Forthcoming in Faith and Philosophy.

Copyright © 2001 Faith and Philosophy

Reprinted on web site with kind permission of Faith and Philosophy.




Alexander R. Pruss

Alexander R. Pruss


Louis Pojman has argued that Christian faith does not entail belief, or even assigning a probability of 1/2 to the claims of Christianity.  However, this conclusion fails in many cases because of its ethical consequences.  A Christian is committed by his faith to acting in accordance with Christian teaching.  However, there are circumstances when it is morally impermissible to act in accordance to beliefs to which one assigns epistemic probability smaller than 1/2, namely when the action is prohibited by ethical claims that one takes to be more probable.  It is argued that in most cases such considerations preclude a person who assigns a probability of less 1/2 from being both committed both to Christianity and to the moral life.  Matters are particularly clear in the paradigmatic faith-action of Abraham’s sacrifice of Isaac: this action would clearly be immoral if Abraham assigned a probability less than 1/2, or in fact any probability not very close to 1, to the existence of God.  A moral version of Pascal’s wager is also discussed.

I. Introduction

In his paper “Faith without Belief?” Louis Pojman has argued that faith does not entail belief.[1]  It is possible to have faith without assigning a high epistemic probability to the claims of Christianity or even to the existence of God.  Instead, what Christian faith requires is the hope that the central doctrines of Christianity are true.  For Professor Pojman, hope is a complex voluntary attitude that involves four essential features: (1) a belief in the possibility of the state of affairs in question obtaining[2], (2) a lack of certainty about it, i.e., the matter having an epistemic probability strictly less than one, and (3) a desire for it which (4) motivates one to run risks for its sake (pp. 161–163).  Moreover, (5) hoping is under our direct voluntary control and (6) is subject to direct moral evaluation: it is possible to have morally unacceptable hopes.[3] 

Professor Pojman expressly states that hoping that p does not require believing that p has more than a 50% chance of being true (p. 165), but in fact is committed to the stronger claim that hoping that p does not require even believing that p has a 50% likelihood of being true.  This is clear both in the case of the ordinary usage of the word “hope” as people often do hope for things they believe to be unlikely and in that of Pojman’s own usage since he explicitly gives the example of a person who believes that Happy Dancer has a ten percent probability of winning the race but who hopes nonetheless, “against belief, against the available evidence” (pp. 166–167).  On the other hand, hope does entail “a subjective probability index greater than 0” (p. 162).

Professor Pojman’s motivating concern is to respond to the claim, traditionally made by many Christians, that certain propositional beliefs are “a necessary condition for eternal salvation” (p. 172).  However, he argues, we have an ethical duty to believe only on evidence[4] and so “it would seem we cannot be judged unrighteous for not believing in these propositions, if we justifiably find the evidence inadequate” (p. 172).  Therefore, he concludes that in fact propositional beliefs cannot be a necessary condition for salvation, and proposes hope as his alternative to propositional belief.

It is worth noting, however, that even if Professor Pojman were right about the duty to believe only on evidence, nonetheless it could still be the case that some propositional beliefs are a necessary condition for salvation, in the sense that God could ensure that everyone who is saved also believes.  For, it is logically possible that God’s plan of salvation should contain the divine decree: “Solid epistemic grounds for belief are to be eventually given to every person who has the kind of hope that Professor Pojman describes,” and given such a decree, if Pojmanian hope were necessary for salvation, so would belief be.[5]  Alternately, Professor Pojman’s difficulty with the idea of that belief is necessary of salvation could be alleviated by a careful formulation of the notion of an “implicit belief,” which might perhaps be had even by someone claiming to be an atheist.

           In this paper, I shall argue that at least for people in fairly common contemporary epistemic predicaments, being both fully a Christian and a moral person does require that one accept Christianity, or at least the existence of God, with an epistemic probability of at least 50%, thereby showing that mere Pojmanian hope, which does not require such an epistemic probability, is insufficient.  This argument does not prove that the requirement holds for everyone, but just that it holds for many and probably most.  Finally, I shall relate the issue to the story of the ‘Aqedah, the Sacrifice of Isaac, which story will suggest that in fact an epistemic probability close to 100% may be required for fully being a Christian.

II. The argument

Being fully a Christian entails having a commitment to do everything that Christianity says is obligatory and to refrain from everything that Christianity says is wrong.  While the claim that one cannot be fully a Christian and still sin would be very controversial, it is uncontroversial that a person who is in a full sense a Christian must be committed to acting in accordance with Christianity.  Faith without at least commitment to works is dead.  Otherwise the requirement of metanoia (“repentance” or “change of heart”) involved in becoming a Christian has no content.  One could put it differently:  By choosing to be a Christian, one commits oneself to acting in accordance with Christianity, and hence if one acts differently, one acts against one’s commitment.

Now, sometimes there is moral conflict between Christian ethics and non-Christian ethics.  By a “moral conflict” between two bodies, A and B, of normative claims, I mean a case where A entails that some action under some physically possible circumstances is impermissible (obligatory) while B entails that that same action under the same circumstances is obligatory (impermissible).  In this usage, if one of the bodies of normative claims merely says that an action is permissible and the other that it is impermissible, there is no conflict: thus, Christianity and Islam do not have a moral conflict over eating pork, since Christianity does not entail a duty to eat pork.

In fact, I claim that for a very large number, perhaps a majority, of reasonable Christians, there is moral conflict between Christian ethics and non-Christian ethics, where by “Christian ethics” I mean the body of those normative beliefs that the reasonable Christian thinks are very likely true if Christianity is true, and by “non-Christian ethics” the body of those normative beliefs that our reasonable Christian thinks are very likely true if Christianity is false.  Some examples will help to support my claim:

1. An epistemic predicament found among a number of Christians is that they believe that if Christianity is true, then it is very likely impermissible to have an abortion under any circumstances, and that if Christianity is false, then there very likely are circumstances in which an abortion is a duty.  With regard to the second part of this claim, the Christian might think that if Christianity is false, then it is very likely that utilitarianism is true, and utilitarianism would say that in a case where both the woman’s and the fetus’s lives are all but certain to be lost through carrying on with the pregnancy, then an abortion is a moral duty even though the fetus’s life will be cut somewhat shorter.

2. More generally, any Christian who believes in the conditional claim “Were Christianity false, utilitarianism would be true” will surely find many examples of moral conflict between Christian ethics and non-Christian ethics.  The classic cases of deliberately killing one innocent person to save many lives come to mind.  According to Christianity, one may not do evil so that good might come of it (Romans 3:8).  According to utilitarianism, such killing can be a duty.

3. Christians believe that it is impermissible to offer sacrifice to Caesar.  Indeed, many early Christians lost their lives rather than to offer such sacrifice.  However, it seems plausible that many if not most Christians will accept that if Christianity is false (and almost certainly if theism is false), then offering such sacrifice is permissible and even obligatory in some circumstances.  For instance, a Christian can reasonably think that if (and only if) Christianity were false, then a person who was offered the option of either offering sacrifice to Caesar or having all of her family tortured to death would have the obligation to offer sacrifice to Caesar.[6]

4. A Christian is likely to believe that if Christianity is true, then she has a duty to go to church on Sunday unless some particularly grave circumstances detain her.  Let S be an illness of one’s spouse, of a gravity sufficiently low that if Christianity is true then a married person still has a duty go to church, leaving the spouse who has S alone at home.  Then, even if S is not very serious, the Christian will reasonably think that if Christianity is false, then such a married person has a prima facie duty to stay at home with the spouse suffering from S.

5. One can presumably find circumstances F of family financial need which are not sufficiently grave to justify working on Sunday if Christianity, which has a prima facie prohibition against working on Sunday, is true, but which are sufficiently grave that they require working on Sunday if Christianity is false.

Many more such examples can be given, and it is likely that an epistemic predicament that leads to one or more such moral conflicts is one that most if not all Christians find themselves in.  I shall henceforth assume that we are talking of a Christian who finds herself in such a predicament.  Thus, she believes that if Christianity is true, then some action φ under circumstances C is obligatory, and that if Christianity is false, then φing under circumstances C is impermissible.  (I am counting refrainings as actions so we do not have to consider the alternative where fing is prohibited if Christianity is true and obligatory if it is false.)  Moreover, I will make the following assumption: The Christian believes that the moral weight which is had by the prohibition against φing under circumstances C if Christianity is false is at least as great as the moral weight which is had by the obligation to φ if Christianity is true.  It is reasonable to think that this condition can be satisfied in a number of examples that can be found for most Christians.  For instance, in Example 5, if F is chosen to be sufficiently grave, but not grave enough to remove the Christian prohibition against working on Sunday, this will be true.  Or, in Example 3, a Christian might believe that if Christianity is false, then protecting one’s family from all being tortured to death is her highest duty, a duty equal in weight to that which Christianity assigns to refraining from idolatry.

How ought a person choose in cases of moral conflict?  Recall that a case of moral conflict is not just when one body of normative claims implies that something is impermissible, and another claims it is permissible.  In such a case, there is no moral conflict because one can just refrain from the action and then one violates neither ethical system.  But a real moral conflict is when one is deciding whether an action is obligatory or impermissible in a case where each choice violates a moral system that is a serious epistemic possibility.  The following principle seems prima facie plausible: In the case of a moral conflict, one ought to follow that body of normative claims which one believes to be more probably true.  This principle, however, as it stands is in need of refinement.  For, it might be that while the normative system that claims that the action is obligatory gives that obligation a very great moral weight (e.g., a moral weight equal to that had by our prohibition against murder), the body of beliefs that claims the action to be impermissible only gives the prohibition a smaller moral weight (e.g., the moral weight equal to that had by our prohibition against stealing one piece of candy from a store).  However, the following Moral Conflict Principle seems difficult to dispute:

(MCP) If after sufficiently thorough investigation (i) one believes that one body, A, of moral claims implies that an action φ is impermissible and another body, B, implies that it is obligatory, and (ii) one believes that the moral weight assigned by A to the impermissibility of φ is at least as great as that which B assigns to the permissibility of φ, while (iii) no other body of moral claims is a serious contender in one’s epistemic predicament and (iv) one takes A to be more probable than B, then (v) one is morally prohibited from φing.

           A bit of explanation is needed of the force of the “morally prohibited” in (v).  There are two kinds of moral obligations.  First, there are objective moral obligations.  These specify what is permissible, obligatory or prohibited in the specified circumstances.  However, there are also subjective moral obligations, which deal with what one, after sufficiently thorough investigation, sincerely thinks is permissible, obligatory or prohibited.  Thus, if after sufficiently thorough investigation one comes to believe that something is immoral, then one has a real, albeit subjective, moral obligation to shun it.  One does wrong if one ignores this real subjective obligation.  (See St. Paul’s the discussion of the “weak” brother in Romans 14.[7])  To put the distinction in historical context, Catholic moral theologians have traditionally used the terms “material” and “formal” for what I’ve been calling “objective” and “subjective” obligations.

           In ideal cases, one knows what one objectively ought to do.  However, we often find ourselves in less than ideal cases.  We sometimes do not think we know what we ought to do, and yet urgently need to decide.  Some cases in medical ethics have a particular power of inducing this kind of perplexity while being coupled with urgency.  In those cases, one has a moral obligation to decide as best one can, and the MCP is one of the principles that one’s decision morally ought to follow.  The force of the “morally prohibited” in the consequent of the MCP is that of a real, albeit subjective, obligation, one that it is wrong to ignore.  More precisely, the MCP is itself an objectively correct moral rule, but the consequent of the conditional in the MCP specifies what one is “subjectively” or “formally” obliged to do in one’s circumstances of ignorance.  The rightful recognition of there being objective moral obligations cannot blind one to the need for moral rules, like the MCP, to be followed when we do not know what the objective moral obligations in a given situation are.

           Now, given the MCP, recall the case of our Christian who believes that if Christianity is true then φing in circumstances C is obligatory, but that such φing is impermissible, with at least as great moral weight, if Christianity is false.  If such a Christian, moreover, assigns a less than 50% epistemic probability to Christianity, then she must think that the negation of Christianity is more likely true, and hence by the MCP she is morally prohibited from φing in circumstances C.  Now, a moral person is committed to acting in accordance with morality.  Hence, our Christian will be committed to refraining from φing in circumstances C.  But since Christianity claims that there is an obligation to φ in circumstances C, it follows that such a person cannot be fully a Christian—unless she is to have incompatible commitments—since we had said that being fully a Christian entails being committed to fulfilling Christianity’s obligations.  Hence, if someone is fully a Christian and committed to morality, and if she finds herself in the above described epistemic predicament with regard to the obligation/impermissibility of φing, then she must (at the pain of having incompatible commitments) believe with at least 50% probability that Christianity is true.  And, as I have argued, most if not all Christians will for some φ and some C find themselves in such a predicament.

           To put the matter simply, a Christian who accepts the judgment in Example 3 that if Christianity is right then one may never offer sacrifice to Caesar but if Christianity is wrong then sometimes one must do so will have to be a person who believes that Christianity is at least 50% likely if she is to avoid the Scylla of a failure to be committed to the MCP and the Charybdis of a failure to be committed to never offering sacrifice to Caesar.  Since most if not all Christians find themselves in such an epistemic position with regard to some issue—many of them in fact with regard to precisely this one—it follows that probably Professor Pojman’s Christian who hopes but does not even believe Christianity to be more likely than not cannot be fully a Christian, i.e., cannot fully commit herself to all the things that Christianity requires her to commit to, unless she takes the immoral route of being committed to break the MCP.

III. The ‘Aqedah

In Fear and Trembling, Kierkegaard’s pseudonymous author Johannes de Silentio talks of the “teleological suspension of the ethical” which happened to Abraham.  De Silentio’s claim was that what Abraham did in attempting to sacrifice Isaac was in fact contrary to ethics, but that ethics was suspended for the sake of its Telos, namely God who commanded Abraham to act to the contrary.  Abraham was placed precisely in the kind of predicament that I describe above.  He could see that if his faith in God and in the associated faculty of recognizing when it is God that speaks were correct, then he would be obliged to try to kill Isaac; but if his faith were wrong, then in doing this he would be acting immorally.  Hence unless he held that his faith and his having the capacity to recognize when God speaks had at least 50% epistemic probability, he acted immorally.  Thus much the MCP gives us.  Moreover, in light of the great seriousness with which the prohibition against killing, and especially against killing one’s children, is endowed, surely the action would be immoral unless the epistemic probability of his faith were extremely high, maybe even 100%.

           Unless one wishes to claim, as de Silentio seems to, that Abraham acted immorally, a claim which would imply that God has led Abraham into sin and which would thus impugn God’s supreme goodness, one needs to say that Abraham must have placed a very high epistemic probability on the proposition that God exists, is to be trusted and his voice can be recognized by Abraham.  (I am assuming that if God commanded Abraham, then it was not wrong of Abraham to do the deed, since God has not given but only lent life to Isaac and has the right to terminate the lease; if something like this assumption is not granted, then Abraham must indeed have done wrong.)  But Abraham’s faith is a paradigm of faith in the New Testament.  And even though Christians believe that God in the time of the New Covenant does not literally require the kind of human sacrifice that Abraham was to offer, nonetheless Christians should have a commitment to acting in Abrahamic ways under Abrahamic circumstances, and some such circumstances, as in Example 3 of Section II, above, are physically possible.  This commitment, as I argued in Section II, is in many instances incompatible with giving a low epistemic probability to the objects of faith.

           On the other hand, there is a rabbinical interpretation of the ‘Aqedah on which Abraham was not commanded to kill Isaac, but simply asked.  This interpretation observes that God said something like “please” (Hebrew: na) when he asked Abraham to kill Isaac (Genesis 22:2).[8]  But if this is so, then in fact Abraham’s moral dilemma is increased.  If he were wrong about there being a God who asks him to kill Isaac, he would be doing a grave moral wrong by killing Isaac.  If he were right, then he would not be doing any moral wrong by failing to kill Isaac, because he was not commanded but merely asked.  But if this is so, then unless the epistemic probability of there being a God who asks him to kill Isaac was fairly close to 100% if not perhaps actually 100%, Abraham did wrong in killing Isaac.  This rabbinical interpretation thus only strengthens the case against Professor Pojman’s account.

IV.  Five objections

What the argument in Section II, if it is sound, has established is that generally the Christian needs to assign least 50% epistemic probability to Christianity.  But what if someone were to modify the Pojmanian position and say that what is required for faith is not just hope, but in fact hope and the assignment of at least a 50% epistemic probability to Christianity?  This would allow for a Christian to sit on the very edge epistemically, considering the evidence for and against Christianity to be exactly balanced.  Call such a person, if she is a possible person, a “50/50 Christian”.

            Note first that if the argument from the ‘Aqedah is sound, this position is ruled out, since I have argued that the ‘Aqedah shows that faith requires a high epistemic probability assignment.  But there is another argument against the proposed alternative:  Being Christian is supposed to be a stable state, whereas holding that Christianity has exactly a 50% probability of being right is a very unstable epistemic perch.  Any new piece of evidence, however small, against Christianity will push one off the perch. 

           For instance, since the claim that Christianity is true entails that there are no gratuitous evils, i.e., evils which God would not be morally justified in permitting, any evidence for the gratuitousness of an evil will at least slightly decrease the epistemic probability of Christianity, as long as this probability was already less than one.  This decrease need not worry a Christian who assigns an epistemic probability significantly greater than 50% to Christianity, but it will be a disaster for the 50/50 Christian, since the 50/50 Christian will then no longer count as having faith on the modified Pojmanian account.  Similarly, any new evidence, however slight, found by a source critical scholar for a later dating of the Gospels will at least slightly decrease the probability of the historical accuracy of the Gospels and therefore decrease the probability of Christianity, unless of course these probabilities are already one.  Thus, one imagines the 50/50 Christian trembling while scanning new issues of scholarly journals, as any new piece of evidence against Christianity, howsoever small, would rob her of her faith. 

           In general, since at least an intelligent 50/50 Christian can be expected to have already factored into her probability assignment most of the major apologetic arguments for theism and for Christianity, any piece of evidence that weakens the force of one of these arguments, or strengthens some argument against Christianity, should decrease the epistemic probability of Christianity for her, thereby pushing her over the edge into lack of faith.  This is not a stable state, whereas Christian faith is supposed to be a stable state since one who has Christian faith is committed to keeping that faith.  Indeed, it is difficult to see how a 50/50 evidentialist Christian could possibly be committed to keeping the faith, since she has the incompatible commitment to lower her epistemic probability below 50%, and by implication to cease having faith, the next time a piece of evidence against Christianity comes in.

           But perhaps there is a way that someone could be a stable 50/50 Christian, namely by thinking that there is no real evidence either against Christianity or for it.[9]  However, this is not a position that a rational person can take.  For instance, it would be unreasonable to deny that the New Testament does provide one with at least some evidence for the resurrection of Jesus, since the conditional probability that a written work from the first century of our era would report a resurrection is clearly greater on the assumption that a resurrection occurred than on the assumption that a resurrection did not occur.  It thus does not appear to be rationally possible to claim that none of the apologetic arguments for Christianity have any epistemic weight, and hence one cannot rationally claim that there is no real evidence either way with respect to Christianity.

           The second objection to my argument of Sections II and III is more difficult to answer.  It might be claimed that Christian moral demands always have higher moral weight than the non-Christian ones, at least assuming that the non-Christian ones are in fact non-theistic, because the Christian body of normative claims makes each immoral act have infinite demerit as a sin against the infinite goodness of God, while the non-Christian body of normative claims can only assign finite demerit to each immoral act.  The idea here is that the magnitude of a moral offense is to be measured by the goodness of that which is offended against, and God has infinite goodness.  If this is so, then the MCP fails to apply in the cases I discuss in Section II and my argument in that Section fails, though the argument from ‘Aqedah on the rabbinical interpretation continues to hold.  Of course, one might retort by questioning the inference from “a sin against the infinite goodness of God” to “infinite demerit”, though there certainly are prominent strands in the Christian doctrine of redemption that would accept it.

           Note that if this objection is correct, then a person for whom the only choice is between Christian and non-theistic ethics, and for whom both are live possibilities, has the obligation to follow Christian ethics by something like a moral version of Pascal’s wager, with rewards/punishments between replaced with moral merit/demerit.  After all, if it is an infinitely great moral offense to omit φing if Christianity is true, and only a finite moral offense to φ if Christianity is false, then one morally ought to φ, as long as the epistemic probability of Christianity being true is greater than zero.

           However, even if this is true, it is likely that many people do not reason this way and hence are not in the corresponding epistemic predicament.  They may simply assign equal moral weight to those obligations which according to Christianity are highest and to those which according to the non-theistic ethics are highest (here I am not considering people who think that if Christianity is false, all is permitted).  The argument above would still apply to the epistemic position of these people, and there will surely be many in this epistemic position.  These people will take it that the evil in worshipping Caesar if Christianity is true is on par with the evil of acting in a way that leads to the painful death of all of one’s family if Christianity is false.

           And for those who do accept the judgment that it is infinitely worse to transgress against Christian moral norms should these be true than against non-Christian ones should those be true, I can apply the moral version of Pascal’s wager, as described above, to show that such a person should believe in God (and the same argument should work for other central doctrines of Christianity).  For, on the non-theistic alternative, unjustified belief in God, while contrary to ethical norms, is not an offense against God, and hence is not an infinite offense.  Now, if Christianity is true, then it is prima facie permissible to believe in God.  And most will admit that there is a non-zero probability that if Christianity is true, then in fact God commands belief in His existence.  Admittedly, Professor Pojman has arguments against this thesis, but the arguments are not apodeictic, while the weight of Christian Tradition and the Bible is against him, thereby bestowing a non-trivial probability on the conditional that belief in God’s existence is obligatory if Christianity is true.

           Hence, we have three epistemic alternatives, on our general assumption that the only two live epistemic options are Christianity and non-theism, as indeed is the case for many.  These are:

(a) the non-theistic alternative is true;

(b) Christianity is true but it is false that God commands us to believe in His existence; and

(c) Christianity is true and God does command us to believe in His existence.

It is assumed that each of these is a live epistemic option.  Now, if (a) is true, then belief in God is only finitely immoral.  If (b) is true, the such belief is prima facie permissible, and since there are no pernicious consequences, we can remove the “prima facie”.  And if (c) is true, then lack of such belief is an infinite offense against God.  Hence, in order to avoid committing an infinitely immoral act, given that (a), (b) and (c) are all live epistemic possibilities, one is morally obliged to go by the morally safer way and believe in God.  Of course this is all predicated on the assumption that violating one’s duties is infinitely worse if Christianity is true than if Christianity is false—but if this assumption fails, then my original argument in Section II against Professor Pojman’s view applies.

           Crucial to this response is my claim that if Christianity is true, then it is not immoral for us to believe in God.  The claim has initial plausibility, but this might simply be caused by the fact that we are apt to confuse it with the truism that if we know Christianity to be true, then it is not immoral for us to believe in God.  But fortunately we can defend the claim.  If Christianity is true, then the purpose of a human being’s life is to love and commune with God.  Such loving communion involves as a crucial component of itself the belief that God exists.  Thus, the belief that God exists is a part of the human good.  But it is prima facie permissible, even meritorious, to perform an act which directly achieves a part of the human good.  Hence, it is not impermissible to act directly to get ourselves to believe that God exists, unless the means are wrong—but if the willing is direct, then there are no means distinct from the act and hence the means cannot be wrong.

           Thus, if one rejects the claim that the non-theistic moral considerations are of infinitely lesser weight, Professor Pojman’s view will be liable to the objections in Sections II and III, while if one accepts the claim, then the argument above shows that if one accepts that there is a positive probability of Christianity being true, then one ought to believe that God exists.  Hence, either way, Professor Pojman’s view fails.

           A third objection might be introduced, and this is the objection that if Christianity is false, then there are no ethical or moral truths.  This would undercut the argument in Sections II and III.  I could reply to this objection in two ways.  First, I could limit my argument to those who believe that there are moral truths even if Christianity is false.  But secondly, I could once again suppose the triple disjunction of (a), (b) and (c), as above.  On that supposition, now, cases (b) and (c) are as before: if (b) is true, then it is not wrong to believe that God exists, and if (c) is true, then it is wrong not to believe it.  But case (a) changes, in that now it is to be said that if (a) is true, then it is not wrong to believe that God exists, since nothing is morally wrong.  But once again the same conclusion follows as above, namely that one should believe that God exists on the off-chance that it is obligatory to do so, given that there is now no live possibility of such belief being immoral. 

           It is important to note that the wager arguments here are moral ones, unlike those of Pascal.  Issues of divine punishment or reward are irrelevant here: doing what is right counts as its own reward and doing what is wrong counts as its own punishment.  If the arguments merely concluded that belief that God exists is justified from the point of view of self-interest, then they would make no headway against an evidentialist who would say that the moral prohibition against believing something evidentially unwarranted trumps self-interest.  However, the arguments conclude that believing in God maximizes the moral worth of the outcome, and thus they do not beg the question against an evidentialist like Professor Pojman who only defends his prohibition of unwarranted belief “as a prima facie moral principle, one which can be overridden by other moral principles, but which has strong presumptive force.”[10]  The possibility of such overriding opens up the way for wager arguments as above—the wager argument cannot be made if the evidentialist imperative is taken to be a rock-bottom exceptionless absolute, but unlike Clifford[11], Pojman does not argue for such a strong and implausible claim.  Note, however, that all these wager arguments are predicated on the assumption, made by someone objecting to my arguments from Section II, that Christian imperatives, if Christianity is true, are of greater weight than non-Christian alternative imperatives.  If this assumption is false, then my arguments from Section II stand, of course.

           An anonymous referee said that one might run the many-Gods objection to the Pascalian Wager in this context as well.  Note, however, that this cannot hold if the wager is confined simply to the choice whether to believe that the God of theism (a perfectly good, all-powerful, all-knowing creator) exists or not to believe that this God exists.  If one attempted to extend the above wager to more specifically Christian beliefs, one would either have to find a way around the many-Gods objection to the Pascalian Wager, perhaps by saying that in a many-Gods situation one should believe the most probable set of doctrines, or one would have to limit oneself to a case where there are only two live epistemic possibilities, e.g., Christianity and atheism.

           Next, there is a fourth, and perhaps strongest, reply which Professor Pojman has himself made in response to a variant of my argument in Section II.[12]  The moral law is reasonable, in the way that natural law theorists have it.  There is certainly evidence in the Bible and in Tradition that Christianity is committed to a claim that there is a law written even in the hearts of non-Christians, a moral law one can know by reason.[13]  Therefore, Pojman argues, there will never be a moral conflict between Christian ethics and rational non-Christian ethics, and so my argument cannot get started.

           However, as an empirical matter of fact, it is easy for human beings to ignore their conscience, to silence it with unconscionable deeds, and to distort it over time.  Even apart from this, our consciences are often apt to err, as can be seen by the disagreement that seemingly rational and intelligent persons have over various moral issues.  Consequently, many Christians (perhaps not including Professor Pojman—or the author of this paper, for that matter) are in the epistemic predicament that some of their moral beliefs rest solely on Christian faith, with these Christians realizing that they would have different moral beliefs if they thought Christianity false.

           But there is a better answer to Professor Pojman’s objection.  Recall three of the examples in Sections II and III: sacrificing to Caesar, missing church on Sunday, and the ‘Aqedah.  In all of these cases, to have epistemic reason to think that the Christian moral view of the matter is right requires that one have epistemic reason to think that God exists.  After all, the wrong in sacrificing to Caesar constitutes precisely in robbing God of the glory due to him alone—and if there were no God, then quite possibly Caesar would be the supreme personal power in the universe.  Going to church on Sunday is a prima facie obligation only if there is a God to be worshipped on Sunday.  And, the permissibility of Abraham sacrificing Isaac comes precisely from the fact that the command or request is given by the God who lent existence to Isaac.  If one does not have epistemic reason for believing in God’s existence, one likewise does not have epistemic reason for thinking that the Christian view of these matters is the right one.  Thus, Professor Pojman’s answer applies in these cases only if one has epistemic reason to believe in God’s existence.  But if one has such epistemic reason, then Professor Pojman’s story of hope instead of belief is moot at least in the case of belief in the existence of God—for given such epistemic reason, everyone would agree that then one could just believe.

           Finally, an anonymous referee has made the objection that my arguments at most establish that one cannot be an ideal Christian without believing Christianity to be at least 50% probable.  After all, the referee notes, all the argument argues for is a conflict between one’s commitments if one is committed to Christianity and yet does not believe Christianity to be 50% probable, and conflicts between commitments are commonplace. 

           Two things can be said in response.  First of all, the conflict here is a particularly insidious one because it is a conflict between Christian commitment on the one hand and the moral commitments of the MCP on the other.  For instance, it is questionable whether one is subjectively morally permitted to become a Christian, or even continue to be one, if one thinks that this would involve one in a conflict with morality. 

           Secondly, although to be a Christian and to be an ideal Christian are not the same, to be a Christian involves being committed to being a morally perfect Christian.  For instance, when an adult becomes a Christian in a liturgically oriented denomination, she makes various baptismal vows, such as to reject all of Satan’s empty promises.  Making these vows commits her to following the totality of Christian morality.  A less liturgically oriented Christian might say that the Christian is one that says that Jesus is Lord.  This acknowledgment of Jesus as Lord cannot be a purely intellectual affirmation that Jesus, de facto, is a Lord—it is also a performative act that commits one to living as if Jesus was “Lord” in the sense of “the Lord God.”  Because of the supremacy of God over everything, this commitment is implicitly a commitment to put aside all commitments incompatible with it (cf. Matthew 5:30 on cutting off the hand that causes one to sin).  Thus, unless one is committed to follow all that the Christian God has commanded, and to put away all commitments incompatible with this, one has not fully become a Christian, since to fully become a Christian involves undertaking this commitment.  I am not making the strong claim here that failure to follow a Christian command involves a failure to fully be a Christian, but the weaker one that a failure to be committed to following a Christian command involves a failure to fully be a Christian.[14][15]

Georgetown University


[1]. Faith and Philosophy 3 (1986), pp. 157–176.  See also Pojman’s more recent “Faith and Doubt or Does Faith Entail Belief?”, in Richard M. Gale and Alexander R. Pruss, eds., The Existence of God, (Aldershot, England: Ashgate Publishing, forthcoming).

[2]. Not to trivialize Pojman’s definition, I shall assume that the “possibility” in question here is only epistemic and not logical possibility.  For, it is clearly Pojman’s point that to be a Christian one does not need to believe such doctrines as the existence of God.  But, arguably, traditional Christian thought makes the existence of God logically necessary since God is held to be a maximally great being (in the terminology of Alvin Plantinga’s ontological argument in God, Freedom and Evil [Harper & Row, 1986]).  But if it is logically possible that God’s existence is logically necessary, then God’s existence is logically necessary by S5.  Hence, there is a simple entailment from the logical possibility of traditional Christianity to the actual truth of the doctrine of God’s existence.  Likewise, since traditional Christianity conceives of God’s trinitarian nature as an essential property of God, it follows that if traditional Christianity is logically possible and if God exists, then God is a Trinity, since if a being has an essential property in one world, it has it in all the worlds in which it exists.  Therefore, at least for the philosopher who grasps this entailment, it is impossible to believe in the logical possibility of traditional Christianity without believing in the actual truth of some of its central doctrines.

[3]. “Faith and Doubt or Does Faith Entail Belief?”, Section IV.

[4]. On page 172, this is stated in merely conditional form, but Pojman also affirms it unconditionally in “Faith and Doubt or Does Faith Entail Belief?”

[5]. Both “necessities” are presumably hypothetical, of course;  i.e., generally when Christians affirm that “faith is necessary for salvation”, they merely mean that God made a certain decree D such that the conditional if God decreed D, then only those who have faith are saved is logically necessary, and they do not intend to claim that the consequent of this conditional is logically necessary.  (Of course it may be that someone will go beyond the weaker claim and demonstrate that the concept of salvation is such that it is logically necessary that only those who have faith are saved.  But in that case, if Pojmanian evidentialism is true, then there is still an epistemic possibility that God’s essential goodness would make it logically necessary that anyone who has Pojmanian hope also has sufficeint evidence.)

[6]. Of course, if the Christian had been, say, a Jewish convert, then he might think that if Christianity is false, Judaism is still true, and hence the sacrifice is impermissible.  But if the Christian is one for whom Christianity is the only plausible monotheistic option—and there surely have been many such Christians—then the description as given will apply.

[7]. The weak brother believes that some things that objectively are morally permissible, such as eating pork or meat sacrificed to idols, are morally prohibited.  St. Paul warns that one should not tempt the weak brother into doing something the weak brother thinks is immoral, because this could bring about the weak brother’s ruin: “he who has doubts [this weak brother] is condemned, if he eats, because he does not act from faith;  for whatever does not proceed from faith is sin” (Romans 14:23, Revised Standard Version).  I take St. Paul to be arguing that because the weak brother believes that it is contrary to faith to eat some meats, therefore his action of eating the meats is contrary to faith, and hence is a sin.  Exactly the same reasoning applies in the case of something that is a matter of conscience rather than of faith, strictly speaking.

[8]. See Nahum M. Sarna, Genesis: The Traditional Hebrew Text with the New JPS Translation, The JPS Torah Commentary (Philadelphia / New York / Jerusalem: Jewish Publication Society, 1989), ad loc.

[9]. I am grateful to an anonymous reader for this suggestion.

[10]. “Faith and Doubt or Does Faith Entail Belief?”, Section III.

[11]. “The Ethics of Belief”, reprinted in: Michael Peterson, William Hasker, Bruce Reichenbach and David Basinger, Philosophy of Religion: Selected Readings, second edition (Oxford / New York: Oxford University Press, 2001), pp. 80–85.

[12]. The variant is forthcoming in the introduction to Richard M. Gale and Alexander R. Pruss, eds., The Existence of God, (Aldershot, England: Ashgate, forthcoming).the Introduction to Richard M. Gale and Alexander R. Pruss (eds.), The Existence of God, Aldershot, England: Ashgate Publishing (forthcoming).

[13]. See, e.g., Romans chapter 1 and 2:15.  Note that “the heart” in Hebrew thought is not just the seat of emotions but also the seat of will and thought. 

[14]. I am grateful to Professors Richard Gale and Louis Pojman for interesting discussions of these issues, and to two anonymous readers for a number of insightful comments that have contributed to significant improvement of this paper. 

[15] I am grateful to discussions with Richard M. Gale concerning this paper and to an e-mail exchange with Professor Pojman.