Emotive meanings are essentially different from referential ones, for they are not structured according to series of dichotomies or contrasts. Rather, emotive meanings consist of polar contrasts separated by a graded series with a high percentage of usages for most words clustering around the neutral position.

To measure the emotive meanings of a word we need a complex matrix for each word. The dimensions of such a matrix could include, for example, a ten-point graded series with such dimensions as good-to-bad, pleasant-to-unpleasant, favorable-to-unfavorable, happy-to-sad, lovable-to-hateful, beautiful-to-ugly, and acceptance-to-rejection. Such dimensions would change from word to word, depending upon patterns of applicability, but if we had the judgments of an adequate sampling of people's reactions to verbal symbols plotted on such a matrix we would have at least a profile of the major emotive feature of such words. Obviously words such as delicious, mother, honeymoon, cat, dog, whore, nasty, and bastard would have quite different profiles.

To test the appropriateness of certain stylistic usages of these and similar words we could measure the match between the emotive values of the words and the types of discourse. For example, in so-called scientific writing we would expect a choice of words which would have a predominantly neutral or central profile. For expressive writing we would expect words whịch would fit the mood, c.g. joyous or depressed. For "slanted" writing or speech we could judge the extent to which the profile of the words was in keeping with the intent of the message.

Some communications are, of course, purposely mixed in emotive meanings, as in the expression of irony and sarcasm, e.g. charming rascal, damnably sweet, and deadly attractive. Some communicators simply want to shock гeceptors, while others, including a number of existentialist novelists, communicate some of their theme of "absurdity" by the lack of match in the emotive meanings of words.

Within a particular language there are often quite radically different profiles for words, depending upon local usage and individual associations, e.g. revolution, communism, Republican, Democrat, cotton-pickin', and beatnik. Such differences must simply be treated as dialectal variants.

It is important to note that not all languages have the same dimensions for classification of emotive meanings. For example, in a number of languages in Latin America the distinction of hot vs. cold is a highly important dichotomy with important emotive overtones.

It must be recognized, of course, that languages tend to differ more radically in emotive meanings than in referential significations. In a number of areas in Africa, for example, there is very little taboo connected with the names of body parts ånd organic functions, while the names of certain animals, ancestors, and kin involve very heavy emotive significance. As noted in the study of hierarchical structures, the more languages depend upon perceptual models the greater is the

agreement between them, and conversely the more they depend on conceptualization the greater are their differences. Patterns of emotive meaning, however, show even wider discrepancies than distinctions based on conceptualization.

责任编辑：admin