Though Christianity has never treated belief (understood in its standard philosophical sense) as sufficient for salvation, the temptation to think that such belief is necessary for salvation is quite powerful. In the Christian tradition, it is rather natural to assume that a person can be saved only if she believes that various relevant propositions are true.Footnote 1 Louis Pojman has, however, attacked this necessity claim with some force.Footnote 2 As one sympathetic to the conclusion that Pojman reaches, my project here has two elements. First, I will rehearse Pojman's argument and raise a pair of concerns with it. Second, I will construct a parallel argument to the same conclusion that avoids these concerns.
Before turning to the arguments, however, we should note that there is a common way of objecting to the necessity claim; a counter-example strategy. Many Christians will, after all, appeal to infants or young children who die before the ‘age of accountability’ as counter-examples to the claim that belief that the salvation-relevant propositions are true is necessary for salvation. Similarly, many Christians will appeal to the infamous tribesman who has never had the opportunity to hear the Christian teaching. I am inclined to join these counter-example strategists in thinking that these sorts of cases can show the necessity claim to be false. But, following Pojman, I will be arguing for something more than the falsity of the general necessity claim. I will be arguing that it can be false even for mature adult persons with access to the relevant claims. Let's call the claim that belief in the salvation-relevant propositions is necessary for salvation, at least for such people (NC*). The denial of (NC*) is clearly more controversial than the denial merely of the general necessity claim. The argument I develop to demonstrate that even (NC*) is false will, as it turns out, apply to the counterexample cases. Indeed, I hope it provides something of an explanation for our intuitions in these cases. But, like Pojman's, my argument will also extend beyond these cases to include a more surprising and less orthodox class of people.
The ethics argument
Pojman's defence of the claim that (NC*) is false emphasizes the ethics of belief. His argument can be reconstructed in the following way:Footnote 3
(1) God is perfectly good.
(2) If God expects a person to violate a moral obligation in order to achieve salvation, then God is not perfectly good.
(3) Therefore, God does not expect a person to violate a moral obligation in order to achieve salvation.[From (1) and (2)]
Further, Pojman's argument appears to depend on a fairly robust Cliffordian principle such that:
(4) It is always morally wrong to believe a proposition on insufficient evidence.Footnote 4
From here, the argument follows intuitively.
(5) For some people, the salvation-relevant propositions appear to be insufficiently supported by the available evidence.
(6) Therefore, it would be morally wrong for such persons to believe these propositions (even if they could bring themselves to do so). [From (3), (4), and (5)]
(7) If God expected such persons to believe these propositions, then God would be expecting them to violate a moral obligation in order to achieve salvation.[From (6)]
(8) Therefore, God would not expect such persons to believe these propositions in order to achieve salvation.[From (3) and (7)]
From this argument, we can conclude that belief that particular salvation-relevant propositions are true is not necessary for salvation. There are some people so situated with respect to the evidence and so constituted with respect to the ethics of belief that God simply cannot blame them for their doxastic hesitance. Indeed, Pojman goes so far as to suggest that God might uniquely honour the commitment and consistency of such hesitant non-believers. He notes colourfully that, ‘[o]n this basis it might well be the case that in heaven (or purgatory) Calvin, Barth, Billy Graham and Jerry Falwell may have to be rehabilitated by taking catechism lessons in the ethics of belief from such archangels as David Hume and Bertrand Russell’.Footnote 5 This quote brings out crisply the dependence of Pojman's argument on the sort of demanding ethics of belief I have attributed to him in premise (4) above. And for this reason, I refer to Pojman's argument as ‘the ethics argument’.
Again, let me be perfectly clear about the fact that I am in agreement with Pojman's conclusion that the kind of faith required for salvation in the Christian tradition need not entail propositional belief in the standard philosophical sense. My concern is with certain limitations of the ethics argument itself. Furthermore, I do not deny the validity of the ethics argument. Rather, I believe that the argument depends unnecessarily on a controversial premise and that its conclusion (though true) is unnecessarily narrow. Both of these concerns will ultimately motivate the exploration of another argument to the same principle conclusion.
The first concern, then, with the ethics argument is that it relies on the controversial claim that there is always something morally wrong with believing a proposition on the basis of insufficient evidence. Getting all my cards on the table, I simply do not believe that this is true. By my lights, then, premise (4) above is false. I do not have the space (or ability, or inclination) to treat in detail the nuanced debates about the ethics of belief and the viability of evidentialism. Nor do I believe that such a treatment would yield the result that all reasonable people would agree with me in rejecting premise (4). But many philosophers are already prepared to join me in rejecting a stringent evidentialism. Even those philosophers with the greatest confidence in evidentialist intuitions are bound to grant that their interlocutors are not completely irrational in their resistance. The disputable nature of premise (4), then, gives us a reason to welcome an argument against (NC*) that does not depend on it.
The second concern with the ethics argument concerns its scope. Notice that the conclusion of the argument is that a certain kind of person cannot properly be blamed by God for failing to believe the salvation-relevant propositions. Strictly speaking, it is only those epistemically ethical persons who are also in the position of finding the available evidence for the salvation-relevant propositions insufficient who demonstrate that belief is not necessary for salvation. On Pojman's argument, then, it seems doubt can be compatible with faith only if it is rooted in evidential failure.
My concern is that, given this implication, the ethics argument proves too little. I believe that there can be salvation without belief and without evidential failure. This is to say that a person might fail to believe the salvation-relevant propositions even while recognizing that belief is sufficiently supported by the evidence. For example, I can imagine someone saying the following honestly: ‘I think there is enough evidence to believe – I mean, if I encountered someone in precisely my same evidential circumstances with respect to the truth of the salvation-relevant propositions, I would think her perfectly reasonable in believing. It's just that I don't find myself believing the propositions – and I can't simply make myself believe them by force of will.’ This imaginary appeal prefigures my later argument. But for now, I hope it illustrates the limitations of Pojman's argument. If we reach the conclusion that belief is not necessary for salvation on evidentialist grounds, then it appears that only evidential failure can ground faith without belief. By contrast, it seems to me that there are ways of failing to believe that are consistent with faith even though they do not involve evidential failure.
So, consider Thomas who does not trust the police. He has been raised in circumstances in which trusting the police has been unjustified. Perhaps he had the misfortune of encountering an inordinate number of ‘dirty cops’. Now, however, he is confronted with a police officer attempting to save him from a dangerous situation. The officer announces that Thomas needs to jump down from a precarious position so that the officer can catch him. Furthermore, Thomas has seen this officer behave in ways that powerfully suggest he is reliable. Because of Thomas's upbringing, he does not, indeed cannot, believe the officer is trustworthy (this does not necessarily mean he believes the officer is not worthy of trust. He might, after all, withhold all belief about the officer's trustworthiness). But Thomas's failure to believe is not grounded in evidential failure. The officer's prior behaviour was, we can imagine, sufficient to support Thomas's belief in his trustworthiness. Thomas, we can further suppose, is willing to grant this point. Nevertheless, the force of his early childhood circumstances prevents his forming the belief in the officer's trustworthiness (and, therefore, in the proposition that the officer will catch him when he jumps). Still, let us assume, Thomas jumps. In this case, it seems reasonable to me to say that Thomas has put his faith in the officer even though he didn't believe the officer was trustworthy, and even though his doubt about this was not rooted in evidential failure.
In light of the controversial nature of Pojman's evidentialist premise and the limited scope of his conclusion, I would like to propose an alternative argument against (NC*).
The involuntariness argument
This argument starts with the recognition that, in the Christian tradition, failing to receive salvation is not like failing to receive any other personal goods. It is not like, for example, failing to win the lottery or failing to receive an academy award. This is because in failing to receive salvation one is thereby under condemnation (i.e. from God). Nothing like condemnation is properly associated with being a failed lottery winner or a failed Oscar nominee. Furthermore, condemnation has its own distinctive internal logic. In particular, it entails moral blameworthiness.Footnote 6 When one is condemned, one is blamed. If a person cannot properly be blamed for what she has done, then this fact makes condemnation inappropriate.
Like condemnation, blameworthiness has its conditions. One crucial condition is that a person can be blamed only if she has violated a moral obligation. Keep in mind that it is moral blame with which I am concerned. We can, in some sense, blame the Cubs' [team's] woes on injuries, but this will not be moral blame. And this is precisely because the Cubs' [team's] players who get injured violate no moral obligations in getting injured (though, to some Cubs' [of the team's] fans, it may appear almost immoral). If a person can demonstrate that her action was within the bounds of duty, then fairness demands that we withdraw our attitudes of blame. These considerations lead to the conclusion that:
(1) A person can be fairly condemned for doing something only if in doing it she has violated a moral obligation.
The second component of my argument is the deeply entrenched moral intuition that a person can be obligated to perform an action only if the person is able to perform the action. The traditional aphoristic formulation of this putatively Kantian principle is that ‘ought’ implies ‘can’ (OIC for ease of expression). The plausibility of this principle depends, in part, on what is meant by ‘can’. I suggest the following meaning. A person ‘can’ do something, in the morally relevant sense, just in case either he has it in his power to do it at the time of action, or he would have had it in his power at the time of action save for some prior negligence on his part. On this definition, the drunk driver who kills the innocent pedestrian still could have avoided killing the innocent pedestrian (he had the relevant ability), since the unavoidability of the bad outcome is due to the driver's earlier negligence. We can suppose that had the driver not chosen to get drunk he would have been able to avoid killing the pedestrian.
These reflections on OIC and the nature of moral ability license the following premise:
(2) If a person cannot do something, then she does not have a moral obligation to do it.
(2) is simply the contrapositive of OIC. And since, as we've already seen, there can be no blame without obligation, it follows that,
(3) If a person cannot do something, then it would be unjust to blame her for failing to do it.[From (1) and (2)]
Furthermore,
(4) God is perfectly good.
Thus, given the relationship between condemnation and blame developed earlier,
(5) If a person cannot do something, then God will not condemn her for failing to do it.[From (3) and (4)]
Notice, now, that if we accept even a fairly uncontroversial form of doxastic involuntarism, then it follows that those who do not believe the (putatively) salvation-relevant propositions do not have the power (here and now) to believe them. Such people cannot simply will to believe them, even if they could will to withhold belief about these matters (as one version of doxastic voluntarism would suggest). Further, by my lights, there is good reason to accept that among those who do not believe these propositions, there are some who have not come to be in this state of unbelief through negligence. To deny this would be to insist that all unbelief is due to some dereliction of doxastic duty. I can find no grounds for this insistence (though there surely is much unbelief with this causal history). After all, there are legitimate controversies about the sufficiency of the evidences for various of the propositions in question, and many non-believers will simply not have had the grounding experiences or been exposed to the very best evidential arguments.
This means that we have good reason to believe that there are people who do not have the power to believe the (putatively) salvation-relevant propositions, and who do not lack this power because of prior actions or wilful omissions they ought to have avoided. This is to say that we have good reason to believe that:
(6) There can be people who cannot believe the salvation-relevant propositions.
(7) There can be people whom God will not condemn even though they fail to believe the salvation-relevant propositions.[From (5) and (6)]
It is a common view in Christian thought that those who are not condemned are saved. Ultimately, there are only two groups of people; those who are saved and those who are condemned. Thus, we get:
(8) There can be people whom God will save even though they fail to believe the salvation-relevant propositions.
This is to say that (NC*) is false. Even some mature adults who are appropriately aware of the propositions associated with salvation in the Christian tradition can be saved without believing those propositions.
Conclusion
Of course, there are various points at which to take issue with the involuntariness argument. One obvious target of critique will be the principal premise that ‘ought’ implies ‘can’. I have not attempted to defend this premise, nor will I do so in any substantial sense. This is not to say that I believe it to be beyond reproach. Rather, I am confident that the intuitions supporting it are quite deeply entrenched in our ordinary moral thinking. OIC is, at the very least, less controversial to many philosophers and theologians than Pojman's evidentialist premise.
One reason to resist the premise is distinctively theological.Footnote 7 That is, there are traditions within the Christian community that are almost explicit about their rejection of OIC. I have in mind here stringent versions of reformed thought that emphasize a form of providence that entails strict predestination. Philosophers and theologians with such sympathies will no doubt reject the involuntariness argument on these grounds. I can see no way to dissuade them short of challenging their larger theological commitments. I have no interest in taking up that project. Suffice it to say, then, that the rejection of OIC strikes me as too jagged a pill to swallow.
In any event this means that, my aspirations notwithstanding, the involuntariness argument is not, alas, theologically neutral. I find the premises deeply plausible on broadly philosophical and theological grounds. And I do not think that I am at all alone in doing so. Furthermore, I imagine that many of those who share the theological and moral intuitions to which the involuntariness argument appeals will be somewhat surprised by its conclusion. After all, the claim that belief in some set of salvation-relevant propositions is necessary for salvation, at least for informed adults, is widely thought to be true. What follows from the falsity of the necessity claim (for, say, Christian apologetics and evangelism) will have to be the subject of another project.Footnote 8
