The analysis of meaning is an investigation of the inferential processes speakers engage in when communicating verbally. It is useful to distinguish between three types of such inferences, which I shall call linguistic, implicit, and implicative. Linguistic inferences pertain to the elaboration of those relationships expressed by syntactic manipulations such as subject-verb, verb-object, complement, relative clause, question, passive transformation, nominalization, tense, number, transitivity, and so on. Studies on the acquisition of language (see Smith and Miller, 1966) reveal that children very early master this kind of competence, perhaps as early as age three-and-a-half or shortly thereafter. Implicit inferences pertain to the semantic or conceptual implications of morphemes, words, and larger linguistic constructions. They presuppose knowledge of criterial and nonessential attributes of words (denotation and connotation) and their relation to the world. For instance, the verb "to eat" preserves its criterial attributes in "the man eats bread" and "the man eats soup," but one must also be aware of the facts that in the latter case, but not in the former, a "spoon" is implied, otherwise one would not realize the unusual implications of "the man eats bread with a spoon," or the difference between "the man eats soup" and "the dog eats soup."
Contrastive differences of a particular sort, which I have called elsewhere (James, 1968) "aspectual qualities" of words, are implicit in sets such as "to stare" vs. "to notice" (extended vs. short duration), "obsolescent" vs. "obsolete" (inception), "to nag" vs. "to criticize" (iteration), "to hire" vs. "to rent" (directionality), "filthy" vs. "dirty" (intensity), "to toss" vs. "to throw" (nonchalance), and so on. Another instance of implicit inferences pertains to the "affectivity" of words: the fact that "mother" is good, "elephant" is strong, "old man" is passive, "airplane" is fast, "rabbit" is weak but fast, "pyramid" is strong but passive, and so on (see James, 1969a). At the sentential level, implicit inferences recover such facts that in "She'll make someone a nice husband" (C. E. Osgood) the attributes of masculinity and dominance are implied and not sex.
Implicative inferences pertain to the function of utterances and their psycho-logical (as well as psychological) implications. To communicate effectively, an individual must know the difference between, say a promise and a request, and must be able to recognize which of these a speaker intends by a particular utterance. He must also recognize the logical implications of discourse "hinges" whether overt such as "therefore," "thus," "it follows that," etc., or covert and given by juxtaposition or sequencing (e.g., "Oh! We can't go out. It's raining" where causation is implied; cf. also the psychologic involved in the construction of paragraphs, Koen, 1968). Utterances also provide clues to the momentary psychological state of the speaker (e.g., emotional involvement, manipulative intentions) as well as more permanent characteristics (such as geographic origin, education, etc.). The quality of communication (its breadth, scope, effectiveness) will be a function of the quality of the inferences produced by the speaker and hearer on the three levels just outlined.
Communicative competence is a reflection of the quality of the linguistic, implicit, and implicative inferences of which an individual is capable. A theory of communicative competence, which is an attempt at specifying the nature of these inferences, will draw upon the knowledge developed in a wide body of interdisciplinary fields: linguistics (for clarifying syntactic relations), philosophy and psychology (for clarifying logical implications of statements and the acquisition and use of conceptual cognitive structures), sociology (for revealing characteristics of speakers that accrue to them by virtue of the social organization they belong to), and so on. The techniques which these and other disciplines have developed provide useful tools for the investigation of communicative inferences: on criteriality of semantic features (cf. componential analysis in cultural anthropology, Tyler, 1969), on commonplace background expectations in ordinary conversation (cf. the ethnomethodological procedures discussed by Garfinkel, 1968), on the management of impressions in interpersonal interaction (cf. Goffman, 1967, 1959), on the function of utterances (Searle, 1969; Skinner, 1957), on affectivity (cf. Osgood, James, May, and Miron, in press), on inferential reasoning (cf. Bruner, Austin, and Goodwin, 1957), and many others.
I purposefully omit mentioning here the traditional works on rhetoric, stylistics, and literary criticism, not because they are unhelpful, but because they represent where we want to go, but by a different road than has been used so far in those areas. These previous attempts, important as they are in revealing the richness and complexity of verbal communication, have failed to develop a systematicity that would insure the development of a cumulative body of knowledge upon which new insights can build and gradually chip away at the iceberg of ignorance. There is a critical need for accumulating facts and phenomena of verbal communication which can be recognized and accepted by scholars independently of their particular traditional orientation. For this we need to develop systematic tools of observations and classification by borrowing from the work carried out in the various disciplines mentioned.
Back to The Titlepage