously 3PP DUR love Obviously, they are in love

d. Duoban tamen zai xiangai

quite probably 3PP DUR love Quite probably, they are in love

e. Jushuo tamen zai xiangai

it’s said 3PP DUR love It’s said that they are in love

f.

Dagai tamen zai xiangai probably 3PL DUR love Probably, they are in love

g.

Yexu tamen zai xiangai perhaps 3PP DUR love Perhaps, they are in love

h.

Haoxiang tamen zai xiangai seemingly 3PP DUR love Seemingly, they are in love

Epistemic pragmatic markers exist in a language for the purpose of indicating one’s personal assessment of the degree ofreliability of a piece of knowledge or the likelihood of its being a fact. This is inevitable, just as one is bound to have a spatialangle of vision, because we are aware that not all knowledge is equally reliable.

5.1.2. Evaluative pragmatic markersAn evaluative pragmatic marker bears a conventional implicature potential suggesting an evaluation of some sort with

respect to a proposition. Any utterance of a sentence containing an evaluative pragmatic marker potentially implicates thatthe speaker has a personal evaluation with respect to the proposition expressed. This can be presented schematically as in(11):

Page 7: A neo-Gricean pragmatic analysis of Chinese pragmatic markers

Fig. 2. Commitment scale represented by epistemic adverbial pragmatic markers.

G. Feng / Language Sciences 33 (2011) 417–434 423

(11)

RPM(evaluative) +> the speaker’s evaluation with respect to p

An actual conventional implicature does not emerge until p is fixed. This can be illustrated by (12) and (13):

(12)

a. RPM(huajideshi)

+> the speaker’ thought that it is ridiculous that p

b.

Huajideshi, Zhang Hua cheng le zuqiu dui de jiaolian ridiculously Zhang Hua become PFV football team MM coach Ridiculously, Zhang Hua was appointed coach of the football team +� The speaker thinks that it is ridiculous that Zhang Hua has become the coach of the football team

(13)

a. RPM(yingmingdeshi)

+> the speaker’s thought that it is wise that p

b.

Yingmingdeshi, Zhang Hua liang nian qian ba gongsi mai le wisely Zhang Hua 2 year before BA company sell PFV Wisely, Zhang Hua sold the company two years ago +� The speaker thinks that it is wise that Li sold his company two years ago

In (12), RPM(huajideshi) bears a conventional implicature potential that the speaker thinks that it is ridiculous that p. This po-tential is enriched, in the case of (12b), into the conventional implicature that the speaker thinks that it ridiculous that ZhangHua was appointed coach of the football team. In (13), RPM(yingmingdeshi) bears a conventional implicature potential that thespeaker thinks it is wise that p. This potential is substantiated in the case of (13b) into an actual conventional implicaturethat the speaker thinks that it was wise of Zhang Hua to have sold the company 2 years ago.

An evaluative pragmatic marker is suggestive of the speaker’s positive or negative evaluation of the proposition expressedor the agent in terms of what he did or will do. The adjective within the marker indicates the speaker’s positive or negativeevaluation. The following are some examples:

(14)

a. RPM(mingzhideshi)

+> the speaker’s thought that it is sensible that p

b.

Mingzhideshi, Zhang Hua qu nian mai le fangzi sensibly Zhang Hua last year buy PFV house Sensibly, Zhang Hua bought a house last year +� The speaker thinks that it was sensible of Wang Hong to have bought a house last year

(continued on next page)

Page 8: A neo-Gricean pragmatic analysis of Chinese pragmatic markers

424 G. Feng / Language Sciences 33 (2011) 417–434

(15)

a. RPM(yingmingdeshi)

+> the speaker’s thought that it is wise that p

b.

Yingmingdeshi, Zhang Hua zou le beifen wenjian wisely Zhang Hua do PFV copy file Wisely, Zhang has made a spare copy of the file +� The speaker thinks that it is wise of Zhang Hua to have made a spare copy of the file

(16)

a. RPM(gaoxingdeshi)

+> the speaker’s thought that it is happy that p

b.

Gaoxingdeshi, Zhang Hua zhaodao le gongzuo happily Zhang Hua find PFV job Happily, Zhang Hua has found a job +� The speaker is happy that Zhang Hua has found a job

(17)

a. RPM(yuchundeshi)

+> the speaker’s thought that it is foolish that p

b.

Yuchundeshi, Zhang Hua ba mima xiezai le xingyongka shang foolishly Zhang Hua BA pin number write PFV credit card on Foolishly, Zhang Hua wrote the pin number on the credit card +� The speaker thinks that it is foolish of Zhang Hua to have written the pin number on the creditcard

(18)

a. RPM(jiaohuadeshi)

+> the speaker’s thought that it is cunning that p

b.

Jiaohuadeshi, Zhang Hua gai le mima cunningly Zhang Hua change PFV pin number Cunningly, Zhang Hua changed the pin number +� The speaker thinks that it is cunning of Zhang Hua to have changed the pin number

(19)

a. RPM(shangxindeshi)

+> the speaker’s thought that it is sad that p

b.

Shangxindeshi, Zhang Hua shiqu le gongzuo sadly Zhang Hua lose PFV job Sadly, Zhang Hua has lost his job +� The speaker thinks that it is sad for Zhang Hua to have lost his job

(14)–(16) are positive evaluations (as marked by the adjective within each marker: mingzhi ‘sensible’, yingming ‘wise’,

gaoxing ‘happy’) while (17)–(19) are negative evaluations (as marked by the adjective within each marker: yuchun ‘foolish’,jiaohua ‘cunning’ and shangxin ‘sad’).

To digress for a moment, whilst English evaluative pragmatic markers, which are identical in form with their verb phraseadverbs, can be used as fragmentary utterances as shown in the dialogue in (20), Chinese evaluative pragmatic markers, onthe other hand, cannot be used in such a way. This is shown in (21), though the adjective within the marker can be so used ifit is prefaced by an adverb of degree (e.g., hen ‘very’), as shown in (22) (due to Blakemore, 2002, p. 87).

(20)

Mum: Have you got much homework to do?

Son:

Unfortunately

(21)

A: Jintian you hen duo zuoye ma? today have very much homework Q Have you got much homework today?

B:

?Buxingdeshi Unfortunately

(22)

A: Jintian you hen duo zuoye ma? today have very much homework Q Have you got much homework today?

B:

Hen buxing Unfortunately
Page 9: A neo-Gricean pragmatic analysis of Chinese pragmatic markers

Back from the digression, RPM(conceptual) bears a conventional implicature potential, namely, the speaker’s thoughtwith regard to a proposition. While RPM(epistemic) potentially implicates assessment of the degree of likelihood of p,

and thus the degree of confidence in it, RPM(evaluative) potentially implicates positive or negative evaluation of a prop-osition. For actual assessments or evaluations (conventional implicatures of the utterance) to emerge requires actualcontexts.

G. Feng / Language Sciences 33 (2011) 417–434 425

5.2. Non-conceptual pragmatic markers

Traditionally, Chinese non-conceptual pragmatic markers are analysed as lianci or lianjieci ‘connectives’. For each connec-tive, a list of different functions is presented. A defect of such an analysis is that it ignores the contextual factors and thusconfuses its specific meaning in utterances with its stable conventional meaning. Contexts are inexhaustible, and thereforethe contextual meanings of a certain connective cannot be exhausted either. A more plausible account is to say that a con-nective has a basic meaning, which is constant, and that when it interacts with a particular context it acquires a particularmeaning, which is context-sensitive. On our view, a non-conceptual pragmatic marker bears a conventional implicature po-tential suggesting a perception of a connection between propositions. Any utterance of a sentence with a non-conceptualpragmatic marker potentially implicates that the speaker thinks that a connection holds between p and q. This can be pre-sented schematically as in (23):

8

(23)

Paralin

RPM(non-conceptual) +> the speaker’s thought that a connection holds between p and q

Actual conventional implicatures emerge when a sentence containing a particular conceptual marker is uttered in a specificcontext. In what follows, I will analyse in turn how each type of non-conceptual pragmatic markers interacts with contextsto bring forth actual conventional implicatures.

5.2.1. Contrastive pragmatic markersA contrastive pragmatic marker bears a conventional implicature potential indicating a contrast of some sort between p

and q. Any utterance of a sentence with a contrastive pragmatic marker potentially implicates that the speaker thinks (atleast at the time of speaking) there is a contrast between p and q. This can be schematized as in (24):

(24)

RPM(contrastive) +> the speaker’ thought that a contrast holds between p and q

For an initial idea, consider the following dialogue:

(25)

A:

guist

Ta1 laogong zhuyuan, ta1 que tiantian da majiang

3PS husband stay hospital 3PS but everyday play majiang Her husband is in hospital, but she is playing majiang everyday

B:

Ta weishenme bu neng da majiang? 3PS why NEG can play majiang Why can’t she play majiang?

In this dialogue, A conventionally implicates that he thinks that a contrast holds between one’s husband being in hospitaland her playing majiang (a popular traditional Chinese game) everyday. However, B does not share the idea. If the contrastivemarker que ‘but’ is removed, the contrast would disappear. This is why B’s retort in (250) does not make sense and the dia-logue sounds a bit odd.8

(250)

A:

ic fa

Ta1 laogong zhuyuan, ta1 tiantian da majiang

3PS husband stay hospital 3PS everyday play majiang Her husband is in hospital, she is playing majiang everyday

B:

?Ta weishenme bu neng da majiang? 3PS why NEG can play majiang Why can’t she play majiang?

However, a contrast can be read into A’s utterance in (250) even though it has no contrastive marker, as shown in

(26):

ctors such as intonation and tone need not concern us here.

Page 10: A neo-Gricean pragmatic analysis of Chinese pragmatic markers

426 G. Feng / Language Sciences 33 (2011) 417–434

(26)

A: Ta1 laogong zhuyuan, ta1 tiantian da majiang 3PS husband stay hospital 3PS everyday play majiang Her husband is in hospital, she is playing majiang everyday

B:

Ta tai bu xianghua le 3PS very NEG good PRT She is horrible

A:

Weishenme? Why Why?

B:

Ta yinggai tiantian zai yiyuan shou zhe 3PS should everyday in hospital attend DUR She should attend him in the hospital every day

Here B’s reply in the first conversational turn suggests that he has read a contrast into A’s utterance asserting her husbandbeing in hospital and her playing majiang every day. Obviously, this reading is not intended by A. Otherwise, he would nothave asked for clarification in the second turn. The requirement for clarification will be nonsensical when the contrastivemarker is used, as indicated in (260), in which A’s request for clarification sounds self-contradictory.

(260)

A: Ta1 laogong zhuyuan, ta1 que tiantian da majiang 3PS husband stay hospital 3PS but everyday play majiang Her husband is in hospital, but she is playing majiang everyday

B:

Ta tai bu xianghua le. 3PS very NEG good PART She is horrible

A:

?Weishenme? Why Why?

B:

Ta yinggai tiantian zai yiyuan shou zhe 3PS should everyday in hospital attend DUR She should attend him in the hospital every day

The above analysis has successfully demonstrated the contrast between p and q exists in the mind of a conceptualizer ratherthan the external world. In other words, it is subjective in nature. This can also be supported by the fact that the same prop-osition can be construed to be in contrast with different, or even opposite propositions, as shown in (27):

(27)

a. Zhang Hua shiqi sui, danshi hai xiang ge haizi Zhang Hua seventeen year but still like CL child Zhang Hua is seventeen years old, but he is still like a child

b.

Zhang Hua shiqi sui, danshi yijing xiang ge daren le Zhang Hua seventeen year but already like CL adult PFV Zhang Hua is seventeen years old, but he is already like an adult

Note that the propositions in the first half of the two sentences are identical, but the propositions which they are thought to

contrast with are in direct opposition. In (27a), the proposition that Zhang Hua is seventeen years old is construed to be incontrast with the proposition that he is still like a child (immature for his age), and in (27b), the same proposition is taken tobe in contrast with the proposition that he is already like an adult (more mature for his age).

The subjective nature of conventional implicature in association with a contrastive pragmatic marker is more noticeablewhen we contrast two utterances, which are propositionally equivalent, but different in that one contains a contrastive prag-matic marker, the other does not.

(28)

a. Wang Hong jingchang bangchu taren, ye bu xiwang dedao huibao Wang Hong often help others also NEG hope get reward Wang Hong often helps others, and he does not hope to get any reward

b.

Wang Hong jingchang bangchu taren, dan bu xiwang dedao huibao Wang Hong often help others but NEG hope get reward Wang Hong often helps others, but he does not hope to get any reward
Page 11: A neo-Gricean pragmatic analysis of Chinese pragmatic markers

G. Feng / Language Sciences 33 (2011) 417–434 427

(29)

a. Nainai hai jiankang, muqin ye fendan yixie jiawu grandma still healthy mother also share some housework My grandma is still healthy, and my mother shares some housework

b.

Nainai hai jiankang, dan muqin ye fendan yixie jiawu grandma still healthy but mother also share some housework My grandma is still healthy, but my mother also shares some housework

(30)

a. Yuanchu zhan zhe yi ge nüren, shi wo muqin far away stand DUR one CL woman be 1PS mother In the distance stood a woman, she was my mother

b.

Yuanchu zhan zhe yi ge nüren, que shi wo muqin far away stand DUR one CL woman but be 1PS mother In the distance stood a woman, but she was my mother

The two propositions in (28a) are potentially contrastive. The speaker expresses them as a simple conjunction because

he does not think that they contrast. In (28b) the speaker by using dan ‘but’ foregrounds his thought that offering muchhelp to others contrasts with not hoping for reward. In (29a), the speaker simply describes two states of affairs regardinghis grandmother’s health and her mother’s offer to do some housework. He does not show any personal comment. How-ever, in (29b) the speaker’s point of view is easy to be noticed: he personally thinks that while his grandmother is stillin good health, his mother does not need to help with the housework. In other words, the utterance conventionallyimplicates, by making use of the conventional meaning of dan ‘but’, that his grandmother’s good health contrasts withhis mother’s offer to share some housework. The difference between (30a) and (30b) is even more noticeable. In theformer, the speaker does not exhibit any personal state of mind towards the event described, while in the latter heenvisages a strong contrast between the fact that there was a woman standing in the distance and the fact that thewoman was his mother.

Notably, danshi ‘but’ can be used to display different degrees of subjectivity. Consider the utterances in (31), all of whichcontain the contrastive pragmatic marker danshi ‘but’.

(31)

a. Dongtian beifang hen leng, danshi nanfang hen nuanhe winter north very cold but south very warm In winter, it is very cold in the north, but it is very warm in the south

b. Wang Hong man tou bai fa, danshi jingli hen hao

Wang Hong all head grey hair but energy very good Wang Hong’s hair is all grey, but he is full of energy

c.

Zhexie nian ta chi henduo ku, danshi, zongsuan chenggong le these year 3PS suffer many hardship but after all succeed PFV He has suffered a lot these years, but he has succeeded after all

The conventional implicatures in association with danshi ‘but’ in the above utterances are all subjective in nature. Theirdifference is a matter of degree rather than of kind. In (31a), the speaker basically presents a direct contrast betweentwo weather situations in two different areas. The contrast involves the least degree of speaker involvement, and thusthe least subjective. In (31b), the contrast between having a head of grey hair (being old) and being energetic is lessdirect. It is what the speaker perceives to obtain. In (31c), the contrast between the hardships that he has sufferedand his final achievement of something is the least direct. It carries the greatest degree of speaker involvement, andis thus the most subjective.

To sum up, contrastive pragmatic markers can be used to demonstrate the speaker’s perception of contrastive relationsranging from the least subjective to the most subjective. The greater degree of subjectivity it is of, the greater the need for thepresence of a contrastive marker. Whatever degree, the choice of a contrastive pragmatic marker is not determined by thepropositions involved but by how the speaker conceives of them (whether the speaker has perceived a certain contrast). Thisis because two actual states of affairs cannot be said to clash at a real world level if they coexist in the real world.

To characterize Chinese contrastive pragmatic markers as conventional implicature potential bearers enjoys two advan-tages. In the first place, it better explains why Chinese contrastive pragmatic markers dan(shi), keshi and que can co-occurwith pure conjunctions like ji . . . you ‘not only . . . but also’ and yifangmian . . . linyifangmian ‘on the one hand, . . . on the otherhand’. Here are two examples.

(32)

a. Zhang ji xiang chuguo, you xiang baozhu muqian zhe fen gongzuo 3PS JI want go abroad also want keep present this CL job On the on hand, Zhang wants to go abroad, on the other he wants to keep his present job

(continued on next page)

Page 12: A neo-Gricean pragmatic analysis of Chinese pragmatic markers

9

sedo

428 G. Feng / Language Sciences 33 (2011) 417–434

Grice (nse, and: don’t m

b.

1989the oulti

Zhang ji xiang chuguo, dan you xiang baozhu muqian zhe fen gongzuo

3PS JI want go abroad but also want keep present this CL job On the on hand, Zhang wants to go abroad, but on the other he wants to keep his present job

(33)

a. Laoban yifangmian chengren wo you caigan, lingyifangmian shou wo boss one hand admit 1PS have talent other hand say 1PS bu neng luyong NEG can employ On the one hand, the boss admitted that I’m talented. On the other hand, he said that he cannot employ me

b.

Laoban yifangmian chengren wo you caigan, lingyifangmian que shou boss one hand admit 1PS have talent other hand but say wo bu neng luyong 1PS NEG can employ On the one hand, the boss admitted that I’m talented. But on the other hand, he said that he could not employme

The a-sentences in (32) and (33) conjoin two propositions, while the b-sentences not only conjoin the propositions, but alsobring in the speaker’s involvement. In (32b), by using dan ‘but’ in the second conjunct, the speaker foregrounds her personalthought that Zhang’s idea of going abroad contrasts with his idea of retaining his present job. In (33b), the use of que ‘but’suggests that the speaker is thinking the thought that his boss is self-contradictory in saying that he is very talented and thathe will not be employed.

In the second place, to characterize Chinese contrastive pragmatic markers as conventional implicature potential bearersis compatible with Grice’s general grounds of economy and his principle of Modified Occam’s Razer (Senses are not to bemultiplied beyond necessity).9 As we have demonstrated, the different shades of meaning of danshi/keshi ‘but’ are accommo-dated in terms of a single frame.

5.2.2. Elaborative pragmatic markersAn elaborative pragmatic marker bears a conventional implicature potential indicative of an elaborative relationship be-

tween propositions. Any utterance of a sentence with an elaborative pragmatic marker potentially implicates the speaker’sthought that the upcoming proposition is an elaboration or clarification of the foregoing proposition. This can be schematizedas in (34):

(34)

RPM(elaborative) +> the speaker’s thought that q is an elaboration of p

Three subtypes of Chinese elaborative pragmatic markers can be identified. First comes what I call the explanation type. Itcomprises a set of expressions that have been traditionally referred to in English as ‘paraphrastic phrases’, and alternatively‘explanatory markers’ (Greenbaum, 1969; Swan, 1985) or ‘appositive conjuncts’ (Quirk et al., 1985). Prototypical examplesare yejiushishuo ‘that is to say’, huanyanzhi ‘in other words’, and biru ‘for example’. Any utterance of a sentence with a para-phrastic pragmatic marker potentially implicates that the speaker thinks that p is an elaboration of q. This potential impli-cature can be substantiated into actual conventional implicatures in specific contexts. This can be exemplified by theutterances in (35b)–(37b):

(35)

a. RPM(yejiushishuo)

+> the speaker’s thought that q is an elaboration of p

b. Zhang shi jihuizhuyizhe, yejiushishuo, ta zongshi zai zuijia de shijian

1 1

Zhang be opportunist that is to say 3PS always at best MM time

chuxian zai zuijia de didian appear at best MM place Zhang is an opportunist, that is, he always turns up at the right place at the right time +� The speaker thinks that Zhang’s turning up at the right place at the right time explains Zhang’s being anopportunist

(36)

a. RPM(huanyanzhi)

+> the speaker’s thought that q is a clarification of p

, p. 291) said, ‘‘On general grounds of economy, I’m inclined to think that if one can avoid saying that the word so-and-so has this sense, thatther sense, or this meaning and another meaning, and if one can allow them to be variants under a single principle, that is the desirable thing toply senses beyond necessity.

Page 13: A neo-Gricean pragmatic analysis of Chinese pragmatic markers

G. Feng / Language Sciences 33 (2011) 417–434 429

b.

Fanyi shi neng zhiyi de difang jiu zhiyi, bu neng zhiyi de difang translate when can literal MM place JIU literal NEG can literal MM place jiu yiyi. Huanyanzhi, neng zhiyi jinliang zhiyi JIU translate meaning that is to say able literal try literal translate When translating, translate literally where you can. Where you cannot, translate its meaning. In other words,try to translate literally as much as you can +� The speaker thinks that trying to translate literally as much as one can explains translating literally whenone can and translating its meaning when one cannot

(37)

a. RPM(biru)

+> the speaker’s thought that q is an elaboration of p

b. Zhang de shenghuo xiguan hen hao. Biru ta jianchi duanlian,

1 1

Zhang MM life habit very good for instance 3PS keep exercise

congbu chouyan hejiu never smoke drink Zhang has a very clean life. For instance, he keeps doing physical exercises, and he never smokes and drinks +� The speaker thinks that Zhang’s keeping doing exercises and his neither smoking nor drinking explain hisliving a clean life

The second type consists of a set of expressions which indicates that q contains a more important or more relevant piece

of information than p. Typical members are erqie ‘moreover’, jiazhi ‘in addition’, and gengyoushenzhe ‘on top of it all’, etc. Anyutterance of a sentence containing such an expression potentially implicates that the speaker thinks that q is of greaterimportance or relevance than p. Here I would like to use erqie ‘moreover’ as a case in point.

(38)

a. RPM(erqie)

+> the speaker’s thought that q is of greater importance than p

b.

Wo buzhu zheli, li daxue tai yuan, erqie fujin mei gongche 1PS NEG live here from university too far moreover nearby NEG bus I will not live here. It is far from the university. Moreover, there are no bus services nearby +� The speaker thinks that a lack of bus services is of more relevance to his decision not to live there

It is worth noting here that erqie ‘moreover’ is often used in collaboration with budan ‘not only’. Thus (38b) can be changed

into (39) without any additional message.

(39)

Wo bu zhu zheli, budan li daxue tai yuan, erqie fujin mei 1PS NEG live here not only from university too far moreover nearby NEG gongche bus I will not live here. It is far from the university. Moreover, there are no bus services nearby +� The speaker thinks that a lack of bus services is of more relevance to his decision not to live there

In order to emphasize the relevance of the second message, an adverb lian ‘even’ can be used after erqie. Look at (40):

(40)

Wo bu zhu zheli, budan li daxue tai yuan, erqie fujin lian 1PS NEG live here not only from university too far moreover nearby even gongjiaoche ye mei you bus too NEG have I will not live here. It is far from the university. Moreover, there are even no bus services nearby +� The speaker thinks that a lack of bus services is of more relevance to his decision not to live there

The subjective nature of this group has gone unnoticed in the literature. In other words, they have been treated as simpleconjunctions (e.g., Xing, 2001). If they were, the propositions (p and q) could be used interchangeably with what is meanttotally unaffected. Consider the pair of utterances in (41):

(41)

a. Budan muqin chang shuohuang, erqie erzi ye chang shuohuang not only mother often lie but also son too often lie Not only the mother but also her son often lies

b.

Budan erzi chang shuohuang, erqie muqin ye chang shuohuang Not only son often lie but also mother too often lie Not only the son but also his mother often lies
Page 14: A neo-Gricean pragmatic analysis of Chinese pragmatic markers

430 G. Feng / Language Sciences 33 (2011) 417–434

These two utterances are propositionally equivalent. However, in (41a), the speaker personally thinks that the son’s lyingis of greater relevance (e.g., to the current topic) than his mother’s lying. In (41b), it is just the opposite. This difference can-

not be ignored.

The third type is what I call conclusion type. The most typical member of this type is zongzhi ‘in conclusion’. Variants in-clude zongeryanzhi and yijuhua. Any utterance of a sentence containing this elaborative marker potentially implicates thatthe speaker thinks that q is a conclusion from p. Actual conventional implicatures will not come up until it is used in a spe-cific context. This can be illustrated by (42):

(42)

a. RPM(zongzhi)

+> the speaker’s thought that q is a conclusion from p

b.

Tamen qidian bi women gao, women jinbu, tamen ye jinbu 3PL start point than 1PL high 1PL progress 3PL also progress zongzhi women yongyuan gan bu shang tamen in a word 1PL forever catch NEG up 3PL They have a much higher starting point than us. When we progress they progress as well. In a word, we cannever catch up with them +� The speaker thinks that we can never catch up with them because they have a higher starting point andthey progress when we progress

5.2.3. Inferential pragmatic markersAn inferential pragmatic marker indicates a sort of inferential relation between propositions. Any utterance of a sentence

with an inferential pragmatic marker potentially implicates that the speaker thinks that the upcoming proposition followsfrom the foregoing proposition or the other way round. This can be formulated as in (43):

(43)

RPM(inferetnial) +> the speaker’s thought that q follows p

The most prototypical member of this type of pragmatic markers is suoyi ‘so’, which can occur on its own (as shown in (44a))

or in collaboration with yinwei ‘because’ (as shown in (44b)):

(44)

a. Zhang Hua1 shi Zhongguoren, suoyi ta1 hen qinlao Zhang Hua be Chinese so 3PS very hard-working Zhang Hua is Chinese, he is, therefore, very hard-working +� The speaker thinks that Zhang Hua’s diligence follows from his being Chinese

b. Yinwei Zhang Hua shi Zhongguoren, suoyi ta hen qinlao

1 1

Because Zhang Hua be Chinese so 3PS very hard-working

Zhang Hua is Chinese, he is, therefore, very hard-working +� The speaker thinks that Zhang Hua’s diligence follows from his being Chinese

The two sentences above are identical in both propositional content and conventional implicature.At this point, the following three points merit mention. First, inferential relations can be ordered along a scale from min-

imal to maximal speaker involvement, i.e., the degree to which a rational participant is implicitly involved in the conceptionof the inferential relation. On the minimal side of the scale, the participant is only minimally involved in the conception,while on the maximal side, he has the maximal role to play in the conception. It seems that a single pragmatic markercan indicate different degrees of involvement. In other words, for inferential relations there is no one-to-one mapping be-tween a certain degree of speaker involvement and a particular pragmatic marker. For example, any instance of suoyi ‘so’indicates that the speaker thinks that there is an inferential relation between p and q. The utterances in (45)–(48) are illus-trative of the degrees of speaker involvement.

(45)

Wang Hong caidao xiangjiao pi, suoyi shuaidao le Wang Hong step banana peel so fall down PFV Wang Hong stepped on a banana peel, so he fell down

(46)

Wang Hong1 jia liang zhe deng, suoyi ta1 zai jia Wang Hong home on DUR light so 3PS at home The light in Wang Hong’s house is on, so he is in

(47)

Wang Hong1 shi Zhongguoren, suoyi ta1 hen qinlao Wang Hong be Chinese so 3PS very hard-working Wang Hong is Chinese, he is, therefore, very hard-working
Page 15: A neo-Gricean pragmatic analysis of Chinese pragmatic markers

G. Feng / Language Sciences 33 (2011) 417–434 431

(48)

Wang Hong1 hen qiong, suoyi ta1 hen chengshi Wang Hong very poor so 3PS very honest Wang Hong is very poor, therefore he is very honest

The degree of speaker involvement in (45) is minimal, because human reasoning is minimally effective there. What is effec-tive in the inferential relation is the actual state of affairs. The speaker recognizes rather than attributes an actual world cau-sal connection (stepping on a banana peel results in one’s falling down). The degree of speaker involvement is greater in (46)since the inferential relation is an inference of the speaker. There is no direct natural causal link between the light being on inone’s house and one’s being at home. The argument that the light in Wang Hong’s house is on cannot be taken as a real-worldcause for a conclusion that he is at home at the same time. It is temporary within the mental domain of the speaker. Theinferential relation in (47) is even more detached from real-world causality. The degree of speaker involvement in (48) ismaximal. It sounds a bit odd for many, but the speaker may have envisaged a sort of inferential relation between being poorand being honest. The more detached it is from the real world, the greater degree of speaker involvement it enjoys. The sameis true with yinwei ‘because’, which goes with a subordinate clause.

(49)

Wang Hong1 shuaidao le, yinwei ta1 caidao xiangjiao pi Wang Hong fall down PFV because 3PS step banana peel Wang Hong fell down because he stepped on a banana peel

(50)

Wang Hong1 zai jiali, yinwei ta1 jia liang zhe deng Wang Hong at home because 3PS home on DUR light Wang Hong’s is in because the light in his house is on

(51)

Wang Hong1 hen qinlao, yinwei ta1 shi zhongguoren Wang Hong very hard-working because 3PS be Chinese Wang Hong is very hard-working because he is Chinese

(52)

Wang Hong1 hen chengshi, yinwei ta1 hen qiong Wang Hong very honest because 3PS very poor Wang Hong is very honest because he is very poor

It can be seen from the analysis above that neither yinwei nor suoyi or yinwei . . . suoyi merely encodes an actual causal rela-tion. There is a degree of subjectivity inherent to them. In actual verbal interactions, the speaker makes use of them to dis-play different degrees of subjectivity.

A simple test for the subjectivity of the inferential relation is to see whether na jiushi . . .d e yuanyin ‘that’s . . . reason’ canbe used to substitute for suoyi ‘so’.

(490)

Wang Hong1 caidao xiangjiao pi, na jiushi ta1 shuaidao de yuanyin Wang Hong step banana peel that EMP 3PS fall down MM reason Wang Hong stepped on a banana peel, that’s why he fell down

(500)

?Wang Hong jia liang zhe deng, na jiushi ta zaijiali de yuanyin Wang Hong home on DUR light that EMP 3PS at home MM reason ?The light in Wang Hong’s house is on, that’s why he is in

(510)

?Wang Hong shi zhongguoren, na jiushi ta hen qinlao de yuanyin Wang Hong be Chinese that EMP 3PS very hard-working MM reason ?Wang Hong is Chinese, that’s why he is very hard-working

(520)

?Wang Hong hen qiong, na jiushi ta hen chengshi de yuanyin Wang Hong very poor that EMP 3PS very honest MM reason ?Wang Hong is very poor, that’s why he is very honest

The oddity of (500)–(520) is due to the fact that the inferential relations suggested by suoyi in (50)–(52) are not real world

ones. They exist, to a large extent, in the mind of the speaker. Therefore, to add wo renwei ‘I think’ to (500)–(520) would renderthem perfect utterances, given here as (53)–(55):

(53)

Wang Hong1jia liang zhe deng, na jiu shi wo renwei ta1 zai jia de Wang Hong home on DUR light that EMP be 1PS think 3PS at home MM yuanyin reason The light in Wang Hong’s house is on, that’s why I think he is in

(continued on next page)

Page 16: A neo-Gricean pragmatic analysis of Chinese pragmatic markers

432 G. Feng / Language Sciences 33 (2011) 417–434

(54)

Wang Hong1 shi Zhongguoren, na jiu shi wo renwei ta1 hen qinlao Wang Hong be Chinese that EMP be 1PS think 3PS very hard-working de yuanyin MM reason Wang Hong is Chinese, that’s why I think he is very hard-working

(55)

Wang Hong1 hen qiong, na jiushi wo renwei ta1 hen chengshi de yuanyin Wang Hong very poor that EMP 1PS think 3PS very honest MM reason Wang Hong is very poor, that’s why I think he is very honest

Wo renwei ‘I think’ can also be added to (490), which has the minimal degree of subjectivity. This is not strange because eventhe actual world causality is perceived via a human mind. As is pointed out by Searle (1992, p. 95), the actual world itself hasno point of view, but our access to it through our conscious states is always from our point of view.

As shown above, both souyi ‘so’ and yinwei ‘because’ can be used to indicate different degrees of speaker involvement, butby using the latter the speaker intends to minimize his involvement and maximize the actual causal relation between theconjoined propositions. In order to place emphasis on the actual cause, the speaker can use the emphatic adverb jiu ‘only’,zheng ‘just’ or zhengshi ‘just’ before the marker. When this happens, it is more detached from subjectivity.

(56)

Wang Hong1 shuaidao le, zhengshi yinwei ta1 caidao xiangjiao pi Wang Hong fall down PFV just because 3PS step banana peel Wang Hong fell down just because he stepped on a banana peel

However, these emphatics cannot occur with yinwei ‘because’ when they introduce obvious volitional inferences. Consider(57)–(59):

(57)

?Wang Hong1 zaijiali, zhengshi yinwei ta1 jia liang zhe deng Wang Hong home just because 3PS home on DUR light Wang Hong’s is in just because the light in his house is on

(58)

?Wang Hong1 hen qinlao, zhengshi yinwei ta1 shi zhongguoren Wang Hong very hard-working just because 3PS be Chinese Wang Hong is very hard-working just because he is Chinese

(59)

?Wang Hong1 hen chengshi, zhengshi yinwei ta1 hen qiong Wang Hong very honest just because 3PS very poor Wang Hong is very honest just because he is very poor

Secondly, there are four pragmatic markers which seem to specifically indicate a minimal degree of subjectivity. They arejieguo ‘as a result’, youyu ‘since’, yizhi ‘so that’ and its variant zhishi. They are primarily concerned with states of affairs inthe real world, thus it is inappropriate to use them when we make a volitional inference. For illustration, consider the fol-lowing three pairs of sentences.

(60)

a. Gua da feng jieguo shuzhi zheduan bushao blow big wind result tree branch break many It blew very hard. As a result, a lot of tree branches were broken

b. ?Wang Hong shi zhongguoren, jieguo hen qinlao

Wang Hong be Chinese result very hard-working ?Wang Hong is Chinese. As a result, he is very hard-working

(61)

a. Youyu gua da feng, shuzhi zheduan bushao because blow big wind tree branch break many Due to the strong wind, a lot of tree branches were broken

b.

?Youyu Wang Hong1 shi zhongguoren, ta1 hen qinlao because Wang Hong be Chinese 3PS very hard-working Because Wang Hong is Chinese, he is very hard-working

(62)

a. Gua da feng, yizhi shuzhi zheduan bushao blow big wind consequently tree branch break many It blew very hard. As a result, a lot of tree branches were broken
Page 17: A neo-Gricean pragmatic analysis of Chinese pragmatic markers

G. Feng / Language Sciences 33 (2011) 417–434 433

b.

?Wang Hong1 shi zhongguoren, yizhi ta1 hen qinlao Wang Hong be Chinese consequently 3PS very hard-working ?Wang Hong is Chinese. As a result, he is very hard-working

Finally, jiran ‘since’, which often collaborates with jiu, seems to specifically indicate a volitional inference. It displays a highdegree of subjectivity. Consider the examples in (63):

(63)

a. Jiran Wang Hong1 shi zhongguoren, ta1 jiu hen qinlao since Wang Hong be Chinese 3PS JIU very hard-working Since Wang Hong is Chinese, he is very hard-working

b.

?Jiran gua da feng, shuzhi jiu zheduan bushao since blow big wind tree branch JIU break many Since it blew very hard, a lot of tree branches were broken

In summary, like contrastive pragmatic markers, inferential pragmatic markers can also be used to mark different degrees ofsubjectivity. First, a single marker (e.g., yinwei, suoyi) can indicate different degrees of subjectivity (from the maximal to theminimal). Secondly, some inferential markers suggest minimal subjectivity (e.g., jieguo and yizhi), while others (e.g., jiran)can only be used for personal subjective inferences. However, no matter how subjective the speaker’s perception of the infer-ential relation between p and q is, the conventional implicature potential an inferential pragmatic marker bears are substan-tiated into different actual conventional implicatures in different contexts.

6. Conclusion

In this article, I have presented a neo-Gricean pragmatic analysis of Chinese pragmatic markers. I have argued that apragmatic marker only bears a conventional implicature potential, which requires linguistic and/or contextual informationfor substantiation. This analysis is Gricean in spirit, but it goes a considerable way beyond Grice’s (1989) provisional dis-cussion of similar linguistic expressions. It enjoys two advantages over Grice (1989). In the first place, Grice seemed toequate the semantic content of a pragmatic marker with the conventional implicature of an utterance, which consequentlybrings forth the question about whether conventional implicature is semantic or pragmatic. To analyse conventionalimplicatures as a contextually substantiated type of meaning helps remove what Bach (1999) calls ‘‘a monkey wrench’’thrown into Grice’s classical distinction between what is said and what is implicated. If conventional implicature is con-textually substantiated rather than stable and context-independent, it will, of course, go to the domain of what is impli-cated. In the second place, it is compatible with Grice’s principle of Modified Occam’s Razor (Senses are not to bemultiplied beyond necessity), because different shades of meaning of some pragmatic markers are accommodated witha single frame.

Acknowledgements

I would like to thank the anomalous reviewers for their comments and suggestions for revision. I would also like to ex-press my gratitude to Dr. Nigel Love, editor of Language Sciences, for his efforts in looking for reviewers for this article. Thewriting-up of this article is partly supported by the National Key Research Centre for Linguistics and Applied linguistics,Guangdong University of Foreign Studies, China.

References

Bach, K., 1999. The myth of conventional implicature. Linguistics and Philosophy 22, 327–366.Blakemore, D., 2002. Relevance and Linguistic Meaning: Semantics and Pragmatics of Discourse Markers. Cambridge University Press, Cambridge.Feng, G., 2008. Pragmatic markers in Chinese. Journal of Pragmatics 40, 1687–1718.Feng, G., 2010. A Theory of Conventional Implicature and Pragmatic Markers in Chinese. Bingley, Emerald.Greenbaum, S., 1969. Studies in Adverbial English Usage. Longman, London.Grice, H.P., 1989. Studies in the Way of Words. Harvard University Press, Cambridge, MA.Halliday, M.A.K., Hasan, R., 1976. Cohesion in English. Longman, London.Horn, L.R., 2001. A Natural History of Negation. CSLI Publications, Chicago.Huang, Y., 2007. Pragmatics. Oxford University Press, Oxford.Lakoff, R., 1975. Language and Women’s Place. Harper and Row, New York.Lyons, J., 1977. Semantics. Cambridge University Press, Cambridge.Palmer, F.R., 1986/2001. Mood and Modality. Cambridge University Press, Cambridge.Potts, C., 2005. The Logic of Conventional Implicatures. Oxford University Press, Oxford.Quirk, R., Greenbaum, S., Leech, G., Svartvik, J., 1985. A Comprehensive Grammar of the English Language. Longman, London.Schiffrin, D., 1987. Discourse Markers. Cambridge University Press, Cambridge.Searle, J., 1992. The Discovery of the Mind. MIT Press, Cambridge, MA.Swan, M., 1985. Practical English Usage. Oxford University Press, Oxford.Xing, F., 2001. Hanyu fuju yanjiu (Chinese Complex Sentences). The Commercial Press, Beijing.

Page 18: A neo-Gricean pragmatic analysis of Chinese pragmatic markers

434 G. Feng / Language Sciences 33 (2011) 417–434

Further reading

Barker, S., 2003. Truth and conventional implicature. Mind 112 (445), 1–34.Fraser, B., 1990. An approach to discourse markers. Journal of Pragmatics 14, 383–395.Fraser, B., 1999. What are discourse markers? Journal of Pragmatics 31, 931–952.Jucker, A., Ziv, Y. (Eds.), 1998. Discourse Markers. John Benjamins, Amsterdam.Koktova, E., 1986. Remarks on semantics of sentence adverbials. Journal of Pragmatics 10, 27–40.Lü, S., 1981. Xiandai Hanyu Babai Ci (Eight Hundred Words in Modern Chinese). Commercial Press, Beijing.Matsui, T., 2002. Semantics and pragmatics of Japanese discourse marker dakara (so in other words): a unitary account. Journal of Pragmatics 34, 691–867.Murillo, S., 2004. A relevance reassessment of reformulation markers. Journal of Pragmatics 36, 2059–2068.Nuyts, J., 2001a. Subjectivity as an evidential dimension in epistemic modal expressions. Journal of Pragmatics 33, 383–400.Nuyts, J., 2001b. Epistemic Modality, Language and Conceptualization: A Cognitive–Pragmatic Perspective. John Benjamins, Amsterdam.Risselada, R., Spooren, W., 1998. Introduction: discourse markers and coherence relations. Journal of Pragmatics 30, 131–133.Schiffrin, D., 2001. Introduction. In: Schiffrin, D., Tannen, D., Hamilton, H.E. (Eds.), Handbook of Discourse Analysis. Blackwell, Oxford.Wierzbicka, A., 1986. Introduction. Journal of Pragmatics 10, 519–534.Zhang, Y., 2000. Xiandai Hanyü Fuci Yanjiu (Adverbs in Modern Chinese). Xuelin Publications, Shanghai.