Realism and antirealism

1. Facets of the debate

Realism became a prominent topic in medieval times, when it was opposed to nominalism in the debate concerning whether universals were independent properties of things or if classification was just a matter of how people spoke or thought (see Nominalism). The impetus for the debate in modern times comes from Kant’s doctrine that the familiar world is ‘empirically real’ but ‘transcendentally ideal’, that is to say a product of our ways of experiencing things, not a collection of things as they are ‘in themselves’ or independently of us. Kant’s ‘empirical realism’, confusingly, is thus a form of antirealism (see Kant, I. §5).

Closely related is ‘internal realism’, as represented by Hilary Putnam, according to which something may be real from the standpoint marked out by a particular theoretical framework, while the attempt to ask whether it is real tout court without reference to any such framework is dismissed as nonsensical (see Putnam, H. §§7–8). This re-affirms the thesis propounded earlier by Rudolf Carnap, that there are ‘internal’ and ‘external’ questions about existence or reality (see Carnap, R. §5). An internal question is asked by someone who has adopted a language of a certain structure and asks the question on that basis. Only philosophers attempt to ask external questions (are there really – independently of the way we speak – physical objects?). But this is either nonsense or a misleading way of asking whether our linguistic framework is well suited to our practical purposes. ‘Internal realism’, it should be noted, is certainly not a form of realism, since it admits only language- or theory-relative assertions of existence.

By the mid-1980s, largely as a result of the work of Putnam and Dummett, it had become common to formulate the distinction between realism and antirealism in a variety of what are prima facie quite different ways. A realist, it was said, thinks of truth in terms of correspondence with fact, whereas an antirealist defines truth ‘in epistemic terms’, for instance as ‘what a well-conducted investigation under ideal circumstances would lead us to believe’. A realist holds that there are, or could be, ‘recognition-transcendent facts’, whereas an antirealist denies this. Also present was the idea that an antirealist believes that there can be a ‘reductive analysis’ (see §2 below) of whatever subject matter their antirealism relates to, whereas a realist holds such analysis to be impossible. Seemingly still further from the origins of the distinction, it was said to be characteristic of realism to accept, and of antirealism to deny, the general validity of the law of excluded middle. Yet another version located the basic difference in the respective theories of meaning: a realist gave the meaning of a sentence by specifying its truth-conditions, an antirealist by specifying the conditions under which it could properly be asserted.

To come to terms with this debate, the reader therefore needs an awareness of the interrelations of the many definitions of the realism–antirealism distinction, and of the inexactness of fit between some of them and others.