Meaning and explanation
We’ve now considered four proposals about the nature of meaning: meaning as reference/denotation, meaning as concepts, meaning as brain states and meaning as use. What conclusions can we draw? One particular conclusion concerns the status of the term ‘meaning’ itself. Even though the notion of a word’s meaning can be used to facilitate many tasks on the level of practical language use (explanation of new words, translation from one language to another, prescriptive regulation of disputes over usage, etc.), and seems indispensable on the intentional level of explanation discussed in 1.6.3, we should consider the possibility that ‘meaning’ is essentially a pretheoretical, informal notion which will not have any precise equivalent in a detailed account of linguistic behaviour on the other two levels.
QUESTION What are some other everyday, pretheoretical notions about language which have to be abandoned for the purposes of ‘scientific’ linguistics?
Perhaps, then, we do not need to choose between the different theories of meaning discussed in the previous section. As suggested in 1.2, ‘meaning’ can be seen as a shorthand way of talking about a whole variety of separate phenomena which are all individually important in our talk about language, especially on the intentional level of explanation, but which do not necessarily correspond to any single entity that will be revealed by careful empirical study. The English language category ‘meaning’, in other words, which in any case only has approximate equivalents in other languages, might have no precise role in a full understanding of language.
By contrast, the various aspects of meaning that we have distinguished in this chapter – reference, conceptual content, connotation, and so on – are all factors for which linguistic semantics does owe a principled explanation, and for which it should try to find theoretical analogues. There is no single way of breaking the definitional circle: ‘meaning’ is many different things, none of which should be ruled out as irrelevant to the eventual explanation of language.
In this context, we can specify an important condition that any principled theory of language must meet. This condition is linked to the idea that the ultimate goal of research into language must be to contribute to the causal explanation of people’s utterances. To achieve a thorough under standing of our linguistic ability, we will eventually need to be able to specify the detailed causal mechanisms which lead up to the production of utterances by speakers in real time. In order to achieve this, we will need a precise account of what the various phonological, semantic, morphosyntactic and semantic properties of different linguistic forms are, and of the ways in which these properties are combined in actual discourse sequences in different contexts of use. This goal is exactly the same as the one aimed at in other sciences: chemistry, for example, specifies the various proper ties of different molecules, in virtue of which they enter into sequences of causal interaction with each other, and embryology aims at understanding the properties of fertilized cells, in virtue of which a step-by-step understanding of their development into full organisms can be achieved. In the case of linguistics, the detailed nitty-gritty of a causal account is a long way off. What is more, the fi ne detail of an account of linguistic behaviour on the implementational level will have to be provided by neurolinguists and other brain scientists who will be able to isolate the physiological underpinnings of linguistic phenomena. The semanticist’s role is an earlier one, which consists in isolating the important properties of the linguistic system, on the intentional and perhaps algorithmic levels, for which these experimental scientists will need to find the physical mechanisms.
The distant goal of a causal account of language behaviour, however, suggests a possible role for the notion of meaning in semantics. From this perspective, we can suggest that to talk about a word’s meaning is a short hand way of talking about whatever property of a word could enter into causal explanations of its use. In our ordinary talk about language, one of the main functions of the category of meaning is to explain word use: we use the words we use because of the meanings they have. But in order to go beyond this pretheoretical level and explain a word’s use in a rigorous way, which might be ideally compatible with a causal account of language, a word’s meaning may include many different explanatory properties and necessitate consideration of referents and concepts and situations of use.
As a result, we do not need any single, categorical answer to the question of whether meaning is denotation or concepts or uses. To phrase the question as a set of exclusive choices like this is counterproductive, since it may well turn out that all of these categories will need to be invoked in order to explain the use of different words. Thus, as we noted at (44), there is a sub class of words and phrases in any language whose use seems particularly closely linked to certain recurrent and specifiable situations; the most obvious way of explaining the use of these words is to associate them with the particular contexts and situations in which they occur, and the use theory of meaning will be the most relevant. Other words, however, seem best explained by the particular conceptual associations they call up; for these, attention to the link between words and concepts will be the most relevant. If I say, for example, The holidays were a nightmare, then the words holidays and nightmare call up a whole variety of specific connotations and associations (see the question below) for which the conceptual theory of meaning will be most appropriate. In still other cases, such as proper names and ‘deictics’ like here, it seems to be a word’s referent which is the most important factor in accounting for the word’s use on a given occasion: if I say that man just fell over, the ‘meaning’ of that man is best described as the actual person to whom I am referring. This is not to say that concepts are irrelevant for expressions like that man or for words like those in (44), or that referents and denotations are irrelevant for words like holiday or nightmare. In most cases, indeed, we will need to attend to all three aspects of a word’s ‘meaning’, in considering how its relations with referents/denotations, associated concepts and uses mutually combine to account for its presence in a particular linguistic context. It is just to say that in all these cases attention to the explanatory purpose of talk about meaning will direct us towards which ever conception of meaning seems to provide the best explanation of the particular semantic phenomenon at hand.
QUESTION Describe the concepts HOLIDAY and NIGHTMARE in as much detail as possible. How much of this detail is relevant to explaining linguistic behaviour?
