關(guān)聯(lián)理論中的模糊和接近

1、Book ReportVagueness and approximation in relevance theoryTan RuiContent1. Background of the paper2. The questions to be answered3.The content of the paper4. The theoretical basis of the paper5. Some important opinions in the paper6. The conclusion of the paper7. Reflections after reading the paper1

2、. Background of the paper Relevance theory (Sperber and Wilson, 1986) analyzes vagueness and approximation as instances of loose talk, which involve less than literal interpretations of thoughts. This approach has been developed by Reboul (1989), who tried to account for the interpretive use of conc

3、epts.2. The questions to be answeredThis paper evaluates these proposals and shows that they (S&W) leave many questions unanswered, especially about the very nature of interpretive use and the identification of speech acts. It is also claimed that vagueness and approximation should be dealt with

4、 separately.First, vagueness cannot be reduced to an instance of loose talk. When examining the evidential basis of vague utterances and when analyzing the interpretation processes at work, one realizes that it is necessary to posit the existence of vague concepts.2. The questions to be answeredSeco

5、ndly, in order to account for approximation, one has to rely on assumptions concerning the way people ordinarily speak about certain things.However, it is possible to elaborate, in the framework of relevance theory, an alternative account of vagueness and approximation which proves more satisfying e

6、ven though it does not succeed in solving all the problems involved.3.The content of the paper In this paper, the author shows that Sperber and Wilsons explanatory framework leaves unanswered many questions concerning the very nature of interpretive use and the identification of speech acts. He clai

7、ms that vagueness and approximation should be dealt with separately, since they involve two different kinds of concepts. However, he does not extend my criticism to the foundations of relevance theory, which he accepts here as a framework for his discussion.4. The theoretical basis of the paper S&am

8、p;W conceive utterance interpretation as involving code and inference and as being mainly ruled by the principle of relevance: “Every act of ostensive communication communicates the presumption of its own optimal relevance” (S&W, 1986: 158). This presumption of optimal relevance requires an opti

9、mal balance between contextual effects and processing effort: “(a) The set of assumptions which the communicator intends to make manifest to the addressee is relevant enough to make it worth the addressees while to process the ostensive stimulus. (b) The ostensive stimulus is the most relevant one t

10、he communicator could have used to communicate”(S&W, 1986:158)4. The theoretical basis of the paper 關(guān)聯(lián)理論認(rèn)為，話語(yǔ)的關(guān)聯(lián)程度依賴于語(yǔ)境效果和處理能力，語(yǔ)境效果與關(guān)聯(lián)成正比，處理能力與關(guān)聯(lián)成反比。并將把處理能力理解為認(rèn)知語(yǔ)言環(huán)境所消耗的腦力，關(guān)聯(lián)性越強(qiáng)，話語(yǔ)就越直接，認(rèn)知所消耗的腦力就越小，反之亦然，交際中說(shuō)話人對(duì)認(rèn)知負(fù)荷增減的利用就表現(xiàn)為一種交際策略的利用。5. Some important opinions in the paperA. S&Ws description o

11、f loose talkS&W argue that few utterances are strictly literal: most instances of communication involve approximate or vague uses, figurative meaning or indirectness. In order to account for this phenomenon, relevance theory resorts to the concept of interpretation. Consider the following two ut

12、terances, which respectively illustrate what is commonly called approximation and vagueness(S&W do not distinguish between vagueness and approximation, the author argues below that these phenomena involve different production and interpretation processes): 5. Some important opinions in the paper

13、(1)(Answering to a friend in chat. You know that you earn 797.32 pound pence a month)I earn 800 pound a month. (From S&W, 1986:233)(2)(Peter still has some hair) Peter is bald. (From Reboul, 1989:287)In (1), the speaker knows his exact salary but he assumes that the exact figure would be less re

14、levant to his friend than the approximate one. Indeed, the approximate utterance allows for lower processing effort, while implying the same assumptions, i.e. The same contextual effects, about the the speakers way of life.5. Some important opinions in the paperExample (2) was made popular by numero

15、us debates about the so-called sorites paradox(連續(xù)推理悖論). Someone, let us say X, can accept that a man with hair is bald, and also that if a man with no hair is bald, a man with one (0+1) hair bald too. Since this inferential step can be iterated infinitely, X eventually comes to the conclusion that a

16、 man who is very hairy is bald. S&Ws solution to this paradox reads as follows. Every concept is well defined(S&W do not claim that there are no vague concepts. However, none of their analyses involve the use of such concepts, which strongly suggests that “in their terms, there is no such th

17、ing as vague concepts”-Reboul, 1989:294), but can be use in a loose way. Consequently, the proposition in (2) is false literally, but the speaker uses it to interpret another thought which shares most of its contextual implications with it.5. Some important opinions in the paperB. On Gouvard (1995)

18、on approximate utterancesAccording to Gouvard (1995), S&Ws analysis of approximate utterance entails that the hearer has to reconstruct the initial thought of the speaker, who is supposed to have an exact knowledge. This would be, in fact, a very unfortunate consequence: obviously, in a communic

19、ative instance like (1), the speaker need not know the exact amount of his salary. Gouvard also points out that people use and expect their conversation partners to use approximate figures in accordance to socio-cultural norms.5. Some important opinions in the paperC. On Reboul (1989) on the interpr

20、etive use of conceptsReboul aims at generalizing S&Ws solution to the sorites paradox. Let us consider examples (3) to (7), taken from Reboul (1989: 287-297), which contain concepts like bald or dead, and scalar modifiers:(3)Peter is bald.(repeated= (2)(4)(Peter still has some hair.) Peter is ve

21、ry bald.(5)Peter is dead.(6)Peter is very dead.(7)Peter is less bald than Tom.5. Some important opinions in the paperScalar modifier like very may modify terms which express absolute concepts: dead is undoubtedly absolute in (5) or in (6), Reboul says, it cannot be used interpretively. Now, this fac

22、t turns out to be crucial if we remember that, according to S&W, bald expresses the well-defined or absolute concept of having no hair. The only difference between dead and bald would then be that the latter can have interpretive uses. Following S&W, Reboul concludes that not only bald, but

23、also qualitative concepts often classified as vague(intelligent, good , ect.) should be viewed as well defined concepts which are prone to interpretive uses. 6. The conclusion of the paper As far as the author concerned, S&Ws analysis of vague and approximate utterances as instances of loose tal

24、k is not entirely convincing. There is no good reason for grouping vagueness and approximation under the generic label of “l(fā)oose talk”. We have to assume the existence of vague concepts, while there is no similar necessity to posit approximate concepts: vague thoughts include vague concepts, but app

25、roximate thoughts include precise concepts.In order to account for approximation, one has to rely on assumptions concerning the way people ordinarily speak about certain things. One may wonder whether the principle of relevance will be able to explain such talking habits, which can become social-cultural norms. 6. The conclusion of the paper Althoug