ABSTRACT
The change of attitude from the last chapter on meaning is slight, but there nevertheless is a change. We are now looking at the problem of meaning away from the narrower context of the “classical philosophical” approach. The result may not be so dramatically different, but one result is undoubtedly to point us in the direction of pragmatics. This chapter and the next therefore should be read together, since the one shades over into the other.To initiate proceedings, though, we say a few words about syntax since the large-scale developments in syntax of the last few years cannot be ignored and they do represent a strong influence on linguistic analysis. SYNTAX The structure of languages - at least Indo-European languages - is essentially that of collections of sentences in subject-predicate
form, which in turn means that the sentence is broken down into the subject itself and its amplifications by adjectives and adjectival phrases, and the predicate contains the essential verb and enlargements of the verb in adverbs or adverbial phrases, with the possibility of an object - whether direct or indirect - which can, like the subject, be suitably amplified.The most important types of words used in the syntactical analysis of language are nouns, verbs, adjectives and adverbs, although connectives also play an important part in language.In a very interesting analysis of the development of language, Diamond (1959) showed that with the evolution of language the following state of affairs existed: We may, then, sum up the vocabulary of Modern English by saying, with growing emphasis, that nouns are generally formed from verbs, adjectives are overwhelmingly formed from nouns, and adverbs in practically every case from adjectives . . . Diamond also pointed out that there is, with the development of language, “a universal decrease in the proportion of the verb” (to nouns especially). We might expect this on the grounds that an increase in human knowledge is vastly more dependent upon an increase in “things” and “events” (or a knowledge of things) than it is on an increase in what we can perform by way of actions on such things. Thus modern science, by virtue of its experimental techniques, can bring more and more structural detail to light but can only slightly increase the operations we can perform.The questions raised by Diamond and others who have worked only on Indo-European languages do raise a lot of questions which are difficult to generalise upon. Many of the arguments that have led to the postulation of so-called “deep grammars” reflect the fact that some languages have very different “superficial structure” and perhaps none more so than those languages which are essentially pictorial.We should now try to distinguish between syntax and semantics prior to a more detailed discussion of syntax. The basic distinction is between rules that decide the structure of sentences (syntax) and those that we must be clear are by no
means complete and precise (i.e. English and other natural languages are by no means completely formalised) and semantics which is the science of meaning. We cannot say easily, or even attempt to say (apart from what we have already said in the last chapter) as yet, what we mean by ‘meaning’ since this is one of the central problems of semantics. We could think of semantics as being concerned with rules of designation which means that the words used in English sentences must have their precise references stated. In other words, we want to be able to say that ‘Berlin’ designates Berlin, ‘women’ designates the class of all women and so on.Where we have to be careful in drawing a firmer line between syntax and semantics is where we have to deal with statements such as:
“women are men” which immediately raises the point as to whether the exclusion of such sentences should be on the grounds of syntax or semantics. We could, for example, have rules that statement forms of the kind “A are i?” should entail (syntactical) rules as to what are acceptable as far as A and B are concerned. As the above case makes clear, it should be possible to have rules that “B must not be a class which is mutually exclusive with respect to A”. We can, of course, formulate many more rules along the same lines which would (or could) be capable of eliminating all “meaningless” statements.The above rules for the elimination of meaningless statements would achieve their object by constraints upon the acceptable statement forms, but we could equally well take a semantic view of the problem and demand that certain statements are unacceptable on grounds of offending rules of meaning. We can at this point see clearly that the rules of meaning that were offended would be stated in terms which were much the same as the syntactical rules, so that we might agree that the boundaries between what we call syntactic matters and what we call semantic matters are at the very least shadowy.G. A. Miller (1965) has made a very relevant point here:
The syntactic structure of a sentence imposes groupings that govern the interactions between the meanings of the words in that sentence. This clearly underlines the fact that syntactic structures help to determine meaning, and that, for example, the meaning of particular terms (if we can talk thus) is contextually determined, at least in part. Let us now look briefly at some forms of syntactic analysis.The simplest type of syntactical analysis, which we already mentioned, breaks sentences down into subject and predicate, where the subject is made up of the subject word or phrase (term) and its enlargement in the form of adjectives or adjectival phrases which refer to the subject (this all refers, of course, to Indo-European type languages). The predicate is more complicated having possibly a direct as well as an indirect object or no object at all but rather some predicative phrase as well as a verb. The objects (if any) and the verbs can be enlarged upon in precisely the same way as the subject. Sentences can, of course, have sub-sentences coming in the above categories and this makes for considerable complication. Furthermore, as we have already said, it is not always easy to separate syntactic from semantic issues.Postal (1964) has made the point that linguistic skills are of great complexity. He says: In short, I hope to have shown that the results of generative linguistics are not an obscure oddity, of interest only to the specialist in linguistics, but rather provide the kind of knowledge which is prerequisite to the understanding of the domains of the entire range of language studies.
