J. L. Schellenberg's study aims to provide a prolegomena to the philosophy of religion in two respects. It explores some fundamental issues that the author thinks should lie at the foundations of any study of the subject. It also lays down some thoughts re those issues that are to serve as starting points for his own philosophy of religion (to follow in later, promised volumes). In bare outline, his eight chapters cover: the definition of religion; the character of belief; the nature of religious belief and disbelief; types of religious scepticism; the features of religious faith (considered over two chapters); the aims of the philosophy of religion; and the principles for evaluating responses to religious claims.

The conclusions reached about the definition of religion in the first chapter can be summed up in one word: ultimism. Ultimism is summed up in the proposition ‘that there is an ultimate reality in relation to which an ultimate good can be attained’ (37). Schellenberg is concerned to define what Williams James had in mind by ‘personal religion’. Religion, in this sense, is introduced by way of four main features (13–14): (1) thoughts of a transmundane reality; (2) emphasis on a significant good that may be realized through relationship to this reality; (3) the cultivation of such a relation; (4) a disposition to totalize or ultimize the central elements of features (1) to (3). This fourth feature means treating, for example, the good in (2) as ultimate. Schellenberg claims (33) that his definition provides a point of orientation for the philosophy of religion. It fits major forms religion in existence today, while allowing us to recognize new forms of religion as such.

In his second chapter, Schellenberg turns to the concept of belief. He offers a dispositional analysis which signals a sharply non-voluntarist view of belief and separates the notion of belief from any essential connection with feelings of confidence or the like. He discusses and criticizes Swinburne's linkage between belief and estimates of probability (62–64). His definition of ‘S believes that p’ draws upon the work of L. Jonathan Cohen and comes down to: ‘S is disposed to apprehend the state of affairs reported by p, when that state of affairs comes to mind, under the concept reality’ (50). This definition is linked to earlier discussions in the chapter that connect believing that p with regarding the thought that p as a ‘world-thought’, as representing an actual state of affairs.

Chapter 3 explores the definition of religious belief-that and religious belief-in, and of their opposing forms of religious disbelief. Schellenberg thinks that to define religious belief we need to define the notion of a religious proposition. A religious proposition (following from the work done in Chapter 1) turns out to be a proposition entailing the existence of an ultimate and salvific reality (77). Religious belief-that and disbelief-that are easily characterized, given this move. But Schellenberg also wants to pick out a notion of affective religious belief and a notion of affective religious disbelief. These notions pertain to the possession of a positive or a negative attitude toward the value of an ultimate, salvific reality (85–86, 91–92). This enables him to present a broader picture of the territory of belief and disbelief.

Chapter 4 is concerned with the characterization of religious scepticism, understood not as the denial of religious belief but the attitude of doubt as to its truth or falsehood. It produces four types of religious scepticism (97ff.) which are then allowed to exist in either passive or active forms (101–105).

Chapters 5 and 6 form a connected discussion of the nature of religious faith. A brief summary cannot do justice the richness and detail of Schellenberg's treatment of the notion of faith and its allied concepts. A striking theme that governs both chapters is contained in this quotation: ‘contrary to popular opinion, religious faith neither is nor entails religious belief’ (106). Schellenberg goes further: in one of its main forms it is compatible with religious non-belief and entails it (106–107). Schellenberg's approaches the definition of faith through the notion of trust. Faith in a religious reality, such as God, is a form of trust in that reality. After some discussion, faith in a religious reality is defined in terms of the following conditions: faith or belief that the putative religious reality will be or do what the person of faith wants, the absence of sufficient evidence for this proposition's certainty, the expectation of bad consequences by the person if this proposition is false, and a disposition to act on this faith or belief (124). Along the route to this definition, Schellenberg rebuts attempts to insert some feeling of serenity or confidence into the definition of faith and contends that faith does not have to have as its object a salvific, ultimate reality that is personal. The Buddhist, no less than the Christian, can be described as having faith.

Schellenberg's second chapter on religious faith contains one of the most striking claims in his book: that religious faith can exist without propositional religious belief. Indeed, having what Schellenberg characterizes as propositional faith that p requires the absence of belief that p. There are already existing defences of non-doxastic religious faith, notably Audi's – as Schellenberg acknowledges (146–147). But Schellenberg's argument for this stance is significant for its depth and thoroughness. A five-fold definition of faith that p is finally offered on 139. To avoid collapsing faith-that into belief-that, Schellenberg has to characterize the attitude to the proposition that is the object of faith carefully. Without thought that p being a world-thought for this person (Chapter 2), the one who as faith that p ‘tenaciously and persistently represents the world to herself’ as including the state of affairs reported by p and ‘voluntarily and committedly adopts a policy of assent toward that representation’. Schellenberg contends that this form of representation and assent is an important alternative to belief (137). The voluntary character of these attitudes is clearly one thing that Schellenberg regards as importantly different from belief proper. If faith is voluntary it can neither be, not entail belief (154).

It will be seen that, having got thus far, that Schellenberg has broadened the range of matters that concern the philosophy of religion. There is more to religion than theism. There are more responses to religious propositions than belief. Thus when he turns, in Chapter 7 to discussing the aims of the philosophy of religion, Schellenberg contends that the subject should look beyond the justification of religious beliefs. There is a range of possible responses to religious propositions (beyond belief) and the philosophy of religion should be concerned with the worthiness of choosing or pursuing such responses (176). The range of responses is summed up at the start of the next chapter: believing, faith-full, purely sceptical, and disbelieving (195). There has been too much emphasis on the justification of religious beliefs, and not enough on that of religious practices, given that practice can flow from religious responses (like faith) that are other than belief (186). There is a general question of whether a religious response to religious claims is preferable to a non-religious one, and the further question of whether, if a religious response is justified, religious belief is justified rather than religious faith. This chapter contains various other suggestions for broadening the scope of philosophy of religion.

The eighth and final chapter of this study is largely devoted to outlining some general principles for evaluating responses to religious claims. Schellenberg comes up with no less than twenty-three of them!

From the above summary, it is clear that Schellenberg's Prolegomena provides a case for a substantive critique of the priorities of much recent English-speaking philosophy of religion. It has a preoccupation, not to say obsession, with the truth and justification of theistic belief, or, even more narrowly, Christian belief. Schellenberg's argument re the definition of religion in his first chapter and his cumulative argument from the second to the seventh chapters for a broader range of responses to religious claims than belief or disbelief, provide an important challenge to these priorities. There thus seem to be two kinds of interest this book might provoke and cater for. Some readers will focus on the commentary on specific issues – such as the definition of religion, the account of faith, or the analysis of religious scepticism. Others will be interested in the broad argument identified above. Both kinds of reader will find the positions and arguments in this book meticulously set out and worthy of very serious consideration.

As a contribution to reflection on Schellenberg's general argument, let me run the following line of criticism: has Schellenberg done enough to make his distinction between a believing and a faith-full response to religious propositions clear? This query can be reinforced from both ends, as it were. On the one hand, I find the account of belief offered in Chapter 2 less than perspicuous. I am not at all clear how far we get in understanding belief by way of talk of being disposed to apprehend a state of affairs under the concept reality. Ditto for talk of regarding the thought that p as a ‘world-thought’, as representing an actual state of affairs. These locutions can be accused of being just fancy ways of saying ‘to believe p is to hold that p is true’. (In a footnote on 45 Schellenberg denies that ‘to believe p is to hold that p is true’; falsely so, in my view.) As such, they would not get us very far. In particular, they do not seem to get beyond whatever we possess in our intuitive understanding of belief.

The doubt I thus raise is not driven by possession of a more adequate analysis of the notion of belief, but by scepticism as to how far any worthwhile analysis of it can be produced. To the extent that Schellenberg's characterization of belief is unclear, then so is his contrast between faith-that and belief-that. Everyone can accept that there are ways of using propositions in thought and action that fall short of belief (that is: holding them to be true). One can act as if p; one can adopt p as a supposition for the sake of argument; one can entertain p as a working assumption; and so forth. The problem I perceive in Schellenberg's version of non-doxastic faith comes about through the very ‘high’ terms in which he describes the manner in which the person who has faith that p assents to and makes use of p. To illustrate this we can note that Schellenberg writes of assent that is in faith:

And further: ‘In responding to the proposition that p in any of these ways, one brings it about that the proposition exerts a certain intellectual influence – even dominance – in those aspects of one's mental life where the truth of falsity of p matters.’ (137)

What Schellenberg appears to rely on in the discussion on 129ff to enforce the contrast between faith and belief is a series of examples where belief is absent, but faith is present. In the light of these examples, general differences are introduced (136–137): faith is voluntary, whereas the assent to a proposition in belief ‘is elicited by one's perception of the evidence, and any tendency to continue saying yes may be expected to be involuntary’ (136). I don't find the examples as clear-cut as Schellenberg. Further, one can ask in this context how far anyone could give the kind of assent to propositions Schellenberg has in mind under his notion of faith in the absence of evidence, that is truth-indicators. And finally, one can wonder how far it is right to contend that belief is as involuntary as Schellenberg makes out.

I raise the above criticisms while fully accepting the main thrust of this book: the philosophy of religion needs a more thorough study of prolegomena and that such a study will lead to the broadening of the subject's horizons.