A moral argument against Christian faith

Before spelling out the argument against religious faith that Zamulinski attempts to dispose of, let me make one preliminary remark concerning ‘the ethics of belief’. Zamulinski uses this notion in a somewhat unusual way. In most writings on this topic, this term denotes either a complex set of issues concerning our doxastic responsibility or the philosophical debate about these issues. In Zamulinski's article, however, the term denotes a specific thesis, viz. roughly, the view that one has an obligation to proportion one's belief to one's evidence. Also, the ethics of belief is usually understood either as a discussion concerning the epistemic norms or one concerning our all-things-considered duties that ought to regulate our doxastic lives, but Zamulinski takes the ethics of belief to be a specifically moral thesis. He does so, because he believes that this is what William Clifford, who supposedly was the first to use the term, had in mind when he titled his essay ‘The ethics of belief’. I will not question Zamulinski on this point, but simply employ his use of the term.

In his response to the moral argument against religious belief, Zamulinski focuses on Christianity rather than other religions.Footnote ² I will adopt the same policy, although I believe something similar is true for at least some non-Christian religions. I think the argument that Zamulinski tries to rebut should be reconstructed as follows:

1. (1) Every human being ought to proportion her beliefs to her evidence. [premise]

2. (2) No human being has sufficient evidence for Christian beliefs. [premise]

3. (3) No human being should have Christian beliefs. [from (1) and (2)]

4. (4) One can have Christian faith in God only if one holds Christian beliefs. [premise]

5. (5) No human being should have Christian faith in God. [from (3) and (4)]

Let us dub this argument the moral argument against Christian faith (MA, for short). Let us call (1), which is just another wording of the ethics of belief, the moral premise, (2) the evidential premise, and (4) the doxastic premise. Notice that the terms ‘ought to’ and ‘should’ in (1), (3), and (5) are all to be understood morally rather than epistemically or otherwise.

It seems, however, that there is an ambiguity in MA. For what does it mean to have sufficient evidence for some belief? On the one hand, one might think that some person's evidence is sufficient for some belief it is meets certain objective standards, standards such as logical entailment or standing in a certain probabilistic relationship. On the other hand, one might consider evidence as sufficient if it meets some person's own subjective evidential standards, standards that, at least partly, differ from person to person. This distinction is important, because it affects our interpretation of (5). In the first case, (5) merely says it is in some sense of the word a morally bad state of affairs if people have Christian faith in God. It does not express the idea that people who have Christian faith are to be held accountable for that. It would do so only if one were to add an argument to the effect that Christians are accountable for having certain false subjective evidential standards. In the second case, (5) says that Christian beliefs do not meet the evidential standards Christians themselves have. On this understanding, presumably the Christian is to be held accountable for his religious beliefs and he is a proper subject of praise and blame for having the belief he has.

Although Zamulinski fails to explain what he means by ‘having sufficient evidence’, he seems to interpret (5) subjectively, that is, as implying evaluation in terms of praise and blame. First, he uses terms such as ‘doxastically responsible’ (337) and ‘beyond moral reproach’ (340). And second, he offers a fairly extensive discussion of excuse conditions for doxastic moral responsibility. I will return to this ambiguity in our discussion below.

We should notice that Zamulinski's response to MA is merely conditional: if the moral and evidential premises are true, then Christian faith can still be morally proper, given the falsehood of the doxastic premise. He adds, however, that he will say enough to show that a Cliffordian ethics of belief is plausible (335) and he has elaborately defended a Cliffordian version of the ethics of belief elsewhere.Footnote ³ In what follows I will argue that, contrary to what Zamulinski claims, we should not reject the doxastic premise. Instead, I suggest two other ways to meet MA, by offering a critique of Zamulinski's Cliffordian ethics of belief and by commenting on the implications of doxastic involuntarism and the possibility of doxastic excuse conditions.

The doxastic premise

According to Zamulinski, Christian faith need not be a matter of belief: one can very well be a Christian without having any religious beliefs about God or the world whatsoever. One can choose to adopt as fundamental or existential assumptions on the basis of which to build one's life the traditional propositions of Christianity, such as the propositions that there is a God, that Jesus Christ incarnated to deliver us from evil, and so on.Footnote ⁴ One can voluntarily commit oneself to these assumptions, even in the face of strong evidence to the contrary.

We should notice two things about this alternative to Christian faith as belief. First, it differs from mere acceptance in that it is an existential commitment that orients and informs one's entire life. Second, one can only commit oneself to certain assumptions if one has some purpose in doing so. According to Zamulinski, a Christian can commit herself to certain religious assumptions about God and the world in the hope of salvation. Unfortunately, he is not very clear about what this hope amounts to, but here is what he has to say about it:

Here, the idea seems to be that the Christian has insufficient evidence to believe the central Christian claims about God and the world, but realizes that, if Christianity is true after all, one can only be saved if one has acted on the basis of those claims, that is, if one has lived as if one believed them to be true. Therefore, the Christian can properly take as fundamental assumptions on the basis of which to live her life the central Christian dogmas without actually believing them. Given that no harm seems to follow from it, there is nothing morally reprehensible about committing oneself to certain Christian assumptions.

Of course, Zamulinski is not the first to offer a non-doxastic account of religious faith. We find similar views in the writings of Immanuel Kant and John Stuart Mill and several contemporary philosophers.Footnote ⁵ Remarkably, any discussion of these is absent from Zamulinski's article. In what follows I will confine myself to Zamulinski's discussion of the doxastic premise, although I believe that some of my criticisms apply to other non-doxastic interpretations of Christian faith as well.

Zamulinski's non-doxastic construal of Christian faith seems untenable to me for at least four reasons. First, assuming that p and, hence, acting as if p over a fairly long period of time will seriously raise the chances of one's coming to (over)believe that p. This is why Blaise Pascal famously urged his readers not to believe that Christianity is true, but rather to act as if it is true by reading one's Bible, attending mass on Sunday, etc. However, if this is true and if (over)believing that p violates a moral obligation, then is one not morally required not to act as if p? Zamulinski may be right that the probability that visualization and incantation lead to belief in God might be even greater (342), but this does not alter the fact that acting as if p will significantly raise one's chances of coming to believe that p and the fact, if it is a fact, that this is morally reprehensible.

Second, on spelling out Zamulinski's account of Christian faith as assuming that God exists, it turns out to be circular: (a) assuming God to exist entails hope of salvation, (b) hope of salvation requires trust in God, (c) one can only trust in God if one assumes God to exist, (d) assuming God to exist entails hope of salvation, etc. As Zamulinski correctly remarks, one can only commit oneself to the assumption that p if one has some purpose in doing so, which, in the case of Christian faith, he claims, is being saved. That is, the hope of being saved is the reason for assuming that God exists. But if that reason itself assumes that God exists, it cannot function as a reason. The account is not circular if we rephrase (c) as follows: (c*) one can only trust in God if one believes God to exist. Obviously, however, that is exactly the position that Zamulinski attempts to avoid.

Third, contrary to what Zamulinski contends (341), the propositional attitude of faith, even if understood as commitment to fundamental assumptions, is evidence-dependent to at least some extent. There is an enormous variety of mutually exclusive religious claims telling me quite different things as to what I should do in order to be saved, to live happily on earth, to have a peaceful (non-)existence in afterlife, etc. That means that I can only reasonably commit myself to certain Christian assumptions in the hope of salvation if I believe that there is a fair chance that the Christian God exists, that He saves certain human beings, that I will be one of them, and so on. If I did not have these beliefs, it would be utterly irrational to commit myself to certain fundamental Christian assumptions about God and the world. However, what else would these beliefs be but religious beliefs about God and the world? Thus, assuming in the hope of salvation seems to require at least some religious beliefs.

Fourth and finally, elsewhere, in defending the evidential premise – a premise that I do not discuss in this paper – Zamulinski argues that religion can exist only if its adherents believe the relevant propositions to be true.Footnote ⁶ But that just flatly contradicts his account of Christian faith in terms of assuming rather than believing in the hope of salvation. Whatever the latter may be, it could not count as a religion on Zamulinski's own terms.

Zamulinski also offers an argument against the view that Christian faith involves religious beliefs. It runs as follows:

Now, it is not exactly clear, at least not to me, how we should understand this argument, but I think the thrust of it is something along the following lines:

1. (6) If Christian faith involves religious beliefs, then Christians damage their standards of evidence. [evidential premise]

2. (7) If Christians damage their standards of evidence, they significantly enhance the chances of not objectively loving their neighbours. [premise]

3. (8) If Christian faith involves religious beliefs, Christians significantly raise the chances of not objectively loving their neighbours. [from (6) and (7)]

4. (9) Being a Christian entails the attempt to love one's neighbours objectively. [premise]

5. (10) If one attempts to love one's neighbours objectively, one will attempt not to raise the chances of one's not objectively loving one's neighbours. [premise]

6. (11) Christians will attempt not to raise the chances of their not objectively loving their neighbours. [from (9) and (10)]

7. (12) Christians should not have religious beliefs. [from (8) and (11)]

Zamulinski does not state (9) explicitly, but I do not see how one can plausibly arrive at the conclusion without assuming something like (9). Regrettably, Zamulinski fails to explain the crucial notion of ‘loving one's neighbour objectively’. Prima facie, one would think that if anything is subjective, love is. The only remotely plausible interpretation I can think of is that some person x objectively loves some other person y iff x acts to the benefit of y.Footnote ⁷

Now, there are at least two problems with this argument. First, I do not see why we should accept (7). It does not follow from my evidential standards' being weakened that the chances of my not objectively loving my neighbour will be raised significantly. For if I can only hold religious beliefs by weakening my evidential standards, I might quite well weaken only my evidential standards concerning the existence of immaterial concrete entities, such as God and angels, and the reliability of documents that are claimed to be divine revelations. I do not see how weakening these standards will somehow influence my objectively loving my neighbour.

Second, (12) does not follow from (8) and (11). If one attempts not to do x, and x is entailed by y, it does not follow that one should attempt not to do y.Footnote ⁸ Normally, people should avoid pain, but going to a dentist might entail having a painful experience. It does not follow that one should never go to a dentist. And this is because the benefits of going to a dentist might outweigh its disadvantages. By driving a car one enhances the chances of injuring another person. It does not follow that we should not drive cars. Thus, the question is whether the advantages of having religious beliefs might outweigh the disadvantages of raising the chances of not loving one's neighbour objectively. It is not clear to me that this is the case, nor that it is not. It all depends on how probable it is that by weakening one's evidential standards one will objectively hurt (not love) one's neighbour, what the possible harm to one's neighbour would consist of, what the esteemed benefits of religion are, and how probable it is that one can acquire certain goods by having religious beliefs.

Interestingly enough, I believe the same idea – the idea that Christians are commanded to love both God and their neighbour – can be used to demonstrate that Christian faith in God entails belief that there is such a person as God. For, how can you love God if you do not believe that He exists? It does not seem to make sense that we can love a person that we merely assume to exist. The same applies to all sorts of other emotional attitudes. How can I feel thankful toward God if I do not believe that there is such a person as God? Could I love my neighbour or feel thankful toward my neighbour if I suspended judgement on whether I have any neighbours and merely commit myself to the assumption that I have them? This seems simply impossible.

The moral premise

Let me now suggest two different ways of meeting MA, starting with a discussion of the moral premise. Zamulinski interprets the Cliffordian thesis that one ought to proportion one's belief to one's evidence as a second-order consequentialist thesis and as an analogue of rule utilitarianism. His argument in favour of the moral premise runs roughly as follows. False beliefs are more likely to result in harm than true beliefs. Therefore, we should avoid having false beliefs. Not to overbelieve is one good way to avoid having false beliefs. Thus, we should try to maintain the best evidential standards and to believe a proposition only if it meets our evidential standards.

Of course, some overbeliefs are harmless or even good. However, we do not have voluntary control over our beliefs, so we cannot decide to have only those beliefs that we think are harmless or even good. Thus, we can only believe some proposition either if it meets our evidential standards or if we weaken our evidential standards in such a way that having the belief does not violate them. If one's evidential standards are weakened, however, there is a twofold danger. First, we impair our ability to judge whether certain beliefs are overbeliefs. And second, we impair our capacity to form an accurate opinion on whether some overbelief is potentially harmful. Therefore, we should accept either a policy of never overbelieving or a policy of doxastic amorality on which one is allowed to believe whatever one wants to. Overall, the former policy seems to have morally better consequences than the latter. Therefore, we should adhere to Clifford's ethics of belief.

I think that there are several problems with this argument, but let me point out just one of them. At several junctures in his article, Zamulinski intimates that one can have some belief only if one takes oneself to have sufficient evidence for it (336, 337, 341). But this raises a serious problem. Imagine that Tony for some reason or other strongly desires to believe that p, but finds himself with insufficient evidence for p. What Zamulinski seems to suggest is that, given that Tony can only believe that p if he believes that his evidence for p is sufficient, Tony will alter his evidential standards (his beliefs about what constitutes good evidence for some belief). And weakening his evidential standards will dangerously affect his belief acquisition and maintenance in the future. If it is true, however, that one cannot have some belief that p unless one believes to have sufficient evidence for p, how can one alter one's evidential standards – themselves beliefs – without sufficient evidence for doing that? In such cases there does not seem to be any epistemic reason (evidential reason) to change one's evidential standards. Changing one's evidential standards at will is (at least psychologically) impossible on Zamulinski's account. Hence, Zamulinski's account of overbelieving runs foul of his own assumption that one cannot believe that p unless one believes to have sufficient evidence for p.

Zamulinski could deny that one normally believes some proposition p only if one has sufficient evidence for p. But if that is true, then one need not change one's evidential standards in order to overbelieve that p. And, as we have seen, this premise is crucial in his argument for the moral premise.

Another objection could be that my counter-argument has the uneasy consequence that nobody believes upon insufficient evidence. Fortunately, it has no such consequence. What happens in cases of overbelief, as it seems to me, is that people, rather than changing their evidential standards, focus on evidence that affirms their belief and that they avoid considering evidence to the contrary, evidence they do possess. They might also purposely avoid considering their evidential standards. But if this is what people do when they overbelieve, then by overbelieving they do not affect or alter their evidential standards. Hence, overbelieving is not morally reprehensible, at least not for the reason Zamulinski puts forward.

Zamulinski might attempt to buttress up his argument by saying that if in certain circumstances some person S overbelieves that p by focusing on some minor part of his evidence or by purposely not considering his evidential standards, she will probably act similarly in like situations. And if she will, then she is more likely to put people at risk in those situations. This revised version of the argument seems to me as unconvincing as the original one. Whether one decides to pay attention to only part of one's evidence or not depends on all sorts of considerations (moral, prudential, epistemic, religious, legal, etc.), considerations that will differ from situation to situation. There are situations in which it is proper (or rational all-things-considered, if you like) to bring to mind only part of one's evidence or to purposely neglect some piece of evidence. In order to keep a good professional relationship with my employee, I might decide to avoid thinking about my knowledge of how he treats his children. Given that my overbelief issues from a voluntary act, upon considering several aspects of the situation, there is no reason to think that I will have a similar overbelief in other situations. Such well-considered overbeliefs, therefore, are not morally reprehensible. And this means that Zamulinski has failed to establish the moral premise.

Excuse conditions and doxastic involuntarism

As I said, Zamulinski distinguishes different excuse conditions for some person S's moral doxastic responsibility: (a) S has been indoctrinated to believe that p, (b) S has overbelieved p, acted in accordance with p, and thus cemented her commitment to p, (c) there is an evolutionary explanation for S's belief that p, and (d) S's belief that p comes with her genetic inheritance (340). I doubt whether all of these conditions – especially (b) and (c) – excuse one entirely, but let us assume they do. Here is how this is relevant: we have good reason to think that in the case of people's religious beliefs, often one or more of these four excuse conditions are instantiated.

First, many religious believers have been indoctrinated by sect leaders, such as certain members of the Church of Scientology and Aum Shrinrikyo. But one might also think of adherents of mainstream religions, such as members of the Roman Catholic Church, that, say, have so strongly been influenced in their youth by religious education, that time and again they find themselves with certain religious beliefs that they cannot get rid of. Second, assuming for the moment that there is insufficient evidence for religious beliefs, many religious believers have indeed acted on the basis of their overbeliefs (and the same seems to apply mutatis mutandis to atheists or agnostics), so that they are firmly rooted in their opinions. Third, several scholars have argued that there is an evolutionary explanation for religious beliefs.Footnote ⁹ Religious believers might disagree about this, but if it is true, then, on Zamulinski's account, they are morally excused for having those religious beliefs. Fourth and finally, a widespread tradition in Christian theology has it that belief in God is somehow built-in. Here, one might think especially of Calvin's idea of the sensus divinitatis.Footnote ¹⁰ Whether or not such beliefs come with our genetic inheritance, it is clear that we are in no way responsible for having them.

One might question whether one or more of these excuse conditions are exemplified in the lives of all religious believers, but remember that the conclusion of MA concerns every human being and that it, therefore, requires that none of these excuse conditions is ever satisfied. And that claim just seems far too strong.

My second criticism, which is closely related to the first, is as follows. At several junctures in his article, Zamulinski acknowledges that we cannot believe at will – he thus endorses what has been called the thesis of doxastic involuntarism – and that we, therefore, cannot overbelieve at will either (337, 341). According to Zamulinski, doxastic involuntarism does not render the ethics of belief senseless, though, because we have indirect control over what we believe by improving or deteriorating our evidential standards, our logical and statistical abilities, etc. And those having insufficient time to do these things can rely on some authority X if they have sufficient reason to believe that X is truthful, that X has the expertise necessary for acquiring a legitimate opinion about the issue in question, and that X has exercised his ability in coming to have the belief in question (337–338).

Now, many Christians and adherents of other religions at a certain point in their lives find themselves with certain religious beliefs. At least some of them will have profoundly reflected on their religious beliefs and believe that they have sufficient evidence for their religious beliefs. In opposition to Clifford's shipowner,Footnote ¹¹ they have not suppressed their doubts. Rather, they have thoroughly tested their evidence for their religious beliefs. But if that is true, how can they be held morally accountable for believing that p? Another option, as I said above in commenting on the ambiguity of MA issuing from different interpretations of (2) and (5), is to argue that Christians are to be held accountable for having the wrong subjective evidential standards. And, again, I would say that there might be some Christians that are blameworthy for having the wrong evidential standards, but that there is no reason to think that this is true for all Christians.Footnote ¹²