Page 45 - Encyclopedia of Philosophy of Language

mitments. These include recognizing a clear theory- observation distinction, maintaining a realistic con- ception of the objects of scientific study, and also holding that the sciences have a hypothetico-deductive structure in terms of which notions like confirmation and testability can be explicated. Under the impact of Thomas Kuhn's sociological approach to the sciences, and of serious qualms about the viability of realism in metaphysics, theory of meaning, and the sciences, many contemporary philosophers of science would be reluctant to endorse Popper's image of an objective science being driven forward by the method of con- jecture and refutation. This is not to deny that scien- tists make conjectures and test them: but this is only
part of a range of activities in which they engage and may play a relatively minor role in determining which portions of current theories are rejected and which retained.
See also: Verificationism.
Bibliography
Hacking I 1983 Representing and Intervening. Cambridge University Press, Cambridge
Kuhn T S 1970 The Structure of Scientific Revolutions, 2nd edn. University of Chicago Press, Chicago, IL
Popper K R 1968 The Logic of Scientific Discovery, rev. edn. Hutchinson, London
Questions about foundations of (or for) linguistics might be either 'ontological,' 'epistemological,' or 'theoretical.'
1. Ontological Foundations
Foundations for linguistics in the ontological sense are established through identifying and describing the ultimate constituents or aspects of reality which linguistic theories seek to refer to and to characterize. (Similar questions arise in relation to other disciplines. Are social facts really, ultimately, facts about indi- vidual human beings? This is a question about onto- logical foundations for the social sciences.) Theorists concerned with this issue have formulated distinct accounts of the ontological foundations of linguistics.
According to 'psychologism,' linguistic theories are meant to characterize the psychological states of lan- guage users and, in particular, their competence to employ their language (see Chomsky 1986: ch. 1).
According to Platonism, the objects which linguistic theories are meant to characterize (namely, sentences) are purely abstract in the same way as are the objects of mathematical theorizing—they have, in other words, no material existence or embodiment per se (though they might be 'represented' by given material objects or events) (see Katz 1981).
According to 'behaviorism,' linguistic theories seek to characterize the actually occurring speech behavior of individual language users and to identify the stimu- lus circumstances and patterns of conditioning which give rise to it (see Skinner 1957).
According to 'conventionalism,' linguistic theories aim at characterizing the socially constituted con-
ventions which regulate individuals' speech behavior (see, for instance, Bennett 1976).
'Instrumentalism' is a radical alternative to these more committed positions on ontological foun- dations, according to which it is unnecessary, when theorizing about linguistic phenomena, to provide any account of deeper realities which might be manifested in these phenomena; linguistic theories, on this account, need only provide a basis for the prediction of phenomena. (Instrumentalism, less fashionable in the early 1990s than previously, was perhaps most appealing in relation to quantum mechanics, where it is notoriously difficult to give, within the framework of classical physical theories and of commonsense concepts, a coherent interpretation of underlying realities.)
2. Epistemological Foundations
Foundations for linguistics in the epistemological sense are established when a category of claims is identified with respect to which all other claims are to be justified. (Foundationalism in this sense is no longer as reputable, in the general philosophical context, as it certainly once was. According to con- temporary thinking, epistemic justification is provided not by establishing links between foundational claims and those which are to be justified, but, instead, by exhibiting the 'coherence' of the claims which are to be justified with other claims already, though only defeasibly, assumed to be justified.)
According to some, it is native speakers' 'intuitions' about grammaticality, synonymity, etc., that are to be used to test linguistic hypotheses. (This position is
Foundations of Linguistics F. D'Agostino
Foundations of Linguistics
23