Being Justified

A common argument in popular apologetics premises the truth of Christianity upon the plausibility of the resurrection. Such arguments draw from various writers such as Gary Habermas (not to be confused with the Habermas you’ve probably heard of), C.S. Lewis, and the omnipresent William Lain Craig. They generally make the case that, due to historical factors [x, y, z], there is no better explanation for the ‘resurrection’ than the one provided by the mainstream Christian tradition (sorry, Schillerbeeckx). In this post, I am going to present an argument drawn from Hume and Ahmed against such approaches. I am then going to indicate the possibility for an alternative approach: a more sophisticated argument along the lines that we might be justified in such a belief in spite of this. I will develop this into something more concrete in a later post.

Central to considerations of arguments of this sort is the issue of the epistemic component in the justifiedness of holding a belief, which I shall (for the purposes of this article, and in full consciousness of the other uses of this term) call warrant. In any case of belief, there are two factors: the belief in a truth claim itself (that is, the assent to that truth claim on the part of the ‘knower’), and warrant.

In the case of abductive arguments (such as that in question), warrant comes from ‘evidence’. We say that there is evidence for a proposition (taking the general form “X is the case”) when a set of conditions Y arise which can be explained by X. We make the leap from evidence to warrant based upon two factors: firstly, is X the best explanation out of all the possible explanations for Y? Secondly, is X a sufficient explanation for Y – after all, our only option could be a completely inadequate explanation. In this case, we still would not be warranted in believing X, even in the absence of alternatives.

Both of these conditions (best and sufficient) rely on being able to distinguish between the qualities of explanations. This is where Bayesian probability comes in. To briefly summarise the relevant aspect of this, Bayesean probability dictates that an event is more likely to have arisen from a set of conditions the more times it has occurred in the past as a result of those conditions. This means that better explanations are ones that have explained similar events before. Warrant, as presupposed by Habermas et al., is derived from probability.

There are multiple ways this can play out with regards to the question of biblical evidence. At a basic level, no one has been resurrected before. However, there has historically been a lot of deception and mistakenness (including on the part of historians), and people will rationalize events and believe absolutely in their accounts. To this end, it is more likely that one of these explains the gospel accounts of the resurrection, rather than an actual resurrection occurring. Indeed, by this logic, we are even more warranted in believing that there is a better explanation that we have just not thought of than to attribute the ‘inspiration’ of the gospel accounts to the actual resurrection of a human body.

We can intensify this and say that belief in the resurrection can never be warranted. This is an argument most famously used by David Hume against miracles (defined as an event which goes against the established ‘laws’ of the universe) of any sort – the probability of a miracle is always less than an event which works within the bounds of those ‘laws’. This argument can be said to cut both ways, and there is some ambiguity in Hume as to whether he is not, in fact, arguing against the idea of natural laws. However, if we wish to keep our understanding of a consistent universe with immutable tendencies, we cannot accept this second reading. I offered this argument to a very friendly and eloquent evangelist with whom I spent several enjoyable hours discussing religion over coffee. His response, which is the standard response to Hume’s argument, is that it is a bad test because it automatically rules out the possibility of the resurrection. Here is an analogy: imagine you have bought a Bayesean fire alarm which, realising that in the vast majority of cases there is no fire in your kitchen, infers that it should not go off in any instance of possible fire. In the vast majority of instances, it would be correct - check your kitchen now: it’s probably not in flames. However, it would still be a pretty useless alarm precisely because we want to acknowledge the possibility of there being a fire. If we are going to take the possibility of the resurrection seriously, then, we cannot employ Hume’s argument.

We might respond that, assuming the coherence of Hume’s argument, what the argument shows is precisely that the resurrection should not be taken seriously in the first place. The believer might still respond that this is essentially begging the question. This may or may not be fair (the Bayesean might argue that it’s just a case of sticking with the best epistemic standards we’ve got). Alternatively, against such responses, Cambridge Philosopher Arif Ahmed develops a variation on Hume’s argument which demonstrates that we can, a priori, rule out miracles without neglecting the question of whether they should be entertained seriously. This gets round the objection that we are not taking the issue seriously because it undermines the justification for raising this objection in the first place. He asks us to imagine a continuum of possible worlds. At one end of the continuum is a world governed by rigid natural laws that never permit miracles. We will call this a ‘maximally lawful’ world. At the other is a world in which there are no natural laws, such that every event is essentially a miracle. We will call this a ‘chaotic’ world. Eliminating the possibility that we live in a chaotic world which appears lawful by chance due to its sheer improbability, when we look at our world, we realise that it is predominantly lawful. Every instance of lawful behaviour increases the probability that we live in a maximally lawful world. Given enough of these instances, the probability of our world being maximally lawful will outweigh all other possibilities such that we are warranted only to posit that the world is maximally lawful. Thus we would have no reason to take a claim for the resurrection (or indeed any other miracle) seriously, although we theoretically could have. Given that we have yet to observe a definite miracle (the Resurrection cannot be considered one without question begging) we are lead to this conclusion.

Note that these arguments do not imply that Christianity is false. Rather, they imply that if we wish to prove the truth of Christianity, we cannot rely so easily upon abductive arguments based on biblical accounts and ‘commonsense’ historical standards. The wider issue here is that there are a host of epistemological issues which must be engaged with if we are to draw any sort of rigorous conclusion. Moreover, as Ahmed shows, they present a particular problem for belief.

This problem, however, can possibly circumvented by taking stock of our notion of warrant. There are two considerations here, which are not entirely divorced from one another: firstly, just what are the conditions for warrant and why are they so; and secondly, when can we be morally justified in holding a belief in general? The first question is one for epistemologists, which I am not. The pertinent issue here, however, is that warrant has an ethical relevance. As such, the first question takes place against the backdrop of the second.

When we make a claim about the warrantedness of a belief, we are generally implicitly claiming something along these lines:

That there is some property, ‘warrant’, which is distinct from truth, that links a belief to truth in such a way that holding it is permitted (or possibly even obligated) according to some distinct ethics of belief which requires the knower to maximise the number of true beliefs, or at least minimise the number of false ones, in the set of beliefs which they hold, and that this property is a necessary condition for such an ethical status.

Note that this notion does not presuppose anything about the nature of the conditions for the possession of warrant, the objectivity of the property or ethics in general, nor about realism with regards to either of these things.

We might want to ask whether warrant itself has an ethical or normative component: is warrant derived from the normative framework which guides us to hold warranted beliefs over unwarranted beliefs? We might want to also question the necessary condition clause. That is, we might want to ask: are we ever justified in belief where we are not warranted?

The answer to the first question is not particularly obvious: on the one hand, we might instinctively want a notion of warrant which ‘stands on its own’. Such a notion of warrant could serve as a condition for justified belief which allows us to avoid wider ethical and meta-ethical complications. However, take, for example, the ethical maxim conveyed by the popular phrase, “Science: it works, bitches”. This exhorts us to adopt a particular kind of empiricist standard of warrant in response to its effectiveness in attaining particular goals. This is ethically driven, rather than purely epistemological, as it is explicitly motivated by ends.

I do not think that the answer to the second question is particularly obvious either. Indeed I think that there are some fairly obvious, plausible arguments to the contrary: e.g. a hedonist might say that a person in no position of intellectual authority (say, a shopkeeper) might be justified in believing in Russell’s teapot (you can see where this is going).

The openness of both of these questions leaves open the possibility for justified belief where warrant is not a necessary condition thereof. This opens possibilities for the apologist: the first question opens the possibility of adopting an ethics under which belief is warranted in the resurrection in spite of objections such as those raised by Hume and Ahmed. The second question raises the possibility of circumventing the question of warrant, dissolving their objections.

In development of the above point, I would like to note that this does not mean that the question of warrant can be circumvented entirely. For example, a scientist, philosopher or theologian might, as a result of their authority, have a duty to hold to a higher valuation of warrant than our shopkeeper. Equally, it may just be virtuous to value warrant. Indeed, it is quite possible that, while not a necessary condition for justified belief in a maximal form, a measure of warrant is still necessary for some or all beliefs to be justified in all cases. In fact, I believe that this is intuitively true (whether I am justified in this or not...)

This is where we might turn back to Habermas et al. Retaining their probabilistic notion of warrant, perhaps, given their arguments, we might say that the resurrection is plausible enough that belief is either warranted by virtue of some higher end to which warrant is subordinate, or justified because it fulfils a condition of warrant that is lower given other ethically-relevant factors. Both of these seem to me to be easier positions for the apologist to defend.

The obvious point to make here is that I have yet to actually produce anything concrete: under what ethical conditions might we suppose that this is the case? And could an actively persuasive case be constructed based on these conditions? I will explore these questions in a later article. Watch this space.