The book is part of what seems to be a project by Swinburne of reviewing the ideas issued in some of his main works in light of the debate they have brought about or in view of new ideas which have occurred to him in the meantime. As with The Existence of God, the differences are considerable, but do not affect either the structure or the main conclusions drawn in the 1981 edition. In the following, I aim to present and discuss the main innovations of the second edition.
The first chapter (‘The nature of belief’) has some important changes. The major one is in the concept of propositional belief itself. Instead of describing it in terms of relative probability (‘believing so-and-so as more probable than such-and-such’), Swinburne has watered things down in the second edition, taking belief as the notion of ‘believing so-and-so as against such-and-such’. The distinction is not as subtle as it may seem to be, and the alteration is due to a criticism originally made by William Alston in ‘Swinburne on faith and belief’, in Alan Padgett (ed.) Reason and the Christian Religion (Oxford: Clarendon, 1994). According to Alston, Swinburne's first definition was too sophisticated, since young children, not to mention complex non-human animals, to whom we easily attribute beliefs, would hardly be expected to make any comparative probability assessment. In addition, for Alston, the original definition involved an infinite regress, since to believe that p would imply believing p as more probable than something else, and this in turn would imply believing that this belief is more probable than some other belief, and so on ad infinitum. As a result, no finite creature would believe anything.
Swinburne acknowledged this problem in Epistemic Justification (Oxford: Clarendon Press, 2001), 36n, and included the alteration in this second edition of Faith and Reason. As a result, propositional belief is now taken by him to be a mental state in which one piece of information is contrasted with another one, not necessarily in a conscious, elaborated form. Despite no longer figuring in the definition, probability is still a possible expression of what is involved in believing for those able to make probability evaluations. Hence, although believing in p does not necessarily mean believing that p is more probable than an alternative, if I do not believe that p is probable, I do not believe that p, which is very coherent with the Bayesian approach Swinburne gives to the problem of epistemic justification. However, having faith is not only a matter of propositional belief, but also a practical commitment in a relationship of trust. As a result, for Swinburne, the rationality of faith depends both on the epistemic justification of the correspondent propositional belief and on the coherence between the ends and means of an action. Despite being different, these types of rationality are correlated, since in order to achieve an end a person must hold some beliefs about this end, and the way to attain it.
Chapter 2 has been almost totally reshaped and expanded. In comparison with the first edition, Swinburne now elaborates more on the thesis that to hold true and justified beliefs is good, and on why it is so important to search for them. Being so important, we need clear criteria for assessing epistemic rationality, and the second edition comes with a new classification of rationality, which is largely based on the one proposed in 1981. The main difference is the discussion of externalist patterns for epistemic justification. Swinburne criticizes this approach, whose leading proponent in the philosophy of religion is Alvin Plantinga. The externalist approach focuses on the way a belief is produced in order to assess its epistemic credentials. These will be positive if the process of belief production is reliable in the sense of producing mostly true beliefs.
The major problem with this type of assessment for Swinburne is that it will give totally different results depending on the way the belief-forming process is defined. So, for instance, memory beliefs may be highly reliable in general, but this rate may be different for aged professors remembering facts which happened a long time ago. In other words, Swinburne argues, since there is no principled way of choosing one way rather than another, the externalist approach to epistemic justification is either arbitrary or unhelpful. It is along this line of argument that he criticizes Plantinga's theory of warrant, and defends his position that the highest level of belief justification depends on the use of correct inductive criteria and of its being probable in view of the total evidence available.
Chapter 3 (‘On the value of rational religious belief’) has also been expanded by the introduction of brand new material, and by the transference into it of the discussion about the application of rational belief criteria to religious belief originally made in Chapter 2. This time Swinburne adds a section on Christian reasons for believing that religious enquiry is pointless, and answers them by arguing that natural theology is central to the Christian tradition. On this line, the most interesting novelty in the chapter is another disagreement with Plantinga. For Swinburne, it is true that some people believe that there is a God as a result of a direct religious experience in a community where the existence of God is not challenged. In this case, their religious belief can be taken as properly basic. However, direct awareness of God is rare, and, in an age when religious beliefs are constantly confronted either by atheism or by religious pluralism, it needs be backed up by argument.
Chapter 4 (‘The nature of faith’) brings an important typology of Christian faith, which has been retained in this edition, although some important improvements have been made in the text. The Thomist view is still interpreted as predominantly a matter of belief-that, but this time Swinburne attempts to clarify it better by distinguishing faith, opinion, and scientific knowledge. Faith would be stronger than opinion, but would not amount to knowledge, because we do not fully understand why its articles are true. Another interesting new qualification to the Thomist view is the idea that faith is not incompatible with doubt, since, for Aquinas, according to Swinburne, the former can be a matter of degree (140). As to the Lutheran account, this edition elaborates a bit more on the concept of faith as sufficient for salvation, a traditional Lutheran stance.
For Swinburne, faith can only be so if it is moved by the love of God, since in order to be accounted as faithful, the Christian will need to have good purposes and perform good actions accordingly. These actions and purposes can be either good in a general moral sense (no matter whether or not there is a God) or in a strict religious sense (for actions like worship or asking God's forgiveness, for instance). So, the trust in God – emphasized by Lutheran faith – amounts to the same thing as faith formed by love – emphasized by Thomist faith. These remarks reinforce Swinburne's conclusion, drawn in the first edition, that the Reformation controversy about the role of faith for salvation is not a real disagreement between Lutherans and Catholics.
The account of the Pragmatist faith is the most reformulated of the three. Swinburne now quotes William James and Kierkegaard to show that faith as acting-as-if God exists requires at least some weak theoretical beliefs (e.g. that probably there is a God as Christianity depicts), otherwise it is simply incoherent. In addition, as a response to a criticism made by Stephen Maitzen (‘Swinburne on credal belief’, International Journal for Philosophy of Religion, 29 (1991), 143–157), he admits a fourth interpretation of what believing a creed amounts to. On this view, it means to believe the whole creed to be more probable than any alternatives to it as regards one item or other. This is a weak form of belief since it does not require that the whole creed is taken to be more probable than its negation as the sole alternative. Now, if the Christian faith requires only the minimal belief that its creed as a whole is more probable than various alternatives, regarding one item or another, then it is possible to take the three views of faith as basically the same. This time Swinburne ends the chapter by criticizing John Hick's thesis that there is a non-propositional view of faith as just seeing the world and automatically acting as if it is God's creation. Not surprisingly, for Swinburne, this should not be regarded as anything special or new, since it can be easily expressed in propositional terms without any loss.
Chapter 5 (‘The purpose of religion’) comes this time with a more extensive discussion about the concept of religion and a deeper account of the Christian and Buddhist ways of salvation. Religion is still defined as ‘any pursuit which dominates the life of many men and cements them together in common dedication to an aim’ (159). The aim and entities involved in the religious activity are generally extra-mundane, which excludes Marxism. Although it still serves as a good contrast to distinguish better the notion of religion, it sounds odd to talk about Marxism today as in the 1980s, since its force as an answer to the problem of the meaning of life looks much less (if at all) attractive now, even in developing countries. A more updated comparison could be with what some people call ‘fundamentalist atheism’ (espoused by authors like Richard Dawkins and Michel Onfray).
In any case, the importance here is in the description of the reasons for following a religious way, which is capital in the book since it comprehends Swinburne's main argument for the rationality of faith. In this edition, these reasons are better clarified and more extensively accounted for. All of them assume a religious point of view, that is, they start from the idea that extra-mundane realities are logically possible, and that there is a reasonable degree of epistemic justification in their correspondent assent by the religious person. From this, Swinburne argues in favour of the Christian way as the most founded religion, given its concept of God, and its depiction of salvation. In other words, for Swinburne there are good reasons to think that following the Christian religion leads to the most fulfilling type of salvation among the world religions, and that its concept of God and the relationship among human beings it commands are the most interesting and satisfactory. In addition, he claims that it is more probable to attain this best salvation through the Christian way than by means of any of its alternatives. In this spirit, it comes as no surprise that Swinburne criticizes Hick's pluralistic thesis that all religious ways have the same value (190–194).
After arguing that Christianity offers a better way of salvation than any of its main rival alternatives, in chapter 6 (‘The role of creeds’) Swinburne attempts to show that it is more rational as regards the credal dimension of religion as well. The question here is not whether the Christian creed is true, but whether it will lead to the goals of Christian salvation better than rival alternatives, which results in the predictable conclusion that the Christian creed is the best explanation of why following the Christian way will achieve the purposes of religion (203). Much more interesting is the argument for the idea that salvation is best achieved by the Christian way than by some other way (204), which comes in an account considerably improved over the first edition.
The criteria for assessing the rationality of religious belief are then summarized: (1) if achieving religious goals is much more worthwhile than achieving only mundane aims, then the search for salvation is rational in a means–end sense; (2) then it will be good to pursue a religious way, even if it is not too probable that following it will bring us to those goals; but it is previously required that (3) the religious creed is probable in an epistemic sense, at least more probable than its alternatives (222); and, in view of the Christian creed this requires basically that (4) the probability of the existence of God is not too low on the evidence available. For Swinburne, given that it is probable that there is a God, and, since this would make probable the deep well being depicted by the Christian way to salvation, and if one can show that there is no better kind of salvation than is offered by Christianity, then ‘it would be a rational5 act to seek it as energetically (for ourselves and others) as possible’ (224), even if based on a weak belief. A weak belief may make it harder to pursue the path of salvation, but this is enough as long as it is completed by strong dedication, and honest doubt is certainly compatible with the Christian creed.
The last chapter (‘The comparison of creeds’) has been almost totally rewritten, despite the fact that the idea is still to compare the Christian doctrine with other world religions' creeds. In this second edition, Swinburne is more explicit in the use of Bayesian epistemology here, although not employing the probability formalism. He first assesses the prior probability of Christianity in view of its internal coherence (especially regarding the doctrines of the Trinity, Incarnation, and Resurrection), and the explanatory power of the evidence available to it relative to rival theistic creeds (deeds reported in the sacred texts, purported testifying miracles, and the fidelity of their respective later churches to the teaching of their founder prophets). The part about miracles and the comparison with non-theistic creeds has been enhanced and expanded, and a section on ‘philosophical religions’ has been added. In another criticism of Hick, Swinburne presents an alternative account for religious pluralism having to do with his claim that investigation of the truth of religious doctrines is part of the rationality of faith.
The epilogue (‘Faith is voluntary’) keeps the original thesis that while belief is involuntary, faith requires voluntary effort, which is why it can be assessed regarding the faithful's merit. However, in this edition the text is more devoted to the argument that the search for religious salvation is not only coherent with the longing for mundane wellbeing but also a natural extension of it. The effort required by faith implies that such an all-important aim (‘the best thing for anyone to do with his life’ (268)) rationally involves constant investigation (266).
This new edition's insistence on the role of investigation for the positive assessment of the rationality of faith may be viewed as a new approach which Swinburne gives to the Bayesian methodology he adopted in the 1970s. Instead of his usual stress on the importance of determining prior probabilities in an impersonal and universal form, the high value he now acknowledges to investigation corresponds to a different aspect of Bayesian epistemology. According to this version, it does not really matter from which prior probability attribution to a thesis someone starts, for, as long as it is not dogmatic (0 or 1), it can always change in view of new data. In other words, one's degree of faith may be updated in view of new religious experiences, readings, discussions, and other forms of learning, which those who are disposed to follow a religious way will experience as part of their commitment.
Even so, Swinburne's account of faith still sounds too cerebral, almost heartless. Faith may not be just passion, as Kierkegaard says in one passage of Fear and Trembling, but a passionate devotion to God is certainly a dimension of it. In this respect, the faithful's attitude is similar to a football supporter's. His loyalty and confidence in his team is not a question of a calculated balance of its chance to win, but a passionate commitment that comes from love and dedication. It might be revealing of Swinburne's ideas that he does use the football metaphor – but not from the point of view of the supporter, but rather from that of the gambler. However, differently from the gambler, the supporter will not ground his faith (understood as his commitment to a football team) on the balance of probability only. He expects his team to win and hopes it will, even knowing that it may be objectively inferior to its rivals. It is not the probability of earning money that moves him, but the loving passionate commitment to the object of his love. In this regard, the worse the team performs, the more meritorious is the supporter's loyalty to it, and this is ‘part of the game’ for those involved in football. In other words, for those who take part in the football milieu, this kind of supporter's passionate commitment does not sound irrational.
However, before this starts to sound too ‘Wittgensteinian’ in D. Z. Phillips's sense, let us remember that faith is not the same as belief for Swinburne. While the latter is rational, as long as is proportioned to evidence and formed along acceptable inductive criteria, faith is also a matter of commitment, of dedication to following a given religious way. In order to be rational, faith needs some backing in probable propositional belief, but in a weak degree only, analogous to the minimum chance to win a match someone needs to believe his team has in order to keep on supporting it. For Swinburne, apart from propositional belief justification, faith is a question of means–end rationality, and in a world of religious scepticism and pluralism, this requires his investigating which is the best end, and the best way to reach it.
In this process of constant learning defended by Swinburne, one may say that the philosopher meets the religious, that is, the philosophical openness implies and is implied by the religious commitment to the search for deep answers and life's meaning. The need of permanent investigation is very much in line with the religious perception that the sacred, although partly revealed in the creeds, is essentially a deep and fascinating mystery. The humble recognition of this turns the religious into a philosopher, for the faithful will see himself in a permanent search for the ultimate that is neither totally incomprehensible nor fully understood. And the permanent search for being in touch with something or someone is a very common meaning one gives to ‘love’.
Apart from not being as heartless as it may seem at first, Swinburne's defence of Christianity as the best religious way, only outlined in this book, should not be seen as unilateral dogmatism. Rather it is, above all, an open contribution to a continous enquiring dialogue, which is much more enriched by clear positions both honestly expounded and liable to revision, than by a pluralistic relativism that is apparently generous, but which is in fact meaningless and unfruitful.
The world has changed considerably since 1981. Religion has come back into the public domain and has become once more the subject of an intense and rich discussion. For all his expansions and improvements, this second edition of Faith and Reason can be read as an entirely new book, very much fitted for this new age, with a considerable contribution to that debate. One thing, however, has certainly not changed from the first edition. Swinburne's style continues the same: arid, technical, rigorous, and not easy to read. But the book is well worth the effort.
