Discussion of the religious realism problem – put briefly, whether religious sentences address a real subject matter – has an odd status in philosophy of religion. Whereas work on issues such as religious knowledge or the existence and nature of God commonly draws on cognate matters in contemporary mainstream epistemology and metaphysics, research on religious realism is more detached from other philosophical work on the topic.
Notwithstanding William Alston's papers on religious reference in the 1990s, the religious realism debate has tended to overlook or make only uncertain reference to the (fairly huge) literature in related fields like philosophy of language or ethics. Seminal papers and books – Michael Dummett and Crispin Wright on truth and Simon Blackburn on expressivism, to name just a few – remain largely unexplored, despite their quite apparent relevance to formulations and arguments about religious realism and anti-realism. Moreover, philosophers working on the realism question have for the most part reciprocated, tending to disregard potentially interesting applications of arguments in the realism/anti-realism discussion to religious cases. So it is a notable merit of Insole's book that it shows an awareness of a realism debate outside philosophy of religion that is also highly relevant to the ongoing discussion about religious realism, and aims to address it. The results, however, are a bit patchy.
Insole's discussion of expressivism in the first chapter – the view that religious commitments are primarily expressive rather than descriptive – recognizes that there are much more sophisticated versions of expressivism than the position advanced by R. B. Braithwaite in the 1950s. In particular, Insole draws on Blackburn's 1984 classic Spreading the Word (Oxford: Clarendon Press) which kick-started contemporary discussion of expressivism, though disappointingly he does not refer to the considerable literature that has followed it in the last ten years and has made expressivism one of the most discussed anti-realist positions, particularly in ethics. Also, the implementation of Blackburn's theory to religious discourse is imperfect.
According to Insole, expressivists claim that religious and ethical claims are ‘exclusively expressive of desires’ (his italics) and are ‘not at all descriptive’. But no (careful) expressivist, and certainly not Blackburn, has insisted that utterances that express attitudes cannot also have a descriptive element. Even A. J. Ayer recognized that an ethical claim like ‘It was wrong of James to lie to his parents’ conveys the belief that James lied to his parents. And one interesting contemporary variety of expressivism, developed by Mike Ridge, allows that moral utterances may express beliefs provided that the truth of a moral utterance is not guaranteed by the truth of the beliefs that it expresses.
Religious expressivism trades in part on the tidy explanation it can offer for the apparent motivational character of religious commitments: the attitudes that are voiced in religious utterances are those that change the lives, behaviour, and dispositions of religious believers. Insole believes that religious expressivism is untenable, but does not locate a compelling argument against the theory. For example, he argues that it is problematic for expressivism that the relationship between a religious commitment and action is ‘subtle’, ‘complex’, and may be ‘diffuse’. But why should the expressivist deny that the relationship is complex, subtle, etc.? Most contemporary expressivists are functionalists about mental states, so some complexity in specifying the relationship between desires and actions should not be surprising.
Insole also argues that expressivism is ‘unable to account for instances where someone expresses a belief but also fails to act on it’, but in fact this is a familiar point to which the expressivist has a ready response. An utterance that typically gives voice to a certain attitude will not change the behaviour of the speaker either if the utterance was insincere, or if the speaker was sincere but has other attitudes that counteract any behavioural effects that the expressed attitude might have. Surprisingly, Insole does not make use of the Frege-Geach argument, widely taken to be the most compelling objection to expressivism (and well advertised by Blackburn), which aims to show that attitudes cannot exhibit the logical features of validity and contradiction commonly found in ethical reasoning. It seems that this problem could straightforwardly be adapted from the ethical to the religious case. Insole may be right that religious expressivism may be unsuccessful, but does not really give it a fair shake or show where it goes wrong.
Insole devotes two chapters to questions about religious truth. He begins by looking at the much-commented-on (but not very clear) idea, pressed by D. Z. Phillips and possibly suggested by Wittgenstein, that religious truth should be understood as internal to a language-game. Insole helpfully links this up with a broader, and much clearer, philosophical debate as to whether truth is epistemically or evidentially constrained, that is, whether truth can be analysed in terms of what it is possible for us to know or to discover. One upshot of such an analysis is that there cannot be unknowable truths or evidence transcendent truths, i.e. truths which outstrip our ability to establish them, however much our information or cognitive abilities are improved. For at least some areas of discourse this seems to capture part of what is at stake in realist/anti-realist debate: the anti-realist sees the truth in the discourse as determined by the practice of discovering truths – truth is a status conveyed on a sentence by virtue of its satisfying relevant standards of warrant. The realist, in contrast, sees the truth of a sentence to be determined by its accurately representing the world, and that this relation may hold independently of the practice of discovering it. Our investigative methods aim to track the truth rather than in any way constitute it.
The interesting prospect of a treatment of how these ideas might play out in philosophy of religion, however, never materializes and Insole instead gets rather bogged down in analysis of the work of D. Z. Phillips. This causes a couple of problems. One is that Phillips takes the argument in some very puzzling directions. For example, Insole spends several pages looking at the (widely discredited) redundancy theory of truth, which he believes Phillips uses in defence of an epistemically constrained notion of truth. But how does one get from the claim that truth is redundant to the conclusion that truth is (something like) warranted assertibility? The other problem is that the arguments attributed to Phillips are uniformly terrible. For instance, the entirely question-begging claim, ‘Truth never marks a property of judgements’, is used as the first premise of an argument Insole alleges Phillips uses for redundancy theory. Moreover, D. Z. Phillips, a ‘paid-up Wittgensteinian’, is generally construed as engaging in mostly hand-waving arguments to the effect that any realist construal of religious language must be ‘confused’. I was unclear, given how little Insole thought that Phillips had to offer, why he engaged in exegesis rather than looking at how more sensible generic arguments for anti-realism might be applied in religious discourse.
After looking at Phillips, Insole argues that epistemically constrained notions of truth must fail on the grounds that, ‘The belief that reality extends beyond the reach of possible human thought is closely analogous to something which we know to be the case’ (63). He gives the example of a blind community, the members of which cannot understand colour concepts, and of a community of nine-year-olds whose limited cognitive capacities mean that various (true) claims are beyond their grasp. Surely, Insole suggests, we could be in a comparable situation, where higher beings know truths that are inaccessible to us? If this is possible, Insole argues, then there could be truths that are unknowable, and therefore truth is not epistemically constrained (64).
To see why this argument does not work, it is worth briefly considering the position that is being criticized. A recent example of an epistemically constrained account of truth has been proposed by Crispin Wright. He contends that truth can be analysed as super-assertibility. A sentence is super-assertible if it is warranted given our current information, and will remain warranted under any further accumulation or other improvements of our information, and arbitrarily close scrutiny. (Wright, it should be noted, claims only that super-assertibility provides the necessary and sufficient conditions for the concept of truth – he is not saying that truth, in all fields of discourse, is nothing more than super-assertibility.) Does Insole establish that there could be truths that are not super-assertible? All his argument posits is that there could be sentences that we do not (and may never) know to be true, but that would be known to be true were we as well informed as the higher beings and our powers of scrutiny increased. So these claims are super-assertible. Insole's argument appears to conflate the anti-realist theory that all truths are in principle knowable, with the claim that no (clearly thinking!) anti-realist should make, that all truths are knowable to us.
In the second half of the book, Insole moves away from semantic realism/anti-realism to metaphysical issues such as Goodman's theories about world-making, and Kant's transcendental idealism. The fairly sympathetic treatment of Gordon Kaufman is notable. Later parts of the book review more theological literature and its Kantian influences, notably work by Paul Janz and James K. A. Smith. In the concluding sections, Insole looks at whether talk of God should be construed analogically (though he does not take up the debate about whether it should be taken metaphorically).
In all, Insole offers a review of relevant literature, and numerous arguments (of varying success) against religious anti-realism. He leaves us rather less clear on what his positive realist theory would involve.
