The standard Mad, Bad, or God argument
A certain argument for Christianity has appeared in various forms since the twentieth century, and it seems to trace back to Pascal. C. S. Lewis, arguably its most popular defender, puts it this way:
A man who was merely a man and said the sort of things Jesus said would not be a great moral teacher. He would either be a lunatic – on a level with the man who says he is a poached egg – or else he would be the Devil of Hell. You must make your choice. Either this man was, and is, the Son of God: or else a madman or something worse. You can shut Him up for a fool, you can spit at Him and kill Him as a demon; or you can fall at His feet and call Him Lord and God. But let us not come with patronising nonsense about His being a great human teacher. He has not left that open to us. He did not intend to.Footnote 1
This sort of argument is sometimes called the Mad, Bad, or God (hereafter MBG) argument. The version I will call the standard is summed up as follows:
P1. Jesus claimed, explicitly or implicitly, to be divine.
P2. If (P1), then either Jesus was right or he was wrong.
P3. If he was wrong, then either
a. he was lying (he was bad), or
b. he was institutionalizable (he was mad), or
c. he was merely mistaken.
P4. He was not lying (not–a).
P5. He was not institutionalizable (not–b).
P6. He was not merely mistaken (not–c).
C1. He was right; Jesus was and is divine.Footnote 2
The idea is that our evidence suggests that Jesus asserted he was divine, and that no explanation of this assertion fits the rest of our evidence about him except the one according to which he really is divine (call this explanation the God option). We have good reason to rule out the leading alternative explanations: that Jesus was the sort of person who would lie about being divine (the bad option), or who would be diagnosable as clinically insane (the mad option), or who would have sanely, sincerely, but mistakenly, claimed he was divine (the merely mistaken option). So our evidence indicates Jesus must have been right – been divine – and so there is prima facie reason to believe that Christianity, or at least one of its central claims, is true.
In the next three sections I point out defects in the standard MBG argument, one of which is critical. Contra Stephen Davis, I argue that Daniel Howard-Snyder has exposed the critical flaw that (P6) is inadequately supported.Footnote 3 Even if we rule out that Jesus was mad or bad, the mere implausibility of him being merely mistaken does not establish (P6). In the final section I sketch a more promising version of the MBG argument – an explicitly ‘inference to the best explanation’ version.
The bad in the standard Mad, Bad, or God argument
‘Merely mistaken’ is too hard to rule out
Davis suggests that the merely mistaken option is improbable, just as the mad and bad options are improbable. Howard-Snyder does not think this is a good enough defence of (P6), since certain stories according to which Jesus was ‘merely mistaken’ in claiming to be divine are epistemically possible and are not significantly less plausible or likely than the God option. Howard-Snyder's view seems to be that the God option must be significantly more plausible or likely than the mad, bad, and merely mistaken options if the standard MBG argument is rationally to support it. If the God option is only slightly more, or equally, plausible, it is most rational to suspend judgement. And Howard-Snyder thinks there are at least two versions of the merely mistaken option that aren't significantly less plausible or likely than the God option.Footnote 4 First he offers the Beelzebub Story (BS), in which Satan tricks Jesus into rationally but falsely believing he is divine by ‘duplicating’ the grounds for rational belief in Jesus' divinity (call this Jesus' divinity belief). These duplicated grounds might include an experience of ‘what it's like’ to be divine or of a clear intimacy with God. And these grounds might be supported by Jesus' seeming to perform miracles. So on BS, Jesus could be merely mistaken in believing, and thus in claiming, that he is divine, while not being a liar or institutionalizable.
The second story is consistent with metaphysical naturalism: the Messianic Story (MS). Here, Jesus, by studying scripture, comes to believe that he is the Messiah, and that the Messiah is divine, and then infers that he himself must be divine. On MS, Jesus again has grounds for rational belief in his own divinity, so his claiming that he is divine would not be mad or bad but merely mistaken. Howard-Snyder does not claim these stories are likely to be true or even logically possible, but he claims that our evidence does not make them significantly less plausible or likely than the God option. It could turn out that these stories are necessarily false, but, for all we know or have reason to believe, they could be true, and presently there is no rational reason to prefer the God option over them. So we do not have sufficient support for (P6), and, thus, do not have sufficient support for (C1).Footnote 5
In reply, Davis suggests BS and MS are improbable and difficult to make plausible.Footnote 6 Against BS, Davis says, first, it's not plausible on our evidence that BS describes what in fact happened to Jesus, and second, Satan couldn't duplicate the rational grounds for Jesus' divinity belief. For example, it's not clear whether Satan could perform miracles, and it is clear that he can't raise people from the dead, yet Jesus raised Lazarus from the dead. And Satan could not know what it's like to be divinity incarnate because only a divinity could know that; so there's no way Satan could have given Jesus an experience of what it's like to be divinity incarnate.
Davis's first point is consistent with Howard-Snyder's objection that (P6) is not sufficiently supported. According to Howard-Snyder, (P6) is sufficiently supported only if we have reason to think all the merely mistaken options are significantly less plausible or likely than the God option. Davis's claim that BS is an implausible account of what happened to Jesus might be right. But that's no reason to think the God option is more plausible, or, for that matter, significantly more plausible. In other words, the claim that BS is implausible is, by itself, not enough to establish (P6).
Davis's second point also seems inadequate even if correct. We can grant that Satan couldn't have ‘duplicated’ certain grounds that would make Jesus' divinity belief rational – that Satan couldn't have raised Lazarus from the dead, and couldn't have shared ‘what it's like’ to be divinity incarnate with Jesus, because Satan cannot know what that's like. However, Satan needn't have known how or been able to do any of these things in order to give Jesus rational grounds for his divinity belief. He merely could have adequately counterfeited each thing, such that each at least in part rationally grounded or confirmed Jesus' divinity belief. For instance, Satan could have made it seem like Lazarus had been resurrected; he could have produced a sufficiently realistic, mass illusion that would have rationally grounded the belief that he had been raised from the dead. Or, Satan could have shared with Jesus an experience that merely seemed like what it's like to be divinity incarnate, such that it would have rationally grounded Jesus' divinity belief. (Even though only God knows what it's like to be divinity incarnate, certain incredible experiences would to anyone else still seem like what it's like.) Both of these are examples of how Satan could have led Jesus to have a merely mistaken divinity belief, while granting that Satan could not duplicate certain events that would rationally ground the divinity belief. So BS still seems like an epistemically possible story according to which Jesus was merely mistaken in claiming to be divine. Davis might be right that BS is implausible and improbable. However, it's not clearly worse than the God option, and on Howard-Snyder's view, the God option must be significantly more plausible or likely than the others if the standard MBG argument is rationally to support it. And this view seems correct.
Now for Davis's objections to MS: he claims, first, that no observant first-century Jew would have condoned the idolatrous belief that one is divine, and second, that there is no evidence for MS. Here, Davis's first point seems moral, not epistemic. Perhaps it would have been immoral for a mere human to believe himself to be divine (or immoral to condone such a belief), but that's consistent with it being epistemically rational to believe oneself divine. But more importantly, if one no longer has reason to believe one is a mere human (because one now has reason to believe oneself divine), then the idolatry issue disappears. Had scripture led Jesus to believe that he was the Messiah and that the Messiah was divine, then he would have had reason to believe he was not merely human and thus would not be idolatrous in believing himself divine. So, concerns about idolatry seem to dissipate in cases where Jesus had reason to believe he was divine. The real issue is whether on MS Jesus would have had rational grounds for his divinity belief. And it seems he would have.
Davis's second objection, that there is no evidence for MS, is ambiguous. On the one hand, it might mean that there is no epistemic reason to believe MS rather than other options – in particular, rather than the God option. However, even if true this would be no indication that MS is significantly less plausible or likely than the God option, so there still is not enough support for (P6). On the other hand, Davis might mean that we have no epistemic reason at all to believe MS. In that case, it seems he would be mistaken. MS tells a coherent, epistemically possible story that explains Jesus' claims. Roughly, a story that entails no obvious contradiction, that is consistent with what one has reason to believe, and that accounts for some phenomenon we have reason to believe existed, has some evidence in its favour. Indeed, part of the appeal of the God option is that it seems evidentially supported in this way. But so does MS. So there is some epistemic reason for believing MS. On neither interpretation of ‘there is no evidence’, then, does Davis's second objection successfully rule out MS and thus help establish (P6).
Overall, Davis is attacking the plausibility of BS and MS, not their plausibility relative to the God option. Merely discrediting the alternatives to the God option is not thereby supporting it. It could be that every conceivable option is implausible. So, it appears that Howard-Snyder's objection to the standard MBG argument stands: (P6) is not adequately supported. That we seem unable to support this premise properly is the argument's chief flaw.
‘Implicit claim’ is obscure
Davis argues for (P1) by citing evidence of Jesus implicitly claiming to be divine. According to Davis, to claim implicitly that P is either (1) to assert explicitly something that entails P, or (2) to assert explicitly something that one ‘can hold’ only if one ‘holds’ P, or (3) to do something that implies one believes P (if one is sensible).Footnote 7 Regarding the first two disjuncts, I will just point out that on neither of them does an implicit claim that P entail or even indicate belief that P, because one can assert what one does not believe. The third disjunct, the one Davis seems to be employing in (P1), I find difficult to understand. The example of it Davis gives is this: if Jones drinks from a drinking fountain he is implicitly claiming ‘The liquid emanating from this drinking fountain is potable’.Footnote 8 But here are two reasons this seems false even when Jones is sensible. First, there are clear counterexamples to it. Suppose Jones believes the water is poisonous and is attempting to kill himself by drinking it. It is hard to see in this case how his drinking the water implies he believes it is potable. Second, Jones might have a disposition to believe the proposition, but this is importantly different from his having either a dispositional or an occurrent belief in it.Footnote 9 Roughly, to have a disposition to believe P is to be in a position to form the belief that P, whereas to have a dispositional belief that P is to be in a position to retrieve or recall the belief that P. A dispositional belief is a belief that has already been formed but is not currently attended to (attended to beliefs are occurrent). Jones might drink from the fountain without ever having formed the belief that it is producing potable liquid, although he is likely inclined to believe that proposition once he considers it. So Davis's third disjunct does not imply belief in the implicit claim, though it might indicate a disposition to believe it. But now it's not clear what it means for Jesus to have implicitly claimed to be divine – is it that he had a mere disposition to believe it? If that's the case, then much of the standard MBG argument is unclear. Consider (P2): in what sense was it right or wrong for Jesus to have a disposition to believe he is divine? Or consider (P3): in what sense is having a disposition to believe sufficient for lying, being institutionalizable, or being mistaken? I submit that Davis's concept of implicit claim is obscure at best, especially given the role it's supposed to play in the standard MBG argument. So, it's not clear how to understand the argument.
Unnatural deduction
The standard MBG argument seems to reflect poorly the reasoning of its defenders. Consider another line from C. S. Lewis: ‘The historical difficulty of giving for the life, sayings, and influence of Jesus any explanation that is not harder than the Christian explanation, is very great … Hence the non-Christian hypotheses succeed one another with the restless fertility of bewilderment.’Footnote 10 Lewis's contention seems to be that the best explanation of Jesus' actions and claims is the Christian one. But this is not a deductive argument. It is an inference to the best explanation. In my view this is not a weakness of Lewis's argument, but it is something the standard MBG argument, which is deductive, poorly captures. The standards for these two types of arguments are different. It is beyond the scope of this article to compare them in detail, but I will point out that believing the best available good explanation of certain data is rational even if that explanation is not entailed by the data. The standard MBG argument poorly captures the reasoning of its advocates in such a way that makes its success more difficult to come by.
Divinity as the best explanation?
Here I state and briefly explain a more promising variant of the MBG argument, and then explain why defenders of MBG arguments should prefer it. Call it the explanatory version:
P7. Jesus behaved oddly for a first-century Jew.
P8. The best explanation of Jesus' behaviour is that he believed he was divine.
P9. The best explanation of Jesus believing he was divine is that his total evidence rationally supported it.
P10. The best explanation of Jesus' total evidence rationally supporting his divinity belief is that Jesus is divine.
C2. Jesus is divine. (P7–P10)
This better captures the reasoning of those who advocate MBG arguments. In general, what motivates MBG arguments is data about Jesus' actions and about certain events before and after his life. This data set is unusual. People in Jesus' time and culture did not act as he did: he claimed to be one with the Father, and to be the judge of humanity in the end times; he addressed God by name; he assumed he had the authority to forgive sins; and he in general behaved with much more authority than the Jewish prophets had. What's more, Jesus allegedly fulfilled prophesies, prophesied truly, and performed miracles. It is facts like these that initially motivate arguments for Jesus' divinity.
It is facts like these, at any rate, that suggest he believed he was divine. Defenders of MBG arguments find alternative explanations of these facts inferior. One alternative is that Jesus was wicked – he was perhaps trying to deceive people into thinking he was divine, or just doing and saying what he pleased regardless of how irreverent he was. On this alternative, Jesus did not believe he was divine. Another alternative explanation is that Jesus suffered from a sort of lunacy, and as a result he spouted nonsense and behaved erratically, and nothing he did had to do with a divinity belief. So, support for (P8) would eliminate the bad option in the standard MBG argument, and would also eliminate certain mad options.
(P9) states that the best explanation of his divinity belief is that his total evidence rationally supported that belief – Jesus on the whole had reason to believe he was divine. Alternatives to this explanation are that Jesus had no reason to believe he was divine, or that he had some (but insufficient or defeated) reason to believe he was divine – maybe he just couldn't help believing it, or on a whim he elected to believe it. These alternatives would suggest that Jesus was mad, or perhaps so epistemically irrational that he could very well count as mad; intuitively, a sane belief that one is divine must have extremely strong rational support. So these alternatives to (P9) demonstrate additional ways Jesus might have been institutionalizable. So, support for (P9) would show that certain mad options should be rejected.
(P10) states that the fact that Jesus' total evidence rationally supported his divinity belief is best explained by his being divine. This is the most contentious step of the argument. In order to support it one must show numerous alternative explanations to be inferior to the one stated in (P10) – that Jesus is divine. Here one must consider various merely mistaken options, according to which Jesus rationally but falsely believed he was divine. Howard-Snyder has offered two versions of that explanation: BS, according to which Satan tricked Jesus into believing he was divine, and MS, according to which Jesus through scripture acquired rational grounds for believing that he is the Messiah, and that the Messiah is divine, and then he inferred that he must therefore be divine. Another merely mistaken option is that Jesus had some cognitive disorder which caused him to hallucinate numerous interactions with angels, who attested to his divinity. Still another merely mistaken option is that Jesus wanted very badly to be divine, and his desire cognitively penetrated his perceptual experience such that he, say, heard the stones crying out his divinity. These versions of the merely mistaken option suggest that Jesus was epistemically rational yet possibly not institutionalizable.
This is not a defence of the explanatory MBG argument, but an outline of what a good defence would need to do. This argument is better than the standard version for several reasons. First, the explanatory argument makes no reference to ‘implicit claims’, and instead makes Jesus' divinity belief explicit. Second, (P9) makes explicit the rationality of Jesus' divinity belief.Footnote 11 It is compatible with the standard MBG argument that Jesus irrationally believed he was divine. Third, its premises are consistent with Jesus not being divine – with him, for instance, being merely mistaken. So the epistemic possibility and relative plausibility of Howard-Snyder's ‘Just So’ stories does not show that (C2) is inadequately supported. Fourth, by operating outside a probabilistic framework, the explanatory MBG argument is not clearly vulnerable to objections from ‘dwindling probabilities’.Footnote 12 Fifth, it isolates the most contentious step of the argument – (P10) – in such a way that helps reveal the kind of support (C2) ultimately requires. MBG arguments in general reason from a set of historical data that is localized to a certain period and region, to a conclusion that implies that there exists a divine being. The significance of this implication is massive. The explanatory MBG argument rightly requires that all background evidence for and against this implication weighs in; prior considerations about the truth or falsity of theism are not neglected. And, the explanatory MBG argument rightly requires that all background evidence for and against the truth of religious views incompatible with Jesus' divinity weigh in. The standard MBG argument, however, offers no clear point at which these relevant bodies of evidence come to bear, and so it obscures a potentially legitimate point of criticism.
I propose that defenders of MBG arguments would do best to adopt the explanatory version. It has all of the virtues of the standard version and none of its vices. The main difficulty for the defender of the explanatory MBG argument will lie in supporting (P10), which will require careful consideration of the enormous body of relevant evidence. But it is a virtue of the explanatory version that it respects this body of evidence, and that it makes obvious the place where this body matters.Footnote 13
