This book expounds a range of familiar wittgensteinian themes with particular focus on Wittgenstein’s account of ‘grammar’ and the grammatical status of religious beliefs. However, McCutcheon interestingly diverges from most other accounts of this material by virtue of the special attention she gives to the question of realism, that is, whether the sentences of a field of discourse (specifically religious discourse) address a real, mind-independent subject matter. Many writers on Wittgenstein’s remarks on religious language tend towards an antirealist reading of his work, though rarely state in any detail the form that this religious antirealism takes. This is surprisingly remiss given that Wittgenstein’s supposed failure to give a plausible account of the apparent ontological commitments of religious believers is widely taken as a decisive objection to his proposals. McCutcheon brings this issue to the fore and takes an unusual realist reading of Wittgenstein’s remarks.

Chapter one traces Wittgenstein’s philosophical development from the Tractatus to Philosophical Investigations. In at least one respect McCutcheon is sympathetic to the ‘one Wittgenstein’ school, the view that there is a relatively smooth transition from Wittgenstein’s early to late (post 1929) writings: Wittgenstein is held to address the same fundamental problems throughout his career. These problems concern the limits of thought and language. Wittgenstein’s point is expressed in a variety of ways:

Such contentions, McCutcheon believes, put Wittgenstein in prima facie conflict with realists who maintain that ‘breaking free of’ or ‘getting outside’ language is precisely what is required if our notions of truth and objectivity have any foundation. Indeed, McCutcheon argues, there is a sense in which Wittgenstein shows this realist view to be mistaken: any appeal to reality or reason must be done ‘in and through’ language. However, McCutcheon does not believe that this commits Wittgenstein to an antirealist position, for even if we cannot ‘get outside’ language we can nevertheless make statements about the world ‘that turn out to be true because they match reality’ (27). Thus recognition that we cannot break free of language in talking about the world does not undermine the possibility of referring to what is real, or of making true claims about the world.

McCutcheon follows this up with a discussion of Wittgenstein on grammar and attempts to elucidate his difficult remarks on the arbitrariness of grammar. On the wittgensteinian model, grammatical rules are rules of language that determine the boundary between sense and nonsense, and are grounded in agreed human practice. McCutcheon proposes that we can identify the grammatical rules of a language game as those principles which cannot be questioned without undermining the game itself. grammatical rules. An analogy here is with a game of chess: to question a rule of chess indicates that one does not understand how to play the game, or else that one wants to play a different game. Moreover, grammatical rules, as with rules of chess, are said to be arbitrary in the sense that they do not depict reality but are principally underpinned by human agreement. This model seems particularly difficult to believe when we come to our notions of what is necessarily true or necessarily false. Since the necessary statements in a language game are unalterable and essential components of that game, they presumably count as grammatical rules. The implication of this wittgensteinian position seems to be that principles such as the laws of non-contradiction or self-identity, or the results of any mathematical proof, are ‘arbitrary’, in the sense that they somehow rely on human agreement for their truth. Indeed, since the main function of grammatical rules is to stipulate good practice within a language game, they are not even descriptive or referential.

Once applied to religious language, this model provides a familiar wittgensteinian picture of religious beliefs and claims. If a religious statement is construed as a rule of grammar rather than a descriptive (perhaps even empirical) claim, then to sincerely assert that God exists is not to state a fact but rather to commit oneself to a religious system of reference. To talk of God is not to posit the existence of a supernatural entity with such and such properties, but rather to adopt a religious form of life which consists in thinking and directing one’s behaviour with reference to one’s concept of God. Even petitionary prayer is construed in a way that strips it of ontological commitments. In general, religious beliefs are understood in terms of the thoughts, behaviour and attitudes of the believer and the community, rather than having any descriptive content. McCutcheon, however, wants to resist this line of interpretation: we can accept the grammatical status of religious claims while also allowing that they entail ontological commitments. McCutcheon’s argument depends on a supposed ‘instability’ in the distinction between grammar and distinction. Unfortunately, the argument is difficult to evaluate because it is not stated in detail, there are few examples and no supportive evidence from Wittgenstein’s writings. If the idea is that the grammar/description distinction is ill formed, it is difficult to see why Wittgenstein placed so much weight on it. If the point is that the statement of a rule of grammar may, in another context, function as a descriptive claim, why should that show that the statement has descriptive content when it is being used as a grammatical rule?

McCutcheon’s concluding chapter discusses expressivism, the theory that the sentences of a field of discourse are not truth-apt, i.e. do not describe or refer to the world. Rather, they express attitudes, emotions or stances of those who assert them. McCutcheon rejects expressivism not only for religion but also for other discourses, such as ethics. Her primary reason is that the distinction between cognitive and non-cognitive discourse is confused; though at one point she suggests that it might be defended from naturalist assumptions. I wonder, however, whether a much better case might be made for the distinction than McCutcheon’s discussion of the issue suggests. To take an example from Simon Blackburn, if I taste something I like and exclaim ‘Yummy!’ then what I say is plausibly taken as a non-descriptive expression of gustatory delight. It remains expressive if I say ‘It’s yummy!’ instead, even though this utterance is a sentence that can be disagreed with or negated. The fact that the utterance has the form of a sentence does not establish that it has a descriptive content. Similarly, the expressivist may argue, all judgements of taste emerge from expressions of pleasure or displeasure. We may regard this position as less plausible when applied to ethical or religious discourse, but it is not clear that the grounds for it are confused.

I had a number of reservations about McCutcheon’s account of Wittgenstein and her approach to the realism problem. For example, if one looks at the various ways in which Wittgenstein’s position on ‘limits’ is expressed (five are given above), it is not clear whether the subject of his remarks is the limits of language or the limits of thought or, if both, how they are related. If the limits of thought and of language are supposed to be the same, then it is not clear what ‘the limits of thought in language’ could mean. Moreover, the idea that language has ‘limits’, and the realist’s claim that we can ‘get outside’ language are metaphors – but of what? We are not given a direct statement of what is at stake in the debate about realism. Nor is much evidence given to show that Wittgenstein was himself much motivated by the question of limits; there are few references to it in his later work. For a book so immersed in wittgensteinian thinking, the arguments are rather distant from Wittgenstein’s own writings; we are given a way of thinking about Wittgenstein’s philosophy without any telling reasons to show that he thought that way too. This would be of only minor concern if questions about limits were shown to be a central concern in the contemporary debate about realism, or a substantive issue in their own right, but they are not. Realism has been a major topic in analytic philosophy of the last forty years and has been extensively written about by first rank philosophers (Michael Dummett, John McDowell and Hilary Putnam come to mind); moreover, a number of these writers are fully informed of Wittgenstein’s remarks on the issue. However, only the views of Roger Trigg are considered in any detail.

The use of some more sophisticated material on realism might have assisted with two other difficulties with McCutcheon’s argument. Although McCutcheon offers various characterisations of Wittgenstein’s position, and contrasts it with Trigg’s, it is difficult to pin down exactly what Wittgenstein is saying. On one reading, Wittgenstein’s supposed resistance to our ‘getting outside’ language on the grounds that we have no access to reality other than through concepts, is simply idealism. Wittgenstein is offering a version of the Berkeleyian argument that we can have no idea of something that is not being conceived of by somebody. On another reading, Wittgenstein is making just the trivial point that one cannot think something about the world other than by having a thought, or say anything about the world other than by using language – vacuously true claims that even the most incautious realist should not resist. The idealist position is, of course, highly contentious (and dubious), while the latter reading seems to make no progress on the realism problem.

A similar difficulty extends into the account of the realist stance. McCutcheon argues that while Wittgenstein disputes the realist idea that we can ‘break free’ of language, he certainly does not deny ‘the ordinary, everyday sense in which we make statements about reality that turn out to be true because they match reality.’ (27) But it is exactly on the notion of truth that the disagreement between realist and antirealist comes into sharper focus. The realist may argue that the ordinary, everyday notion of truth is one that is given by the correspondence theory of truth, so that a true statement is one that corresponds to facts or state of affairs that it describes. For example, religious statements will be true or false even if we are unable in principle to determine their truth or falsity. The antirealist may argue that the truth of a statement is determined by the satisfaction of the standards of discourse in which the statement is made (i.e. that truth is warranted assertibility, or some similarly epistemically constrained notion). If truth is tied to our standards of warrant and verification, then there cannot be wholly unknowable religious truths. Wittgenstein himself may be sympathetic to the latter view given that he adopts what appears to be a redundancy theory of truth in Philosophical Investigation:

‘p’ is true = p

‘p’ is false = not-p (PI 136)

This can be seen as equating the claim that p is true with the assertion of p, and the claim that p is false with the denial of p.

Despite these reservations, both the focus and thesis of McCutcheon’s book are interesting and deserve consideration.