1Ba5b8el 47:2 158–174 © Fédération Internationale des Traducteurs (FIT) Revue Babel Ke Wen-li How Can Semantics Work to Help Translation Ke Wen-li Linguistics is the scientific study of language. Language is essentially a communication code consisting of symbols organized into a system, through which meaning is expressed. It is now generally agreed that semantics, a term used to refer to the study of meaning in language, is a component or level of a linguistic model in which semantics is at one “end” and phonetics at the other, with grammar somewhere in the middle (Palmer 1981: 5). The aim of serious semanticists is to explain and clarify the nature of meaning. (c) John Benjamins The usual meaning (or literal meaning) of a word or a sentence is different from the meDaniengli ivt hears eind ce rbtayin sIpnegcifeicn citracumstances. This point was made byo Fner:d iSnaandt ,d e1 S8au sNsuorev in 2hi0s 0dis6ti n0ct1ion:1 b1etw:4ee7n langue (lan- guage) and parole (speaking). This distinction reappeared in Chomsky as to: University at Illinois at Urbana-Champaign competence and performance, in Lyons as sentence meaning and utterance IP: 130.126.32.13 meaning, in Leech as semantics and pragmatics and in Guy Cook as semantic meaning and pragmatic meaning. Some symbolic logicians, Charles Morris and Charles Peirce, for example, have divided the study of meaning into three main parts: semantics, syntactics, and pragmatics. “Because of difficulties of terminology and definition, it is hard to pin down clear cases of semanticism and pragmaticism. In practice, one notices a preference of a semantic type of explanation to a pragmatic one, or vice versa” (Leech 1983: 6). Though the title of this article does not mention pragmatics, we approach meaning in its broadest sense in our discussion from a point of view which combines semantics and pragmatics. Translation, which, according to Jakobson, may be divided into three types: intralingual translation, interlingual translation, and intersemiotic translation, is eminently a communicative activity. As a process of conveying messages across linguistic and cultural barriers, translation should reflect accurately the meaning of the original text. Definitions of meaning can be almost as numerous and varied as the persons who have undertaken to discuss How Can Semantics Work to Help Translation 159 the subject. John Lyons draws a broad distinction between three kinds of meaning signalled by language: descriptive, social and expressive (Lyons 1977: 174). But according to Geoffrey Leech (1981: 9–23), “meaning” in the wider sense may embrace 1. conceptual meaning (or sense); 2. connotative meaning; 3. social meaning; 4. affective meaning; 5. reflected meaning; 6. collocative meaning; and 7. thematic meaning. To all the seven types of meaning listed, Leech gives the alternative term “Communicative Value.” A truly ideal translation would be expected to translate the whole communicative value of a text into another language. But since this is almost always impos- sible, conspicuously so with literary texts, preservation of one type of meaning is often achieved at the expense of other types of meaning-equivalence. The aim of translation is, in the ultimate analysis, to comprehend the original text and represent it in the target language. The translator, first of all, has to understand the meaning of the words and sentences, particularly anomalous sentences in the original text. A careful translator often has to weigh again and again before he decides which word, which sentence, and (c) John Benjamins what kind of sentence order or structure are to be used in the target language. Delivered by Ingenta This is because he is often confronted with more than one choice, more than one possibilitoy.n W:h Sy daoets, h1e 8ch oNosoe tvhi s2 on0e0 in6st e0ad1 o:f 1th1at :o4ne7? It seems that bteoyo: nUd pnerisvoenarl spirteydi leactti oInl ltihneroe iiss aalwta yUs rsobmaenthiang- Cthaht aliems bpehainidg hnis choice. And why is one version of translation better than the other? IP: 130.126.32.13 As has been mentioned earlier, semantics is the study of meaning. It is true that there is so little general agreement about the aims and precise nature of semantics, which is not yet a single, well-integrated discipline, and not a clearly defined level of linguistics, that much of the discussion is more philosophical than scientific, but interest in and importance attached to se- mantics has been more and more evidently shown in recent decades. Seman- tics is now burgeoning into a discipline of great promise. It is expected that explaining translation should in principle be among the goals of semantics. What follows is an attempt to probe into some aspects of semantic studies which may be of help to the understanding, explaining and solution of some of the problems in translation. Sense and reference Sense and reference are basic notions central to the study of meaning. According to F. R. Palmer (1981: 29), the term “sense” relates to the com- 160 Ke Wen-li plex system of relationships that hold between the linguistic elements them- selves (mostly the words); it is concerned only with intralinguistic relations; and the term “reference” deals with the relationship between the linguistic elements, words, sentences, etc. and the non-linguistic world of experience. Other terms used for the same or similar contrast are: “meaning” and “refer- ence;” “connotation” and “denotation;” “intention” and “extension.” The relationship between sense and reference can be briefly stated as follows: The referent of an expression is often a thing or a person in the world, whereas the sense of an expression is not a thing at all. Every expres- sion that has meaning has sense, but not every expression has reference. The words if, and, almost, and probable, for example, all have some sense, but none of them refers to a thing in the world. It should be pointed out here that reference is an utterance-dependent notion, and whenever we talk of an expression in a given sentence as having reference, we are assuming that the sentence in question has been, or could be, uttered with a particular communi- cative force in some appropriate context. In other words, whenever we say (c) John Benjamins that an expression in a particular sentence refers to a certain entity or group of Delivered by Ingenta entities, the term “sentence” is being employed in the sense of “text-sen- tence”, rathero tnha:n S“syastte,m 1-s8en tNencoe”v ( L2y0on0s 619 707:1 1:810)1. :47 to:T Uhen diivsteincrtsioint yb eatwte eInll isnenosies a nadt rUeferrebnacen ias -taCkehn adimffepreantliyg nby different authors and is not bound to any single philosophical theory of IP: 130.126.32.13 meaning. The classic example furnished by the German philosopher Gottlob Frege (1892), which is frequently used to illustrate the distinction between sense and reference, is “The Morning Star is the Evening Star”. As Frege pointed out, the two expressions “the Morning Star” and “the Evening Star” had the same reference (Bedeutung), since they each referred to the same planet Venus. But they could hardly be said to have the same sense (Sinn). For, if they did, “The Morning Star is the Evening Star” would become “The Morning Star is the Morning Star,” which would be tautological or analytic. Lyons (1977: 198) argues that “the Morning Star” and “the Evening Star” differ not only in sense, but also in reference. He believes that the rather uncertain status of the two expressions makes them less than ideal for the purpose for which they were used by Frege. Lyons prefers the example provided by the German philosopher Edmund Husserl: “the victor at Jena” and “the loser at Waterloo,” both of which expressions may be used to refer to How Can Semantics Work to Help Translation 161 Napoleon. But he admits that Frege’s example has been introduced to illus- trate in a general way the nature of his distinction between sense and reference. Other classic examples used by philosophers to illustrate the distinction, which are similar to “The Morning Star is the Evening Star” are: “Tully” and “Cicero” (Marcus Tullius Cicero 106–43 BC); “Pegasus” and “Medusa” (GK mythology: the Gorgon Medusa); “featherless biped” and “rational animal” (human beings), etc. The above examples will suffice to show that expressions may differ in sense but have the same reference. In other words, two or more expressions may have the same extensions, but clearly different intensions. Failure to understand this and to make the distinction between sense and reference or between intension and extension can lead to paradoxes and mistakes in translation. This is the crux of the matter concerning “the Morn- ing Star” and “the Evening Star.” One might ask how it could ever have been that people did not know that the Morning Star was the Evening Star, or that Venus was Venus. The point, of course, is that the extension (or referent) of (c) John Benjamins these two expressions is identical, but their intensions are different (though, Delivered by Ingenta in fact, the description is inaccurate since Venus is a planet, not a fixed star). Therefore, thoosne :w hSo afatil,e d1 t8o k Nnowo vth e2 c0or0rec6t e0xt1en:s1io1n o:4f t7he expressions wtoou:l dU nont ikvnoewr sthiatty th ea Mt oIrlnliinngo Sitsar aantd Uther Ebvaenninag -SCtarh waerme tphea saigmne. In literary works examples abound where expressions may differ in IP: 130.126.32.13 sense but have the same referent. It will be recalled that in The Mikado (1885) by Sir W. S. Gilbert and Sir A. Sullivan, the titles “First-Lord of the Trea- sure,” “Lord Chief Justice,” “Commander-in-Chief,” “Lord High Admiral,” “Master of Buck Hounds” and many others all refer to Pooh-Bah, a character in the opera who has been described as “the essence of cultivated diplomacy behind which lurks the basest of motives.” In the short story “Luck” by Mark Twain, “Y. C.” (Yeomanry Cavalry), “K. C. B.” (Knight Commander of the Bath), “that demi-god,” “the hero of the banquet,” “an absolute fool,” “this hero,” “young Scoresby,” “the poor youth,” “a wooden head,” “this donkey,” “such green and inadequate shoul- ders,” “the fellow,” “this immortal fool,” “the supremest ass,” and “a shining soldier” all refer to the very same person who is supposed to be called “Lieutenant General Lord Arthur Scoresby.” Careless translators or young entrants who are lacking in such knowledge as mentioned above are sometimes likely to get mixed up with the different expressions having the same referent, and as a result mistakes occur in 162 Ke Wen-li translation. For example, in “The Inspired Chicken Motel” by Ray Bradbury, a white inspired chicken laid an egg, on the shell of which were the words “Rest in Peace, Prosperity Is Near” written in white-calcium outline. A number of different expressions such as “an inspired chicken,” “the chicken,” “my precious pet,” “the white one,” “the fowl,” “her love,” “one grand bird,” etc., are used to refer to the inspired chicken. In translating the story into Chinese, however, some of the students wrongly rendered “her love” (refer- ring to her precious pet here) into ta de ai (literally: “her affection”) and “one grand bird” into yi zhi liao bu qi de niao (literally: a great bird). It is curious to note that in translating from one language into another, different expressions with the same reference in the source language do not always find their counterparts in the target language. In the case of English- Chinese translation, for example, we have TV, television set, telly, the tube, boob-tube, gogglebox, idiot box and (in a given context) the machine or instrument, etc., which are all translated into the single term dian shi ji (literally: television set) in Chinese. (c) John Benjamins Conversely, in Chinese-English translation, we have a number of differ- Delivered by Ingenta ent forms of address, such as Kong Fuzi, Kong Zi, Zhongni, Zhisheng Xianshi, Konog nSh:e nSgraent,, W 1e8nx uNanowvan g2, K0o0ng6 L 0ao1er:, 1D1ao:q4iu7, to refer to the staom:e Usinngliev peerrssoni,t ybu ta atll Ithllei nnaomiess aantd tUitlresb aaren ina m-Costh caasmes ptraansilgatned into “Confucius” in English, except in rare cases when “Kung Chiu,” “Kung IP: 130.126.32.13 Fu-tse,” “Confucius the Sage” are used. (See The American Heritage Dictio- nary of the English Language and Mathew’s Chinese-English Dictionary, revised American Edition) Hyponymy Hyponymy is one of the six paradigmatic relations of sense which, according to Lyons, include hyponymy, synonymy, incompatibility, complementarity, antonymy and converseness. Hyponymy involves us in the notions of “inclusion” or “subordination.” It indicates a relationship between two words, in which the meaning of one of the words includes the meaning of the other word, or in other words, it indicates the relation which holds between a more specific, or subordinate lexeme and a more general, or superordinate lexeme. Take “animal” and “cow” for example. The meaning of “animal” is included in the meaning of “cow.” This inclusion can be shown roughly by a diagram giving a list of the How Can Semantics Work to Help Translation 163 “meaning components” of “cow.” It will be seen that this list includes the component “animal” (see Hurford and Heasley 1983: 106). ANIMAL meaning of “animal” meaning of “cow” BOVINE  FEMALE But paradoxically, perhaps, if we draw a diagram of the extensions of “cow” and “animal,” the inclusion relationship appears the other way round. The set of all cows the set of all animals In the second sense of “inclusion,” we are actually talking about the reference of a term (the set of individuals or objects it refers to), and not the meaning (or sense), which we have( aclr)e aJdyo dhisncu sBseed neajraliemr. Iitn iss because of this inverse relationship between D“meealnivineg rinecldus ibony” aInndg “reefnerteantial inclusion,” which causes much confusion that it would be safer to avoid talking of inclusion on: Sat, 18 Nov 2006 01:11:47 altogether. Instead, we can use hyponymy, a term used by Lyons, Leech, to: University at Illinois at Urbana-Champaign Palmer and other authors, which has been gaining currency. The more spe- cific term (or the “lowerI”P te:rm 1) 3is 0ca.ll1ed2 th6e. h3yp2o.n1ym3 of the more general, and the more general (or the “upper” term) is called the superordinate term; so “cow” is a hyponym of “animal,” “daffodil” is a hyponym of “flower,” and “rose,” “tulip,” “daffodil,” etc., are co-hyponyms since each of them is a hyponym of “flower.” Hyponymy involves entailment, or in Lyons’ words, “hyponymy is defin- able in terms of unilateral implication.” To say that I saw a boy entails I saw a person and John stole a car entails John took a car. Here “boy” is established as a hyponym of “person,” and “steal” as a hyponym of “take” by virtue of the implications. We can formalise the relation of “boy” and “person” as (cid:1)x (B(x) → P(x)), and the relation of “steal” and “take” as (cid:1)x (S(x) → T(x)), where the upturned letter A stands for the universal quantifier (“for all”), x for an individual variable, and B, P, S, T for “boy,” “person,” “steal,” and “take” respectively. The definition of hyponymy in terms of entailment or unilateral implication enables us to define synonymy as a special case of hyponymy, i.e., bilateral or symmetrical hyponymy. 164 Ke Wen-li Hyponymy involves a transitive relation and reflects in a way the hierar- chical structuring of words. The following diagram which is adapted from Leech (1981: 112) can perhaps serve as an illustration. alsatian bird bird terrier fish dog spaniel animal insect horse etc. creature mammal cow human wolf vegetable etc. In this diagram, “alsatian” is an immediate hyponym of “dog,” and “dog” is an immediate hyponym o(f c“m) aJmomhaln,” “Bmaemnmjaal”m is ianns immediate hyponym of “animal,” and “animal” is an immediate hyponym of “creature,” so “alsatian” Delivered by Ingenta is a hyponym of “creature.” on: Sat, 18 Nov 2006 01:11:47 A better knowledge of hyponymy is not only important in language atcoqu:i sUitionni vbuet ralssoit qyu itae th eIllplfiunl otoi sth ea trta nUslartobr.ana-Champaign The same ChineseI Pter:m 1 n3on0gy.e1 i2n 6no.3ng2ye. 1sh3i guomin jingji de jichu (literally: Agriculture is the foundation of the national economy.) and nong- lin-mu-fu-yu huxiang jiehe de fangzhen (literally: the principle of combining agriculture, forestry, animal husbandry, side-occupations and fishery), for example, may refer to different hierarchical semantic levels. Nongye in the first statement which, according to The Dictionary of Modern Chinese, includes forestry, fishery, sideline occupation as well as farming and hus- bandry, is a superordinate term and therefore should be translated into English as “agriculture.” Nong(ye) in the second statement refers to the practice or specific activity of agriculture, or in other words, it is a hyponym of nongye, and therefore should be rendered into “farming.” Probably owing to the failure to make the distinction between superordination and hyponymy, A Chinese- English Dictionary of Neologisms (Beijing Language College Press, 1990) translates nong-lin-mu-fu-yu into “agriculture, forestry, animal husbandry, sidelines, and fishery,” which is an inappropriate rendition, if not wrong. A similar case is the use of “wine.” “Wine” as a hyponym normally refers to alcoholic drinks made from grapes or other fruits or grain, such as How Can Semantics Work to Help Translation 165 “red wine” or “white wine” mentioned in “Luck of Lottery” by Gordon Gaskill. But it can also be used as a superordinate term referring to all kinds of alcoholic drink in general such as “wine shop,” “wine bar,” “a small-time wine seller,” etc., and it can be so applied even if what the speaker refers to as spirits. A typical case is the use of “wine” to refer to spirits in Chapter 8 of the classic Chinese novel of manners Hong Lou Meng (A Dream of Red Man- sions or The Story of the Stone) in both English versions by the Yangs and Hawkes. Without making the distinction, some younger practitioners are likely to translate “white wine” into baijiu (spirits) or vice versa. In classical Chinese, either yue or dao , which is roughly equivalent to shuo (meaning “to say”) in contemporary Chinese, is used to express the action of “saying” something no matter what the implication of “saying” or what the attitude and feeling of the speaker is. If yue or dao was invariably translated into various verb forms of “to say” in English as some students tend to do, the translation would be monotonous and fail to convey the exact meaning. Actually either yue or dao or shuo is a superordinate term with a (c) John Benjamins large number of hyponyms, and such a notion can be applicable in translation. Delivered by Ingenta For example, in “A Merchant,” one of the Chinese fables from Yu Li Zi by Liu Ji (1311–137o5)n, t:h eSre aarte, a 1t le8a stN fiovev p l2ac0es0 w6he r0e 1yu:e1 is1 u:s4ed7 with different stpoea:k eUrsn inicvluedrinsgi tthye maetr cIhllainnt, othies f iashte rUmarnb aandn oath-eCr pheoaplme. Ipnsateiagd nof translating into (the merchant) “said,” (the fisherman) “said,” (someone) IP: 130.126.32.13 “said,” we can, according to the given context, render yue into (the merchant) “shouted,” (the fisherman) “complained,” (the merchant) “answered” (an- grily), (someone) “asked,” (the fisherman) “answered” (or “replied”) respec- tively. This way of rendition is theoretically feasible and well-founded. According to Lyons (1977: 292), when the relation of hyponymy holds between nouns, it is possible to use x and y in place of two nouns to express the hyponymy in the following formula: “x is a kind of y” (where x is a hyponym of y), for example, “A cow is a kind of animal,” “A tulip is a kind of flower.” Although this formula is not applicable to verbs, adjectives, adverbs and other parts of speech without prior nominalization, we can express the relation of hyponymy in some other way. As we can say “A cow is an animal of a certain kind,” so we can say (to take the above fable for example), though perhaps less idiomatically, “to shout/complain/ask/answer/etc. is to say something in a certain way.” The sense of “in a certain way” is encapsulated in the sense of “shout/complain/ask/answer/etc.” 166 Ke Wen-li Changes of meaning Language as a communication system is subject to constant change. When a language, such as classical Latin, ceases to change, we call it a dead language. Such change, which is constantly going on in a living language, can be easily seen in the vocabulary. Old words disappear, new words come into being, and existing words change their meaning throughout the history of a language. Any broadening in the sense of one word often involves a corresponding narrowing in the sense of one or more of other words. As Baugh and Cable (1978: 2, 307) put it, “That words do undergo such change is a fact readily perceived,” and “Change of meaning can be illustrated from any page of Shakespeare.” Nice in Shakespeare’s day, for example, meant foolish (Latin nescius, “ignorant”), and rheumatism signified a cold in the head. According to Bloomfield (1933: 426–7) changes of meaning can be classified into a number of types. These, together with an example and the earlier meaning, were adapted by Palmer (1981: 19): (c) John Benjamins Narrowing Dmeealitvered b“yf oIond”genta Widening bird “nestling” on: Sat, 18 Nov 2006 01:11:47 Metaphor bitter “biting” to:M Uetonnyimvye (nresarinteyss ian tsp aIcleli onr otimise) at Urbana-Champaign jaw “cheek” IP: 130.126.32.13 Synecdoche (whole/part relation) town “fence” stove “heated room” Hyperbole (stronger to weaker meaning) astound “strike with thunder” Litotes (weaker to stronger meaning) kill “torment” Degeneration knave “boy” Elevation knight “boy” In contrast to Bloomfield, Baugh and Cable (1978: 308) generalize from the changes of meaning four well-marked tendencies that words often pursue in their sense development: extension of meaning, narrowing of meaning, de- generation and regeneration. Extension of meaning is sometimes called generalization. The word proposition, for example, primarily means a statement set forth for purposes of discussion, or in mathematics, for demonstration. It was so used by Lincoln in his Gettysburg Address: “… a new nation dedicated to the propo- sition that all men are created equal.” But in America during the last century it How Can Semantics Work to Help Translation 167 began to be used more loosely. Owen Wister says in The Virginian, “Proposi- tion in the West does, in fact, mean whatever you at the moment please.” Today it is acceptable in America if we say “That’s a different proposition” to mean in more formal English “That’s a different matter.” We can also say “He is a tough proposition,” “He was the coolest proposition I ever met,” with proposition in the sense of person. Narrowing of meaning is for a word gradually to acquire a more re- stricted sense, or to be chiefly used in one special connection. A classic example in this respect is the word doctor. There were doctors, that is, learned men, in theology, law, and many other fields besides medicine, but nowadays when we send for the doctor we mean a member of only one profession. Narrowing of meaning may be confined to one locality under the influence of local conditions. Nickel in America means a coin, and democrat and republi- can seldom have their broader significance to an American, but rather imply adherence to one or the other of the two chief political parties in the USA. Degeneration of meaning may take several forms. For example, smug (c) John Benjamins was originally a good word, meaning neat or trim; its present suggestion of Delivered by Ingenta objectionable self-satisfaction seems to have grown up during the nineteenth century. A voerny : inStearetst,i n1g 8fo rNm oovf d2eg0en0e6ra ti0on1 o:1fte1n :4oc7curs in words atsoso:c iUatend iwviethr sthiitnygs awth iIclhl inareo nisot acotn sUidrerbeda npoalit-eC toh taalkm apboauti.g Fnor example, the common word for a woman’s undergarment down to the eigh- IP: 130.126.32.13 teenth century was smock. It was then replaced by the more delicate word shift. In the nineteenth century the same motive led to the substitution of the word chemise, and in the twentieth this has been replaced by combinations, step-ins, and other euphemisms (Baugh and Cable 1978: 310). Similarly, English has such euphemisms as privy, W. C., lavatory, toilet, bathroom, cloakroom, restroom, comfort room, comfort station, loo, etc. Meaning of regeneration involves the opposite process of degeneration. For instance, words like budge, coax, nonplus, shabby, squabble, stingy, tiff, touchy, wobbly were recorded with proper disparagement by Dr. Johnson, but they have since passed into standard speech. The word sturdy originally meant harsh, rough, or intractable. We now use it in a wholly complimentary sense. Even the word smock, which was just mentioned above as losing caste in the eighteenth century, has now been rehabilitated as applied to an outer garment. Now we use it for a certain type of woman’s dress and we speak of an artist’s or a laboratory technician’s smock. Baugh and Cable claim that the tendencies discussed above “are universal in their application and are not confined to the nineteenth century or