The modern problem of faith and reason
One sort of challenge to the rationality of religious commitment in the Judaeo-Christian tradition has it that faith is in some sense unreasonable because it involves believing on insufficient evidence. There is a well-trodden path to this conclusion, a line of argument that I shall call the Modern Problem of Faith and Reason. (If you like, think of it as just one of many problems, albeit an influential one.) Start with a sensible epistemological requirement such as:
(1) For any proposition, p, the belief that p is justified only if it is adequately supported by arguments or evidence.
Next, observe – as all of the accounts of faith to be considered in this essay shall take for granted – that religions such as Judaism and Christianity have cognitive content: a stake in some genuine claims about reality which are either true or false, including the core proposition that God exists. Then, examine the best arguments ever put forward for and against the existence of God and conclude that the project of offering publically accessible reasons for the existence of God is a failure:
(2) The belief that God exists is not adequately supported by arguments or evidence.
Take for granted a widely held principle about the relation between faith and belief:
(3) Faith in God is justified only if the belief that God exists is justified.
Several conclusions follow:
(4) The belief that God exists is not justified. (from 1, 2)
(5) Faith in God is not justified. (from 3, 4)
Clearly similar arguments could be run with respect to epistemic norms other than justification, such as rationality.
This sort of challenge, which I take to be implicit in a kind of cultural crisis of faith that developed in modern western culture, helps us to understand a great deal of work that has been undertaken in the philosophy of religion. There are two influential lines of response. The most straightforward reply argues that the evidence is sufficient, typically by appeal to natural theology, as a way of rejecting premise (2). Alternatively, reformed epistemology maintains that theistic beliefs can be properly basic in the sense that they are not based on argument or inferred from other beliefs but directly grounded, albeit fallibly and defeasibly, in experience. Alvin Plantinga (Reference Plantinga1981; Reference Plantinga, Plantinga and Wolterstorff1983; Reference Plantinga2000), for example, holds that certain sorts of religious experiences can directly ground the belief that God exists, along with a host of other beliefs, in much the same way that experiences spontaneously give rise to justified perceptual beliefs (e.g. I see a tree), memory beliefs (e.g. I had breakfast this morning), and beliefs about other minds (e.g. that person is in pain). On the one hand, we could see reformed epistemologists as contesting premise (1), or at least interpretations of it that do not recognize that spontaneously formed experientially grounded beliefs can be properly basic, by denying that the belief that God exists must be supported by argument or evidence in order to be justified or rational or reasonable or in some sense intellectually respectable. On the other hand, we could see them as endorsing the evidentialist principle articulated in premise (1), construed broadly enough to count things like conscious experiences and memories as evidence, while rejecting premise (2) in a way that is compatible with (though of course does not require) acknowledging the failure of traditional natural theology.
Assessing the adequacy of either of these replies to the problem is not my concern in the present essay. Instead, I show that both the perception that there is a problem and these attempts to reply rest on a particular set of assumptions about the nature of faith which are open to challenge; an account I call the belief-plus model. I distinguish between three accounts of faith (faith as belief-plus, trusting acceptance, and hopeful affirmation) each of which can claim some plausible grounding in the Judaeo-Christian tradition. Trust-based and hope-based models are worthy of careful consideration for several reasons. First, since trust and hope are compatible with doubt to an extent that belief is not, these can arguably make better sense of the fact that faith and doubt can coexist than can belief-based models. Second, questions about the rationality and justification of religious commitment look different on accounts of faith in which trust or hope are the primary basis for these commitments than they do on belief-based models in ways that make them attractive as candidates from which to develop an authentic and intellectually satisfying account of the relationship between faith and reason. On the trust-based and hope-based accounts to be considered, while the person of faith has a stake in the truth of the content (e.g. that God exists), neither requires acceptance of premise (3) in the Modern Problem of Faith and Reason. Voluntary decisions to trust can be made in situations of recognized risk and hope requires far less by way of evidence than belief with the same content. Moreover, non-epistemic considerations as well as epistemic considerations are clearly relevant to normative appraisals of hope and trust even if this is not the case for belief. The intellectual commitments involved in trusting and hoping can be made in situations where there is reason to doubt that their content is true, need not involve self-deception, and are compatible with the idea that one's purely epistemic opinions should be appraised by evidentialist norms.
Faith as belief-plus
While it is widely acknowledged that Christian faith involves more than belief, it is also often taken for granted that faith includes beliefs. For example, in The Reasonableness of Christianity (1695), John Locke takes it that, in addition to the call to do certain things such as repent and obey God's commands, Christians must believe a set of propositions, such as that God exists and that Jesus is the Messiah. Whatever other responses (e.g. behavioural commitments, values, affections, and so forth) faith involves one cannot have faith unless one has a particular attitude, belief, towards the content in question. Call this the belief-plus conception of faith.
We can characterize the belief-plus conception more precisely and also understand a key motivation for premise (3) of the Modern Problem of Faith and Reason by considering the distinction between belief-in and belief-that. Belief-that is a propositional attitude. To believe-that God exists is, roughly, to regard that proposition as true; to affirm a factual claim about what exists; to be of the opinion that God is real (that there is such a person as God). Is there any need for a contrasting concept, belief-in? Some usages of belief-in are simply alternate ways of reporting belief-that. To believe-in aliens or fairies typically just is to believe that aliens or fairies exist. Sometimes this is what people mean when they say that they believe in God. But Christian commitment does not consist in mere propositional assent, for ‘the devils also believe, and shudder’ (James 2.19). Belief-in, where it is not merely interchangeable with belief-that, is used to designate a host of other non-propositional attitudes or responses involving personal relations, including trust or other remaining commitments, affections, and positive valuations. These forms of personal response are sometimes taken to be the heart of what it is to have faith, such as trusting in God, loving God, dedicating or giving oneself over to God, pledging one's loyalty or allegiance to God, submitting to God's will and making God's purposes one's own. Devils might believe that God exists, but they wouldn't go on to do any of that. Notice that with belief-in of this sort the object is something other than a truth-valued proposition. Moreover, although there is some disagreement about what exactly to put on the belief-in side of the distinction and how much of it cannot be paraphrased away using belief-that discourse, the belief-in side of the contrast often serves as a catch-all which is supposed to include well-nigh all of the remaining affections, behaviours, and commitments (including what Price calls ‘esteem, or trust, or loyalty’) that can make up a life of faith.
On the belief-plus account, then, a response of faith requires both belief-in God and belief-that God exists. At least among philosophers of religion who take monotheistic religions to have a stake in some truth-valued claims about reality, the belief-plus model is by far the most widely held conception of faith. The important issue, both for understanding the belief-plus conception and for understanding the main argument for premise (3), concerns the relationship that is said to hold between these contrast classes. Belief-in, it is said, presupposes belief-that (Kenny (Reference Kenny1992), (Reference Kenny2007); Plantinga (Reference Plantinga, Plantinga and Wolterstorff1983); Price (Reference Price1965); Swinburne (Reference Swinburne1969) ). For surely, ‘I cannot trust my doctor unless I at least believe that there is a person to whom the description ‘being my doctor’ applies' (Price (Reference Price1965), 13). Similarly, one might trust God for many things, but one cannot very well trust God that God exists. As Plantinga has it,
One cannot sensibly believe in God and thank him for the mountains without believing that there is such a person to be thanked and that he is in some way responsible for the mountains. Nor can one trust in God and commit oneself to him without believing that he exists. (Plantinga (Reference Plantinga, Plantinga and Wolterstorff1983), 18; see also Kenny (Reference Kenny1992), 66)
Some will find alternatives hardly worth considering. Is not the statement ‘I have faith in God, but I do not believe that God exists’ absurd, logically fallacious, pragmatically incoherent, heretical, or some sort of revoltingly weak tea?
Given the belief-plus model of faith, the step to (3) is a short one. If you think that
(6) Belief in God presupposes the belief that God exists
then it will be difficult to resist some such epistemological requirement as
(7) Belief in God is justified only if the belief that God exists is justified.
(If you prefer, run this with other forms of positive epistemic appraisal such as rationality, reasonability, or proper basicality.) The belief-plus model has it that
(8) Faith in God just is belief in God.
It follows from (7) and (8) that
(3) Faith in God is justified only if the belief that God exists is justified.
Call this case for (3) the Logic of Belief Argument. You might think: right, that is just the way things stand. The Modern Problem of Faith and Reason must be confronted head on. The view that I shall defend, in contrast, is that the appearance of a problem subsides with the recognition of plausible alternative understandings of what faith is or could be.
Two difficulties for belief-plus models of faith: Mother Teresa and Meaning Drift
Faith, trust-in or reliance on God, is the primary term for the relationship of humans to God in the Judaeo-Christian tradition. Our focus is on the question: what sort of attitudes are candidates to play a primary role in the cognitive component of faith? In approaching this question, I grant that a useful distinction can be drawn between faith-in and faith-that. Jesus, as described in the gospel narratives, and Paul, through his epistles, confront us with both proclamation and invitation; with particular claims (e.g. God exists, Jesus is the Messiah) the truth or falsity of which are independent of what any of us think or how we feel and with instructions for appropriate response (e.g. repent, accept the gift of God's grace, seek to align yourself with God's will). A full response of faith involves both (a) in some way embracing, receiving, affirming, or assenting to the propositional content (perhaps believing, hoping, trusting, accepting it, holding it dear, or pledging our allegiance to it) (faith-that) and (b) undertaking to live in light of it; in relation to God (faith-in). But the terminology of ‘belief’ marks the contrast in a way that presupposes a particular view of the propositional attitude that faith requires. One way to have faith is to both believe-that God exists and believe-in God. The belief-plus account has it that this is the only way, asserting as it does that the propositional attitude required for faith is belief-that. The question of whether or not some attitude other than belief can play the role required for faith-that is precisely what is at issue in the alternatives to be considered below.
I shall begin by raising two difficulties for belief-plus models of faith. These difficulties serve both (a) to provide independent grounds for considering alternative understandings of faith and (b) to identify the sorts of roles that attitudes other than belief might be enlisted to play in connection with faith so that we can begin to assess whether they are up to the task.
Belief-plus models and the phenomenology of faith and doubt
Belief-plus models of faith fail to take seriously the phenomenon, surely an empirically well-grounded one, of religious doubt as it coexists with religious commitment. The intellectual content of faith, or part of it, is sometimes alleged to require and even to enjoy certification by high epistemic credentials. It has the status of knowledge, we are told, warranted by demonstration, direct perception, or the alleged infusion of grace. Perhaps it must even be accompanied by a conviction of certainty! Luther tells us: ‘The Holy Spirit is no Skeptic, and it is not doubts or mere opinions that he has written on our hearts, but assertions more sure and certain than life itself and all experience’ (Rupp & Watson (Reference Rupp and Watson1969), 109). Plantinga, following Calvin's understanding of faith as a form of ‘firm and certain knowledge’ instigated by the Holy Spirit (Plantinga (Reference Plantinga2000), 244), apparently finds himself with such confidence in God's existence that in some places he can say, ‘it is not now within my power to cease believing in God now’ (Plantinga (Reference Plantinga, Plantinga and Wolterstorff1983), 34).
Belief-plus models face what I shall call the Mother Teresa Problem, because it is embodied in her response to self-assured claims of the above sort: ‘Jesus has a very special love for you. As for me – the silence and the emptiness is so great that I look and do not see, listen and do not hear’ (this confession and further remarks concerning Mother Teresa's experiences of the hiddenness of God can be found in Kolodiejchuk (Reference Kolodiejchuk2007), 288). The problem is that the extent to which sincere and wholehearted religious commitment is compatible with doubt far outstretches the rather limited extent to which belief is compatible with doubt. For many, faith coexists with doubt, even profound doubts, and sometimes over long periods of life. Faith is clearly not incompatible with a persistent sense of uncertainty, dark nights of the soul, or a pervasive sense of the hiddenness of God. It is indisputable that many who profess to be devoutly religious simply do not find themselves with spontaneously formed confident belief. Moreover, entire movements of religiously committed intellectuals (e.g. Karl Barth's neo-orthodoxy) have rejected the project of natural theology, having examined the arguments for God's existence in the strongest forms on offer and found them wanting. There is a reason why the father's cry of ‘I believe [πιστεύω]; help my unbelief [ἀπιστία]!’ (Mark 9.24) resonates with people, even if they are unsure quite what it means. Some of these figures are greatly admired, seen as inspirational, or even taken as role models within their religious traditions precisely for the faith they exhibit. If deep, sincere, and wholehearted faith coexists with doubt in the lived experience of many religiously committed persons and can do so in a relatively stable way despite fluctuating levels of confidence, surely this fact is one that any adequate account of faith ought to be able to accommodate.
Belief-plus models define faith in such a way as to preclude significant doubt, yet faith appears to be compatible with doubt in a way or to an extent that belief is not. Just how compatible belief is with doubt will, of course, depend on what one takes belief to be. There is disagreement about how to relate talk of flat-out belief to subjective probability judgments (McKaughan Reference McKaughan2007). Plantinga and Alston are inclined to resist belief attribution even where personal probabilities are fairly high (Plantinga (Reference Plantinga2000), 271; Alston (Reference Alston, Jordan and Howard-Snyder1996), 6). Proponents of the belief-plus model need not assume that belief requires certainty, but there do seem to be some clear lower bounds. It is hard to see, for example, how someone could coherently believe that p while believing p to be less probable than its negation.Footnote 1 If we want to understand the various forms of sincere devotion to be found among religiously committed doubters, we therefore have good reason to examine other accounts of what it can be to have faith.
Belief ain't what it used to be
Belief-plus models have long enjoyed a kind of default status in reflection on the nature of faith in the Judaeo-Christian tradition. But the claim that believing – a concept often only loosely defined, poorly understood, and asked to cover far too much of the cognitive landscape (McKaughan Reference McKaughan2007) – is primarily what religious people (‘believers’) do can be challenged and a case can be made that faith is a more fundamental category than belief for understanding religious commitment. Moreover, the assumptions about the relationship between faith and belief that underlie belief-plus models, in particular the claim that propositional belief is the only attitude that can serve as an adequate basis on which to build a religious life within the boundaries set by these traditions, are at least more difficult to defend in light of a complex set of etymological and translation issues that I shall refer to as the Meaning Drift Problem. The uses we make of belief-related language have changed: formerly it was used to express a far broader array of thoughts, feelings, and practices than the much more restricted use it has come to have in contemporary epistemology.
In the Greek New Testament, the concept of faith is commonly expressed by cognates of the noun πίστις (pistis) and the verb πιστεύω (pisteuo; infinitive: pisteuein). In the Septuagint, the Greek translation of the Hebrew Bible, the pistis word family is also used to translate cognates of the noun אמונה (emunah) and the verb אמו (emun) (Bultmann & Weiser Reference Bultmann and Weiser1961). These Greek and Hebrew words are usually translated as ‘faith’ and ‘having faith’. However, English lacks a verb to refer to the act of faith and hence the verbs are often translated as ‘believe’ or ‘trust’. The accuracy of a translation obviously depends both on the meaning of the original words and on the meaning of the words used to translate them.
A case can be made that, given what ‘believe’ has come to mean in contemporary discourse, ‘trust’ or ‘have faith’ is almost everywhere a better translation for acts of faith. In contemporary philosophy and in much of today's ordinary discourse, while the mental experiences or feelings of confidence in a particular proposition (or associated behavioural dispositions) picked out by ‘belief’ are often seen as the causal products of largely passive psychological processes, beliefs are appraised as epistemic opinions. There are good reasons for making evidentialist evaluations of our opinions of the sort reflected in premise (1) of the Modern Problem. If my aim is to form an opinion solely about what is true (or about what other interesting semantic relations, such as empirical adequacy, a set of propositions bears to the world) or if this is the proper function of the faculties that produce the opinion, my attitude should arguably be formed only with respect to the evidence: information relevant to assessing its likely truth or falsity. Applied to the question of God's existence, it is often assumed that there are just three attitudes here and that these characterize differences in sober assessments of the evidential situation. ‘Believers’ are distinguished from ‘atheists’ chiefly by their confidence that God exists (or does not exist) and these from ‘agnostics’ – coined by T. H. Huxley in the nineteenth century – who withhold judgement.
This was not always the case. Quite often, in ancient Jewish and early Christian texts words translated as ‘believe’ served less to report finding oneself with a disinterested epistemic opinion than to express affections and allegiances involved in identity formation. ‘Belief’ was a profession of love for and allegiance to God; a public commitment, say, to follow Jesus. Wilfred Cantwell Smith, for example, mounts an extensive etymological study to argue that the meaning of the English word ‘believe’ itself has shifted in the modern era (Smith (Reference Smith1998a), 41; see also Smith (Reference Smith1998b) ). While belief now refers to a state of mind, a disposition to assent to a set of propositions, even within the early Christian intellectual tradition historically it had as much or more to do with love, loyalty, and commitments akin to pledging one's allegiance to a person as Lord or to a cause or to entering into a covenant such as marriage. The Latin word credo (apparently a compound of cor, cordis ‘heart’ and -do, -dere, ‘to put’ derived from the proto Indo-European root for placing one's heart upon something, *kred-dhē) means ‘I set my heart’ upon the entity or doctrines in question. Even for scholastics such as Aquinas, Smith argues, credo meant to pledge allegiance to, to give one's self and one's loyalty. The Latin terms most closely expressing today's meaning of belief and opinion, opinio (‘opinion, belief, supposition’) and opinor (opinari, to be of the opinion, to believe) played an almost negligible role in Christian thought. The German word belieben (‘as you like’ from the German root Liebe ‘love’ and Latin libido ‘desire’), an etymological counterpart of the English word ‘believe’, means to prize, to hold dear, to cherish, and to love. An increasing preoccupation with belief as a characterization of what religious people centrally do, in combination with a failure to recognize subtle but substantive changes in the use of belief-related language, has arguably distorted the historical fact that faith, rather than belief, has been the fundamental category for religious commitment in these traditions. But what, then, is faith or what else could it be?
Faith as trusting acceptance
Robert Audi (Reference Audi1991; Reference Audi and Hester1992; Reference Audi and Stump1993; Reference Audi and Timmons2007; Reference Audi2008; Reference Audi2011) and J. L. Schellenberg (Reference Schellenberg2005; Reference Schellenberg2007; Reference Schellenberg2009) have both proposed accounts of faith that take trust as the central cognitive attitude. Richard Swinburne's (Reference Swinburne2005) pragmatist conception of faith and William Alston's (Reference Alston, Jordan and Howard-Snyder1996, Reference Alston and Timmons2007) defence of ‘acceptance’ also deserve consideration under this heading. Without pretending to do justice to the details of their own particular views, I shall draw on the fruits of their work as part of a general discussion of the role that trust might play in an account of faith.
What is trust and what sort of role might it play in religious commitment? Start with two observations. First, the accounts we shall consider treat trust as an action or disposition to act; as a decision to rely on the person or object in question. Second, trust and the associated concepts of trustworthiness or faithfulness, in their primary usages, involve personal relations. Thus, focusing on faith in a person, Alston takes it that:
Here the crucial feature would seem to be trust, reliance on the person to carry out commitments, obligations, promises, or, more generally, to act in a way favorable to oneself. I have faith in my wife; I can rely on her doing what she says she will do, on her remaining true to her commitments, on her remaining attached to me by a bond of love. (Alston (Reference Alston, Jordan and Howard-Snyder1996), 13)
Our best examples of trust are of trust-in persons. In Faith and Reason, Swinburne proposes an analysis of trust-in ordinary persons and in God based on the consideration of paradigmatic cases of trust such as lending something valuable to a careless or irresponsible friend, an escaping prisoner who trusts an enemy guard, a patient trusting in a doctor for a cure, and the like. As he has it, ‘To trust someone is to act on the assumption that she will do for you what she knows that you want or need, when the evidence gives some reason for supposing that she may not and where there will be bad consequences if the assumption is false’, where to act on the assumption that p is ‘to do those actions which you would do if you believed the stated assumption strongly’ (Swinburne (Reference Swinburne2005), 144). Since the actions that one performs when acting on the assumption that p depend on what one's purposes are, Swinburne adds the condition that the trust-in God that forms part of Christian faith includes seeking ‘to do those good actions which the love of God (if there is a God) would lead him to do’ (ibid., 148).
Some form of trust-in God is likely to play a role in any account of faith. But for many doubters, their question is not so much whether God, if God exists, is trustworthy. Rather, they are assailed by doubts about whether there is such a person as God. Is there also a propositional attitude, trust-that or something in the neighbourhood of trust, that can help us to understand what is or can be involved in having propositional faith (faith-that)? Extending Swinburne's analysis to the propositional case, to trust-that p, would be (1) to act on the assumption that p, (2) when the evidence gives some reason for supposing that not p and where (3) there will be bad consequences if p is false. I could, then, decide to trust that God exists, even in the face of good though not decisive reasons to doubt, provided that this is an assumption on which I am willing to act; to arrange my life around. This response could even be, as Kierkegaard would have it, an intensely personal and totalizing form of involvement: ‘the idea for which I can live and die’ (Dru (Reference Dru1938), 15). We have not yet seen that this could be a wise or reasonable or psychologically easy thing to do, but I see no purely conceptual reason that this cannot be done.
Another attitude, perhaps even more clearly voluntary and more clearly appropriate to propositional commitment than trust, is acceptance-that. The most influential discussion of acceptance, as a form of voluntary assent explicitly contrasted with belief in the sense of passive conviction, is that of L. J. Cohen (Reference Cohen1989, Reference Cohen1992). Although his full account contrasts these attitudes over seven separate issues, some of which are more controversial than others, the basic idea is straightforward.
In my sense to accept that p is to have or adopt a policy of deeming, positing, or postulating that p – that is, of going along with p (either for the long term or for immediate purposes only) as a premiss in some or all contexts of one's own and others' proofs, argumentations, inferences, deliberations, etc., whether one assents and whether or not one feels it to be true that p. (Cohen (Reference Cohen1989), 368; see also Cohen (Reference Cohen1992), 4)
This premising policy is something Cohen takes us to be able to adopt ‘at will’ by a decision: ‘You answer the question whether you accept p by making or reporting a decision’ (Cohen (Reference Cohen1989), 368). Accepting that p, then, is a voluntary mental act that involves the adoption of a policy of regarding or treating p (and what one concedes to be deductive consequences that follow from p) as true in one's conscious reasoning whether or not one feels p to be true.
Alston (Reference Alston, Jordan and Howard-Snyder1996; Reference Alston and Timmons2007) registers some minor points of disagreement with Cohen, but also recognizes belief and acceptance as two distinct attitudes. He agrees with Cohen's characterization of belief as (a) dispositional and (b) not under our direct voluntary control and the contrasting concept of acceptance as (a) a mental act that is (b) under our direct voluntary control in the way deciding to raise my right arm is under my control. Alston proposes an account of faith in which the propositional attitude involved in faith-that is or can be acceptance. He takes trust to be the attitude involved in faith-in. Since Alston does not ‘see a great difference between “belief in” and “faith in”’, what he is suggesting is that ‘I can believe in God (trust in his providence) while accepting that He exists, rather than firmly believing this’ (Alston (Reference Alston, Jordan and Howard-Snyder1996), 14). Some ‘sincere, active, committed, devout Christians … are troubled by doubts; they ask themselves what reasons there are to believe that all this really happened. They take it as a live possibility that all or some central Christian doctrines are false’ (ibid., 16). Nevertheless, even if they do not believe the doctrines, they may decide to accept them.
To accept them is to perform a voluntary act of committing oneself to them, to resolve to use them as a basis for one's thought, attitude, and behavior. (And, of course, it involves being disposed to do so as a result of this voluntary acceptance.) (ibid., 17)
Notice that acceptance is a truth-oriented attitude; a decision to regard something as true. The idea is that by an act of deliberate choice, even in the midst of doubt and while perhaps lacking the spontaneous feeling of confidence enjoyed by the believer,
the accepter can be as fully involved in the form of life, and not just on an ‘as if’ basis. This is not a matter of resolving to act as if the doctrines are true, while not really taking seriously the idea that they are true. To accept the doctrines is to accept them as true. Since Jesus was resurrected is true if and only if Jesus was resurrected, I can't accept the latter without at least being committed to accepting the former. (ibid., 18)
While one could accept that p for purely epistemic reasons, acceptance differs from belief. Practical considerations can clearly be relevant to acceptance in a way that they are not, at least arguably, to belief. Keith Frankish gives an example in which belief and acceptance come apart for precisely this reason:
Suppose I believe that the gun in my desk drawer is unloaded. And suppose I am now offered a small sum of money for taking the weapon, aiming it at the head of a loved one, and pulling the trigger. In deciding whether or not to accept this offer, I might, quite reasonably, refrain from relying on the proposition that the gun is unloaded. (Frankish (Reference Frankish2004), 134–135)
One can believe that the gun is unloaded without taking it as a premise in one's actions and one can accept that it is loaded without believing that it is. In this case, the combination of desires, beliefs, and tolerance for risk lead me to handle the gun cautiously. In other situations, such as the choice to trust someone who has let me down in the past or the decision of a rescue team to continue the search for workers who were trapped in a collapsed mine, practical considerations might lead me to accept a proposition as a basis for action that I don't have good reason to believe.
It looks, then, as though we have several good candidates for faith-that which involve trust-that or something like it. Call the suggestion that trust-that or acceptance-that God, as portrayed in some particular religious tradition, exists could anchor trust-in God the trusting acceptance conception of faith. We might require, as Audi (Reference Audi2008, 93) and Schellenberg (Reference Schellenberg2009, 209) propose, that the cognitive or propositional attitude which can play a religiously significant role in faith-that involves both (1) a positive evaluative attitude towards the object or content (e.g. S thinks that it would be a vitally or momentously good thing if x exists or if p were the case) and some kind of (2) cognitive commitment (e.g. S is committed to acting on the assumption that p is true in some or all contexts). Trust of this sort, a cognitive resolution to assume that x exists or that p is true in one's intellectual and practical conduct, is clearly compatible with having some reason to doubt that x exists or that p is true. If trust-that involves or is associated with positive value-judgments, this might well make a difference to its practical (non-epistemic) rationality. On this model, faith is seen as a ‘risk-taking behavior on behalf of the good’ (Schellenberg (Reference Schellenberg2009), 40–44).Footnote 2 While trust allows wider latitude for doubt than belief, trust is not entirely unconstrained by one's epistemic opinion. Trusting that x exists or that p is the case coherently, for example, is arguably incompatible with S's also flat-out believing that x does not exist or that it is not the case that p. However, before discussing how epistemic norms such as rationality and justification might apply to faith so construed, there is at least one more alternative to the belief-plus conception that we need to put on the table.
Faith as hopeful affirmation
Now consider hope. James Muyskens (Reference Muyskens1979), Louis Pojman (Reference Pojman1986a; Reference Pojman1986b; Reference Pojman, Loades and Rue1991), and William Lad Sessions (Reference Sessions1994) have each proposed accounts of faith that take hope as the central cognitive attitude. Pojman claims that:
If belief-in, or trusting, can be analyzed in terms of commitment to a course of action or a disposition to act, then it seems that we do not need to believe-that x exists in order to believe-in or deeply hope in the existence of x. (Pojman (Reference Pojman1986b), 224)
But what is hope and is this claim plausible?
Hope is a complex attitude that involves both evaluation and opinion or, at least, some relatively weak constraints on opinion. If I hope for sunny weather on my sister's wedding day, ordinarily this will involve both a desire that the weather be sunny and a belief, say, that this is at least possible. Notice that I can hope for sunny weather even if I believe that alternatives like rain or even snow are more likely. While there are differences of opinion concerning just how hope is to be analysed, quite generally, it seems that, for any subject S and proposition p, to say that S hopes that p involves at least that (1) S desires that p and (2) S does not believe that p is impossible. Clearly hope is also an attitude one can have towards the existence of an object, entity, or person x (e.g. God) or the obtaining of some state of affairs. These conditions are arguably necessary minima for hope. It would make little sense to say Dave hopes that his wound will heal quickly and not become infected but has no desire that this be the case or that he believes that this is impossible. But perhaps a religiously significant sense of hope requires a bit more. As stated, the first condition leaves the nature of the desires quite unspecified (e.g. are these emotions, considered value judgments, or what?); ‘impossible’ in the second condition might mean only logically incoherent. A plausible case could be made, for example, that the second condition for religiously significant hope should be that p is a live option for S or that S believes that the probability that p is true is not so small as to be negligible or that S does not believe not-p.
The combination of desire and opinion that hope involves gives it some interesting semantic properties. Hope has multiple directions of fit. Whereas beliefs are satisfied when they correctly represent how things stand, desires are fulfilled when the world conforms to the content of the desire. In each case we look for a match between content and world, but the direction-of-fit of an attitude or speech act identifies whether to place the blame, if you will, on the content or on the world if there is a failure of match. To be sure, the hope that God exists, the belief that God exists, and the trust or acceptance that God exists are all vindicated if and only if there really is such a person as God. All of these attitudes have a stake in the existence of God and share the same conditions of satisfaction. But their appraisal conditions differ.
Although hopes can be misplaced, the minimal epistemic opinion involved in hope is a very weak one. Indeed, hope is most nakedly apparent in cases where something is hoped for despite its improbability. Moreover, and for this reason, the hope that p requires less, often far less, in the way of evidence to be rational than the belief in that same content p. It can be reasonable to hope that p in cases where belief with the same content would not be. Clearly, I can hope to win the lottery jackpot without believing that I will and indeed while believing that it is extremely unlikely that I will; that the odds of winning are about one in two hundred million. Lying blind and paralysed in a ditch, I might hope to see and walk again. Devastated by the kidnapping of her child, years later, a tearful mother might still hope to be reunited with her son. Enslaved, I might hope one day to be set free. Similarly, one can hope that God exists without believing that God exists.
At a phenomenological or psychological level the fact that hope can serve as a source of strength or powerful motivator to carry on against the odds is familiar. Clearly having a hope that p can sometimes be sufficient to motivate and guide reasonable action and do so in cases where one lacks that belief that p. As Jeff Jordan memorably points out:
A castaway builds a bonfire hoping to catch the attention of any ship or plane that might be passing nearby. Even with no evidence that a plane or ship is nearby, he still gathers driftwood and lights a fire, enhancing the possibility of rescue. The castaway's reasoning is pragmatic. The benefit associated with fire building exceeds that of not building, and, clearly, no one questions the wisdom of the action. (Jordan (Reference Jordan2006), 1)
But it seems incorrect to claim, as Muyskens does, that if one hopes that p one is disposed to act as if p (Muyskens (Reference Muyskens1979), 15). I might hope that I win the lottery, but I do not thereby begin acting as if I will win and it would be ill-advised to do so. Similarly, one could hope that God exists while lacking further commitments to live in obedience to God's will characteristic of religious faith. Hope can motivate, but does not itself require, action. Hopes, like wishes, can be idle. Moreover, a case can be made that hope is not particularly voluntary. If hope involves or even consists in having a certain combination of beliefs and desires, then, to the extent that these are dispositions that we simply find ourselves with or not, neither will hope be under our direct voluntary control. Whereas trust makes up for what it might lack by way of epistemic confidence by leaning heavily on actions or dispositions to act rather than desires, hope is an attitude that places more of the emphasis or ‘weight’, if you will, of the attitude on the evaluative or affective component than on the epistemic opinion.
The mere hope that a particular set of propositions with religious content is true, just by itself, would not constitute a complete response of faith. Still, if the hope is not the whole of faith, it might nevertheless play a crucial role in grounding religious commitment by serving as the cognitive attitude involved in faith-that around which one could build a religious life. The model of faith that I shall call hopeful affirmation combines a hope-that God exists with an accompanying set of affective and behavioural commitments characteristic of trust-in God as discussed above. A deep hope that God exists might, for example, motivate and guide the decision to live and act on the assumption that God exists which characterizes faith-in. Pojman seems to me to be an example of someone who conceives of faith as a kind of hopeful affirmation and the defensibility of such models of faith at least deserves serious consideration.
In Homer's Odyssey, Penelope waits twenty long years for Odysseus' return despite evidence which raised reasonable doubts about whether he was still alive. One could arguably do something like that, even where the cognitive attitude towards Odysseus' existence was something other than belief that he remained alive; even where it had dwindled to a mere hope that this so. Reports of his death or capture might make it unreasonable to believe that Odysseus is alive. But suppose these are not so conclusive as to make it unreasonable not to believe that he is dead. We might eventually come to regard the Penelope figure as pitiable, maybe even foolish. But suppose that she asks:
Where is the foolishness here? There are very different paths I could take in life: remarry or wait. I recognize the force of the counterevidence all too well but choose to continue to live in the hope that he has not perished and will someday return. I acknowledge the risk – the real possibility that Odysseus will never return – and I accept the consequences of a life lived in waiting even if my hopes should never be fulfilled.
I see no offence to reason in her decision; no intellectual responsibilities that she is flouting. The case seems to show that, in terms of intellectual commitments on the issue of whether or not Odysseus is alive, the form of life that Penelope undertakes can be grounded in or accompanied by the mere hope that this is so. Her form of life, which once might crudely have been classified on the belief-in side of the distinction, does not require a belief-that Odysseus exists.
A road less travelled? Faith and reason in a new light
If we could come to see either the trusting acceptance or hopeful affirmation models of faith, or something like them, as both (a) adequate from the perspective of the religious form of life in which a person might desire to be involved and (b) as at least rationally permissible from an epistemic point of view, this would be an important step towards dissolving an apparent tension between faith and reason that has troubled many reflective religious seekers.
With respect to (a), some – perhaps not persuaded by the Penelope example – will find it ‘difficult to see how one can have a sustained policy of using a claim with full weight in one's practical reasoning and not believe it’ (Poston (Reference Poston2009), 152).Footnote 3 We have seen how both the phenomenology of faith and doubt and a properly contextualized understanding of faith within the Judaeo-Christian tradition might provide independent motivations for (a). Perhaps the closing illustration can also speak indirectly to those interests. But judgements with respect to (a) are bound to be highly personal and dependent on one's other values.
With respect to (b), the goal of these final two sections is to explain how seeing faith as either trusting acceptance or hopeful affirmation might open a tenable, though perhaps less travelled, path of response to the Modern Problem of Faith and Reason. On trust-based and hope-based accounts of faith, premise (3) of the Modern Problem – the claim that faith in God is justified only if the belief that God exists is justified – is false. Because faith in God does not require justified belief that God exists, even if one were to accept both the evidentialist requirement articulated in (1) and the claim that the arguments or evidence for God's existence do not adequately support belief as the epistemically appropriate attitude expressed in (2), the conclusion (5), that faith in God is thereby unjustified or in some sense irrational, does not follow. A person of faith need not (but could) appeal to arguments in natural theology or to the self-assured stance of reformed epistemologists. Attitudes such as trust, acceptance, or hope that God exists might still be held without flouting any plausible intellectual responsibilities even if the evidence is not sufficient to justify beliefs with the same content. However, the main target of my critique, premise (3), is supported by several powerful (but, in my view, not insurmountable) considerations to which I shall attempt to respond in this section and the next.
One consideration that might prevent us from seeing alternatives to the belief-plus model of faith as viable options is reasoning along the lines of the Logic of Belief Argument, the subargument for (3) considered above. However, I think that there is a plausible route by which to reject this argument: What I reject is the conjunction of (6) and (8). Does belief in God presuppose the belief that God exists, as (6) has it? If we construe belief-in narrowly, distinguishing from ways that faith-in God is understood on either trusting acceptance or hopeful affirmation models, then (6) may be true but irrelevant since that reading of belief-in renders (8) false: in that case faith-in God would not be interchangeable with belief-in God. Alternatively, if we construe belief-in broadly, using it interchangeably with faith-in, then this renders (8) true but faith as trusting acceptance or hopeful affirmation stand as counterexamples to (6).
A bit more explanation is in order. In order to evaluate whether belief-in x presupposes belief that x exists, where the values of x are entities such as persons, we need to get clear about whether we are talking about ‘belief-in’ in some well-specified sense or merely as a catch-all for all the non-propositional attitudes involved in faith (faith-in). Advocates of principles like (6) face a dilemma. Suppose that we take belief-in broadly, as a catch-all for all the non-propositional attitudes involved in faith including behavioural policies. This would allow for readings that make belief-in, trust-in, and faith-in interchangeable as (8) has it. Some philosophers have thought that there is a logical difficulty with denying principles like (6) and even make the stronger claim that trust-in presupposes belief-that. James Muyskens, for example, holds that ‘Logically, one cannot trust-in someone if he does not take it to be beyond reasonable doubt that that person exists’ (Muyskens (Reference Muyskens1979), 63). But where is the logical relation which tells us that trust-in x requires us to believe that x's existence is ‘beyond reasonable doubt’ rather than ‘more probable than not’ or some other propositional attitude entirely, as our trust-based and hope-based models have it? Lacking the belief that p isn't the same as believing that not-p nor is it incompatible with having some other positive propositional attitude towards p. Once we recognize trust-that, acceptance-that, or hope-that as cognitive attitudes that could accompany a rich set of affective and behavioural commitments the charge that (6) can only be denied on the pain of logical incoherence collapses. Moreover, it is quite clear that (6), broadly construed, does not hold generally. Take R. B. Braithwaite's (Reference Braithwaite1955) non-cognitivist account of religion as a vivid counterexample. Braithwaite's approach to religious discourse recasts apparent claims about reality as statements of intention ‘to carry out a certain behaviour policy’, shaped by emotional responses to religious stories. Having reinterpreted the language in this way, as he sees it, questions about the truth or falsity of propositions such as God exists or there is life after death simply do not arise. This view takes ‘a Christian's assertion that God is love’ simply as a declaration of ‘his intention to follow an agapeistic way of life’ (Braithwaite (Reference Braithwaite1955), 18). Such policies can have content in the sense that they rule out alternative courses of action, but the content is not propositional content. Clearly one can act as if God exists without any cognitive component or truth-valued religious commitments at all. However inadequate Braithwaite's views might strike us as an interpretation of religious language, the problem is not that it is logically impossible to pursue the behavioural policies he recommends in the absence of belief that God exists. At best there would be a contingent motivational difficulty as expressed by the person who responds to Braithwaite with an incredulous stare: why would anybody do a thing like that?
Alternatively, suppose that we construe belief-in narrowly and can satisfy ourselves that there is something clearly identifiable as a belief on the belief-in side of the distinction (perhaps being confident about what the person who is the object of the belief-in will do as opposed, say, to policies of action voluntarily undertaken for practical reasons). Schellenberg (Reference Schellenberg2005, 72–74), for example, distinguishes sharply between belief-in or ‘affective belief’ (which he characterizes as a passive positive emotion or involuntary feeling of approval that attends or accompanies consideration of an object, state of affairs, or propositional content of the belief) and faith-in or ‘operational faith’, where to have faith-in x is, roughly, what Swinburne means by trust-in: to be disposed to act on the assumption that the propositional content believed or accepted as part of faith-that is true, while being uncertain whether it is true, and where bad consequences will ensue if it is false. On the narrow reading of belief-in, principles like (6) might well be true. But, since we can no longer equate faith-in or trust-in God with belief-in God, (8) will be false and we are left with plenty of room for accounts of faith on which faith-that can involve something other than belief that.
Epistemic norms and the relevance of practical considerations to religious commitment
The task of clarifying the precise role that attitudes such as hope, trust, and acceptance can play in connection with religious commitment, reflecting on their epistemological implications, and elucidating plausible constraints on the cognitive dimensions of these attitudes and combinations of attitudes is one of the most exciting projects in philosophy of religion today. If faith does not require belief, we must also take seriously, as Audi suggests, the ‘possibility that faith, as a central element in religious commitment, can be rational even if theistic beliefs with the same content should turn out not to be’ (Audi (Reference Audi1991), 213; see also Audi (Reference Audi and Hester1992), 64).Footnote 4 Each of the models of faith before us – belief-plus, trusting acceptance, and hopeful affirmation – allows faith to be responsive to evidence-based challenges and religious commitments to be appraised in light of epistemic norms. So faith need not be dogmatic, even if the attitudes and values that motivate and guide faith commitments are often non-epistemic in character. One point crucial to the discussion is that hope, trust, and acceptance are only partly epistemic attitudes. Practical considerations, such as one's values and goals can be relevant to hope, trust, or acceptance even though they are not good reasons for taking something to be more likely to be true. Each of these attitudes is compatible with having a wide range of purely epistemic opinions (opinions formed solely with respect to considerations that one takes to bear on the truth or falsity of a proposition). The results of this larger ongoing project will be relevant to evaluating another seemingly formidable challenge to the response to the Modern Problem of Reason that I am proposing – a challenge concerning the epistemic constraints that might attach to trust. In closing, I take up the more modest aim of raising the challenge and responding with an illustration of the kind of package of attitudes that I think can overcome it.
Even if faith-in without belief-that involves no logical incoherence, one might object that it would be silly or pointless or in some way unreasonable to trust-in x while lacking the belief-that x exists. In particular, one might defend premise (3) by arguing that having trust-in God in the absence of belief-that God exists would be an improper cognitive response given one's epistemic opinion about the evidence. Call this the Improper Response Defense of (3) because, according to it, having a combination of attitudes that violates (3) must involve an improper, disorderly, or irresponsible way of conducting one's intellectual life. Plantinga gets to the heart of the matter:
One cannot sensibly believe in God and thank him for the mountains without believing that there is such a person to be thanked and that he is in some way responsible for the mountains. Nor can one trust in God and commit oneself to him without believing that he exists. (Plantinga (Reference Plantinga, Plantinga and Wolterstorff1983), 18)
Terence Penelhum also feels the strong pull of this intuition: ‘I think it is self-evidently absurd to speak of having trust when one has no belief in the existence of an object of it’ (Penelhum (Reference Penelhum1983), 180; see also Kenny (Reference Kenny1992), 66).
Plantinga and Penelhum seem to be making the following claim about trust:
(9) Trust-in x requires the accompanying belief-that x exists.
By itself (9) seems simply to assert what I have called into question, namely, the central tenet of the belief-plus model, which has it that no propositional attitudes other than belief could ground a religiously adequate faith-in. It appears that (9) is put forward as a normative epistemological constraint on trust (and, at any rate, if it is instead construed as a claim about the logical relations between trust-in and belief-that, I have already argued against this in the previous section).
Construed as a normative claim, (9) sets up standards for trust that do not hold even on the level of human relations. Here is one place where one's values and goals are relevant to the appraisal of the rationality of trust and acceptance. To see the problem with (9), start by observing that there are circumstances in which one can have good reasons to trust person x (trust-in x), even in the face of evidence against the trustworthiness of x or even if one lacks the belief that x is trustworthy. Instances of this sort are common. As Penelhum points out, one reason to trust someone, such as a family member or a poorly motivated student, can be to give them ‘a chance to behave better than the evidence suggests they may’ (Penelhum (Reference Penelhum1983), 74). My desire to help a hitchhiker might lead me to trust someone with little, if any, justification for believing that the trust is well placed. Swinburne's prisoner of war might decide that trusting an enemy guard gives him the best chance of getting home, even if his opinion is only that there is some non-negligible probability that the guard will help (Swinburne (Reference Swinburne2005), 144). Next, as a case of trust-in x in the absence of belief-that x exists, consider a voicemail that begins ‘By the time you get this message, I will be dead. But I promise to bury your share of the treasure in the prearranged location and not to tell the killers where to find you’. Surely I could trust this person to have followed through on her promise, even if I believe that she is now dead. There is no logical difficulty here and, in the right context and evidential circumstances, I could be justified or rational in doing so. In the case of God's existence, Swinburne has argued, plausibly, that one can have good practical reason to trust-in God even in the absence of belief-that God exists if one (a) believes that pursuing goals (of, say, entering into a loving relationship with God, attaining eternal life, avoiding damnation, or what have you) is far more worthwhile than more mundane goals and (b) believes that, if there is a God, it is at least as probable that one will attain the goals, say, by accepting the gift of God's grace and living in accordance with God's will as by doing anything else, including doing nothing at all:
If your purpose to achieve X is strong enough (is far stronger than your other purposes) then you will still do A even if you believe that p is not very probable … [even if you are guided by] the belief that there is at least a small, but not negligible, probability that p. (ibid., 143)
But perhaps proponents of the Improper Response Defence of (3) could motivate it with a slightly different assumption about the epistemic constraints that attach to trust as a propositional attitude:
(10) Trust-that x exists requires the accompanying belief-that x exists.
Prima facie, (10) looks plausible and, if it is correct, this would imply that trust-based faith could only be justified or reasonable under circumstances in which belief-based faith was also.
The first point to see about (10) is that a similar charge cannot plausibly be made against hope. It is clear that one can hope that x exists without believing that x exists. So, insofar as hope-that can be seen as a religiously adequate ground for faith-in, premise (3) in the Modern Problem of Faith and Reason can be dismissed. But suppose one thinks that hope-that is not religiously adequate, what then?
The epistemic constraints on trust-that and acceptance-that are less clear.Footnote 5 I do not deny that there are certain evidential conditions – such as those evidential conditions that pertain to Santa Claus – under which faith-in, on the trusting acceptance model, is inappropriate. Our question is rather: are there any circumstances under which one could sensibly, in the absence of belief that God exists, couple whatever affections and behavioural commitments are involved in faith-in with a decision to act on the assumption that God exists?
The kernel of truth in the Improper Response Defence and, I suspect, the source for much of its intuitive motivation is the thought that one cannot sensibly trust God that God exists, because God's existence is precisely what is in question. But why think that alternatives to the belief-plus model require such a manoeuvre? Notice that this kernel of truth does not entail (10) and, unlike (10), it is perfectly compatible with a risky decision to trust someone's testimony playing a role in the formation of propositional attitudes associated with faith. With respect to acceptance, there will be circumstances in which one could reasonably accept-that x exists without the accompanying belief-that x exists. While steering a ship, if an unreliable crew member, also known for his practical jokes, cries out ‘Captain, there is a large reef to our starboard, pull hard to port immediately!’ I might well accept that there is a reef to the starboard as a basis for action, even if I don't believe this to be the case.
Christianity, at least, is centred on the life and teachings of an historical figure who is alleged to serve as a mediator between God and humans, in whom and through whom the person of God is revealed. One is invited to trust Jesus of Nazareth and the proclamation of his followers (whose own existence in history are not in serious doubt), a proclamation that includes as part of its content the proposition that God exists. In this way, a decision to trust-in the person of Jesus or the testimony of his followers can serve as the basis for a trust-that God exists rather than the other way around. Clearly this sort of trust can take the form of a risky decision, which can be acknowledged as such. After all, Jesus' followers proclaim his resurrection, a proposition for which many hearers have a very low antecedent probability. But given certain values, tolerance for risk, and openness to considering counterevidence carefully and with intellectual integrity, I see no reason to think that robust faith, even in the absence of belief, must violate the usual canons of rationality under a broad range of evidential circumstances.
Consider the plight of one imaginary doubter, Femia. Suppose that Femia finds something deeply evocative in the religious narratives of the Judaeo-Christian tradition: on hearing the tales about the life and teachings of Jesus of Nazareth, she is stirred with admiration for Jesus as an inspiring moral teacher, striking religious figure, and more. Femia finds herself not with belief, but with other sorts of experiences that she regards as at least open to a religious interpretation. These need not be deeply mystical and might include such mundane feelings as awe at the stars above, moral conscientiousness, a sense of gratitude for the gift of life, a need for comfort in life's dark times or for redemption or transformation of a life that has become unworkable, a deep and persistent longing for God or a sense of the emptiness of life without God, and so on. Suppose that Femia falls deeply in love with Jesus thus portrayed and with God so described. She is so touched by the way of life to which it calls us that these become constitutive of her most basic values. Though she regards the whole thing as a long shot, she is able seriously to entertain the possibility that, if God exists, Jesus enjoyed a unique and intimate relationship with God. Were it only a matter of love, there is no question that Femia is willing to follow and to assume whatever risks might be involved.
Why couldn't such considerations motivate and guide Femia's actions: a decision to embark on a religious way of life? Why couldn't they serve as grounds – either reasons or causes – for her so doing even in the midst of doubt? Suppose that, along the way, Femia does experience joy, comfort in life's dark tragedies, suffering, and loss, or finds aspects of her life tangibly transformed. She is persuaded that there is something good and right about the goals of seeking to love and aid the poor and the downtrodden and of seeking justice for the oppressed by non-violent means. At times she experiences the power of forgiveness to restore relationships, feels moved and inspired in worship, finds deep fellowship with others serving the same ends, and is sometimes ashamed that she is not seeking more completely to live a better life. While our doubter, Femia, thinks that some of this is at least open to interpretation as part of what it could mean to live in relation to God, if there is a God, from an epistemic point of view she does not take any of this as good evidence for the truth of the content of her cognitive commitments, since she is all too aware both that all of this is open to plausible alternative naturalistic explanations and that adherents of other religious traditions might have similar motivations for their commitments. We could evaluate Femia's continuation in this form of life at least partially with respect to how well her lived experience fits with what the religious teachings lead us to expect. Provided that these do fit tolerably well, even if in Femia's view the evidence remains ambiguous and open to various interpretations, it could still be entirely appropriate for her to find in such a life non-epistemic reasons for pressing on (see Hick Reference Hick1966; Allen Reference Allen1968).
To be sure, the truth of the religious content matters greatly, even to one who trusts or hopes that God exists. Suffering and the hiddenness of God might weigh heavily on Femia and present genuine challenges to her faith. But it is a mistake to think that faith must be solely or even primarily a response to one's assessment of the evidential situation. Even for a substantive factual question such as whether or not God exists, the locus of difference between those who affirm or deny the proposition need not be a disagreement about subjective probability assignments in light of the evidence. Instead, it can be – and I suspect often is – a matter of values, attitudes, and behaviour. Femia might well say to an atheist:
We agree on the observables. So long as we regard both atheism and theism as live options, what separates us is not disagreement in our assessments of the state of the evidence for and against God's existence nor need we differ in the non-zero subjective probabilities we assign to whether Jesus' resurrection occurred. It is not that I am credulous where Bertrand Russell is sceptical. Rather, I am willing to risk accepting an invitation to trust where he scorns it.
I see nothing in the logic of belief-in and belief-that or in plausible epistemic constraints on responsible intellectual conduct to preclude Femia from having the following combination of attitudes. In agreement with most historical Jesus scholars, Femia believes that Jesus of Nazareth lived, was a religious teacher, and was crucified on a cross. Among his teachings, at least as reported to us by his followers, are particular promises and an invitation to trust. Femia could, believing-that Jesus existed and so on, decide to place her trust-in Jesus in spite of her doubts about the truth of these teachings. On the basis of this trust-in Jesus, Femia could trust that God exists (since that is part of Jesus' teaching). But then, if Femia trusts-that God exists and is as Jesus says, it may be reasonable for Femia to place her trust-in God as well.
Maybe this is dangerous. But, for Femia, this might only add to the attraction. It is not as though she cannot look before she leaps or that, having leapt, she must at any point neglect consideration of counterevidence or cease to think critically about the commitments she has undertaken. Given certain sorts of values or reasons of the heart, Femia could do so authentically and in full recognition of what she is getting herself into, convinced that it matters whether or not God is real while giving full weight to considerations that can lead her to serious doubt. Such a decision is not just arbitrary and, if she takes the state of the evidence to be such that the rather minimal epistemic commitments involved in her trusts or hopes are epistemically justified or rational, she may even be reasonable in doing so. Femia could do this in the absence of belief by deciding to take the risk, acknowledged as such, with eyes wide open.Footnote 6
