Word meaning

Word meanings are what are sought and what should be provided in comprehensive dictionaries of a language. For much of the history of semantic studies, and still to a considerable extent today, the investigation of word meaning has been based on the relationships of reference and denotation. Certainly meaning includes the relations between utterances and parts of utterances (e. g. words) and the world outside; and reference and denotation are among such relations. But for the purposes of linguistics it is desirable to deal with meaning by a more comprehensive treatment.

Sentences have meaning, are meaningful; and a child learns the meaning of many words by hearing them in other people’s uttered sentences and practising such utterances himself subject to the correction of others and the test of being understood by those to whom he is talking. The process goes on all our lives, and we learn new words and extend and increase our knowledge of the words we already know, as we hear and see them in fresh utterances and used slightly differently from the ways which we are accustomed to. The meaning of a word, therefore, may be considered as the way it is used and understood as a part of different sentences; what the dictionary does is to try and summarize for each word the way or ways it is used in the sort of sentences in which it is found in the language.

The grammatical structure and certain phonological features such as intonation may themselves give an indication of part of its meaning, as we can easily see when we consider the part played in English and in many other languages by word form, word order, and intonation in the indication of questioning, commanding, and making statements. Though familiar in literate languages and apparently universal in all languages, word divisions are not immediately audible in connected speech, and the formal features that determine words as separable units, and the recognition of such features, intuitively by the speaker and objectively by the linguist, must be examined further within grammatical analysis.

The potential sentences of any language that may be uttered and understood by a speaker of it are infinite in number, but they are formed from the total stock of words known to the speaker at any time. A speaker’s word stock is always variable, but it may be regarded as fixed at any given point in time. Words, therefore, are, in general, convenient units about which to state meanings, and no harm is done provided it is borne in mind that words have meanings by virtue of their employment in sentences, most of which contain more than one word, and that the meaning of a sentence is not to be thought of as a sort of summation of the meanings of its component words taken individually. With many words particular meanings or uses are only found when they are used in conjunction with other words, and these are often scarcely deducible from their other uses apart from such combinations (one need only think of such phrases as cold war, black market, wildcat strike (unofficial strike, particularly in American English), white noise (in acoustic engineering). […]

Reference and denotation are clearly a part of the meaning of many words in all languages. The many problems arising about the nature of these relations have been the subject of much philosophical discussion. Here it suffices to point out that by the use in sentences of certain words one is able to pick out from the environment and from the general knowledge of speaker and hearer particular items, features, processes, and qualities, draw attention to them, give or elicit further information about them, make them the objects of action or speculation, and, most importantly, recall them from past experience and anticipate them in the future provided only that the words used have had such associations in the previous experience of speaker and hearer. These are the words whose meanings may, in part, be learned and taught by pointing. But the relationship between the word and that to which it may be said to refer is not a simple one. Proper names (John, Mary, etc.) refer to individuals as single individuals, however many there may be so referred to; boy, girl, etc., refer to an indefinitely large class of individuals by virtue of being grouped together in some respect; in the same way, climb, fly, swim, and walk refer to four different types of bodily movement in space. In the strict terminology of logic denotation is sometimes used in a specific and technical sense, but in general usage the term is more loosely made equivalent to reference.

It is often said that the meaning of a word is the idea it conveys or arouses in the mind of speaker or hearer. This is associated with a general definition of language as ‘the communication of ideas by speech’ or the like. Such accounts of the meaning of words and the working of languages are objectionable. Idea as a technical term is notoriously hard to pin down with anything like precision. It is often taken as equivalent to mental picture or image, for which drawings are sometimes made in books dealing with linguistic meaning. Mental pictures are no doubt perfectly genuine components of our private experience, but as such they seem of little relevance to linguistics. Firstly, it would appear that they are not aroused by anything like all the words in a language, even of those for which a referential meaning is fairly easily statable in isolation; secondly, even in the most favourable cases, the idea as a mental picture does not help explain one’s ability to use a word correctly and understand it. Any picture is necessarily particular; as Berkeley pointed out, triangle refers to all of the mutually exclusive sorts of triangles (isosceles, scalene, right-angled, etc.), but any picture, mental or otherwise, of a triangle must be of one triangle only. Even if we did recognize what a word referred to by having a mental picture in our mind, we should have to be able to justify the classing together of what is actually observed and the mental picture by some further piece of knowledge. It is best to regard knowledge of the meaning or meanings of a word as part of a speaker’s competence, an ability to use the word in ways other people will understand and to understand it when uttered by other people; this knowledge includes knowing the range of items, processes, and the like to which words that do have referents of one sort of another may be said, often indeterminately, to refer.

The use of words in utterances to focus attention on particular bits and features of the world involves a segmentation and an organization of our experienced environment. Verbalization is not a mere passive labelling of discrete items and objects; the process of classification implicit in the use of what are often called common nouns (boy, girl, tree, house, etc.) has already been noticed. Moreover, the very permanence of names and designations presupposes that we recognize continuing identities in the stream of successively observed phenomena. Recognizing John today as the John of yesterday and this table today as the same as this table yesterday is more than just perceiving what is before one’s eyes; it is imposing some order on such perceptions. Other words with more abstract meanings involve a much more far-reaching organization of the world of immediate experience; words like motion, gravity, inertia, energy, and equilibrium do not refer to things in the way words like table and chair do, nor do even more abstract words like cause and effect (still less do they call up definite pictures in the mind), but they have distinc­tive and important meanings, and their use is a mark of the high degree of order and systematization imposed by us on the world we live in. In the same way the use of words like right and wrong, duty, crime (and many others subsumed under them: property, theft, punishment, reform, etc.), and of comparable words in other types of society, presupposes a social nexus of expected ways of behaviour enforced by precedent and the sanc­tions of disapproval and legal penalties.

Quite apart from the examples just above, many words used quite ordinarily in everyday life, whose meanings are in no sense part of a specifically scientific or philosophical vocabulary, bear much more abstract and complex relations to our world of things, actions, and processes than words like chair, stone, sun, kick, and run. The difficulties, intricacies, and delicacies of much semantic analysis are concealed from the beginner when, as is too often the case, such words are exclusively chosen as examples in elementary semantics, just because of their rather obvious direct­ness of reference, at least in many of their uses. One may, in this context, reflect on what is involved in the semantic analysis of such ordinary words as succeed (success), prepare (preparation), loyal(ty), and persuade (persuasion).

Some of these ways in which human life are experiences are ordered in our languages appear to be universally recognized in the use of words in all languages, and must therefore be regarded as the general property of mankind (for example, the recognition of objects occupying space and persisting through time). In other matters languages differ in the way they most readily tend to organize parts of the speakers’ experience. Relatively trivial instances of such differences are the obvious non-correspond­ences of the colour words in different languages; more significant are the difficulties involved in trying to translate words relating to moral, religious, legal, and political matters between the languages of communities having different social systems in these respects. Such difficulties arise from the differences in peoples’ ways of life, and are made prominent by the work of anthropologists in examining societies far removed historically from the Greco-Roman and Hebraic inheritance characterizing Western Europe and those countries most influenced by it. We do not all inhabit exactly the same world, and differences in the significant items of vocabulary in languages bring this out clearly.

Not only does reference cover a very wide divergence of relationship between words and the bits and pieces of the world, but many words in all languages can scarcely be said to refer to anything by themselves, for which, consequently, pointing is useless as a means of explaining their use. This does not mean that such words are meaningless, which is nonsensical, as they have quite definite uses in languages; words like English if, when, of, all, none, the are frequent and essential components of sentences. But as it has been seen that it is the utterance and the sentences in it that are the primary meaningful stretches, the meanings of the component words must be taken as the contri­bution they make to the meaning of the sentences in which they appear. The fact that the contribution of some words is partly that of reference does not make reference the same as the whole of meaning; and it is not to be assumed that the meaning of a word when it constitutes a one-word sentence is the same as its meaning when it forms part of a larger sentence. There are also quite complex systems of reference involved in the uses of pro­nouns (I, you, he, himself, etc.).

The ease with which a statable meaning can be assigned to a word in isolation varies very considerably, and in part it depends on the degree to which the word is likely to occur in normal discourse as a single (one-word) sentence; and even in the case of such a favoured word one has no right in advance of the analysis to assume that there will always be found a common ‘core’ of meaning underlying all the various uses the word has in the sentences in which it may occur. With words scarcely ever occurring in isolation, like those cited in the preceding para­graph, it is almost impossible to describe their meaning adequately in any other way than by saying how they are typically used as part of longer sentences and how those sentences are used. The question whether a word may be semantically described in isolation is more a matter of degree than of a simple answer yes or no.

Preoccupation with reference and denotation has troubled semantic theory, by putting an excessive importance on that part of word meaning which can be stated easily in isolation and treated either as a two-term relation between the word and the referent or thing meant (or between word-image and concept, significant and signifie), or as a three-term relation between word, speaker or hearer, and referent. The meanings of sentences and their parts are better treated in linguistics in terms of how they function than exclusively in terms of what they refer to. The different types of reference indicated above are then included as part of the function performed, the job done, by certain words in the sentences in which they are used, and the dictionary entry of a word simply summarizes the function or functions, referential or other, of the word in the sort of sentences in which it typically occurs.

In addition to reference, excessive emphasis on historical considerations colours popular discussion on language, especially on word meanings, as when it is urged that the ‘real’ meaning of a word is to be found in its etymology or earlier form and use in the language. Thus it is claimed that the ‘true’ meaning of holiday is ‘holy-day’ or day set apart for religious reasons, and that awful and awfully are wrongly used like considerable, very (there was an awful crowd there, awfully nice of you to come), since its ‘real’ meaning is ‘awe-inspiring’.

If it is accepted that statements of word meanings in descriptive linguistics are simply summaries of the ways words are used in sentences by speakers at a particular time, it is clear that histori­cally antecedent meanings are outside the scope of such state­ments. Without specialized study speakers are ignorant of the history of their language; yet they use it to communicate with each other and they understand each other. Certainly the meaning of any word is causally the product of continuous changes in its antecedent meanings or uses, and in many cases it is the collective product of generations of cultural history. Dictionaries often deal with this sort of information if it is avail­able, but in so doing they are passing beyond the bounds of synchronic statement to the separate linguistic realm of historical explanation.