A Dictionary of Philosophy, Third Edition
. A priori knowledge is that which has its justification independently of experience, though it may presuppose experience from which we can get the concepts it involves; many philosophers (though not all: see Lehman) regard mathematical knowledge as a priori, though children can’t acquire it until they have experience of the world. Knowledge which can only be justified by at least some appeal to experience (basically the five senses, and perhaps introspection) is called a posteriori or empirical. A proposition, judgment, etc. is a priori or empirical according as knowledge of it is one or the other. Originally, in Aristotelian philosophy, a proposition was a priori if it was based on, or inferred from, something prior to it in the sense of being its cause or ground. A proposition was a posteriori if it was inferred from its effects. When it was later assumed that the main way of knowing a proposition through its effects was to know it through sense-experience, ‘empirical’ largely replaced ‘a posteriori’, and ‘a priori’ took on the meaning given above.
The philosophical and epistemological question of how our claims to knowledge can be justified is different from the psychological question how we in fact came to our knowledge. Some of it we may be born with, but this will only be innate knowledge if it could be justified, whether a priori, as above, or empirically, through experience, though innate empirical beliefs, whether true or not, are likely to be called instinctive, especially if they manifest themselves only in action rather than in conscious awareness. However, these philosophical and psychological questions have often been conflated, sometimes through confusion, but sometimes through the thought that the psychological question should properly replace the philosophical one, as in naturalized EPISTEMOLOGY (cf. NATURALISM). Also the way we acquire a belief, especially if we acquire it by reasoning or intuitive insight, may well coincide with the way we could justify it—but not always: we know that a belief we ‘acquire’ by its being innate may well be false (for more on this see INNATE). Kant, in particular, usually talks of our a priori, rather than innate knowledge, meaning knowledge which we cannot get by experience because only if we already have it can we make any sense of experience. Innate ideas or concepts are also often called a priori, and a proposition can be regarded as absolutely a priori if all the concepts in it are a priori e.g. ‘No proposition is both true and false’, and as relatively a priori if they are not, e.g. ‘Nothing can be simultaneously red and green all over’. ‘Relatively a priori’ could also apply to the everyday sense in which an empirical proposition is knowable independently of a given context, as when a detective says, ‘I haven’t yet found any clues, but I know a priori that money is a motive for murder’.
It has usually been assumed that for any given sense of ‘a priori’ and the corresponding sense of ‘empirical’ every proposition is either a priori or empirical. But sometimes a proposition is not justified by experience, nor known a priori, but simply postulated. Those postulated as regulative principles to guide scientific procedure can be called non-empirical, though they are often classed as a priori, or sometimes, ‘weak’ a priori. ‘Non-empirical can also cover the a priori in general. The term pragmatic a priori (C.I.
Lewis) has been applied to propositions we decide by fiat to make immune to falsification by experience, e.g. ‘Through a point not on a given straight line infinitely many straight lines parallel to the given one can be drawn’, as a postulate of a non-Euclidean geometry.
The epistemological a priori/empirical distinction has often been thought to coincide with the metaphysical necessary/contingent distinction (see MODALITIES) and the logical ANALYTIC/synthetic distinction (concerning the structure of propositions). Kant, however, split the third distinction from the other two, calling some a priori and necessary propositions synthetic, while more recently Kripke has split the first two distinctions, arguing that some propositions are both a priori and contingent (‘The standard meter rod is one metre long’), while others are both empirical and necessary (‘Water is H2O’). See also INNATE, RATIONALISM, INTUITION.
Aristotle, Posterior Analytics I. 1, 2.
D.Bostock, ‘Necessary truth and a priori truth’, Mind, 1988. (Complex but often illuminating defence of claim that a priori/ empirical and necessary/contingent distinctions come apart, though for different reasons than Kripke gives.)
D.W.Hamyln, Theory of Knowledge, Macmillan, 1970, chapter 9. (General discussion of a priori knowledge.)
I.Kant, Critique of Pure Reason, 2nd edn, 1787, B1–6.
J.J.Katz, ‘What mathematical knowledge could be’, Mind, 1995. (Defends a priori knowledge in maths etc. as not requiring causal or quasi-perceptual contact with abstract objects.)
P.Kitcher, ‘A priori knowledge’, Philosophical Review’, 1980. (Offers analysis of a psychologistic or materialistic kind. For criticism, and defence of a moderate version of a more traditional type, see D.M.Summerfield, ‘Modest a priori knowledge’, Philosophy and Phenomenological Research, 1991.)
S.Kripke, Naming and Necessity, Blackwell, 1980 (originally published 1972).
H.Lehman, Introduction to the Philosophy of Mathematics, Blackwell, 1979. (Part II defends an empiricist view of mathematics. Cf. also I.Lakatos, ‘A renaissance of empiricism in the recent philosophy of mathematics’, British Journal for the Philosophy of Science, 1976, tracing reactions to various unsettling developments like GÖDEL’S THEOREMS.
C.I.Lewis, ‘The pragmatic conception of the a priori’, Journal of Philosophy, 1923, reprinted in H.Feigl and W.Sellars (eds), Readings in Philosophical Analysis, Appleton-Century-Crofts, 1949.
*P.Moser (ed.), A Priori Knowledge, Oxford UP, 1987. (Reprinted selections.)
A.Quinton, The Nature of Things, RKP, 1973, pp. 132–4 (‘A priori’ and ‘instinctive’.)
M.Thompson, ‘On a priori truth’, Journal of Philosophy, 1981. (Argues that it concerns our thinking itself, not any subject-matter it has.)
This is the complete article, containing 948 words
(approx. 3 pages at 300 words per page).
View More Summaries on A priori and a posteriori (philosophy)
