Hume on prophecy

Of prophecies, Hume writes:

These few lines constitute the entirety of Hume's remarks on the conceptual nature of prophecies.Footnote ⁴ However, brief though they are, these lines are telling.

First, this passage makes plain that Hume is clearly concerned with predictive prophecies, or acts of foretelling future events. While the text supports the claim that Hume, in examining prophecies, is concerned with the foretelling of future events, one would be remiss – irrespective of the textual evidence – simply to assume this. Such an assumption would overlook that fact that many major religions espouse a notion of what is sometimes called ‘non-predictive’ prophecy. As Scott Davison explains, ‘In the great monotheistic religious traditions (Judaism, Christianity, and Islam), for instance, a prophet is a person who makes God's will particularly clear, whether or not doing so involves making any predictions about the future.’Footnote ⁵

Second, this passage also reveals that Hume is principally concerned with religious prophecies: acts of foretelling future events that are called upon as evidence to support religious claims. These acts of foretelling, in order to count as religious prophecies on Hume's account, must ‘exceed the capacity of human nature’. For Hume, champion of philosophical naturalism, understanding human nature is an exercise in the ‘science of human nature’;Footnote ⁶ by experience and observation we uncover the basic laws, understood as causal regularities in nature, that govern the operations of the human mind. So, on Hume's view, to say that prophetic acts ‘exceed the capacity of human nature’ is to say that they violate the laws of nature.

Finally, since on Hume's account prophecies must be violations of the laws of nature, and since these violations are brought about by divine intervention, they satisfy his definition of ‘miracle’: ‘A miracle may be accurately defined, a transgression of a law of nature by a particular volition of the Deity ...’Footnote ⁷ So Hume, in keeping with many of the major philosophical and religious traditions of his time, defines ‘miracle’ as a violation of a law of nature through divine intervention:Footnote ⁸

Thus, Hume writes, ‘all prophecies are real miracles'. But clearly, on any reasonable account, prophecies and miracles are not co-extensive. On Hume's account prophecies are only a species of miracle. What differentiates prophecies from other species of miracle is that prophecies, as already noted, involve the foretelling of future events. So we can sum up Hume's definition of ‘prophecy’ as follows:

And, by simply substituting in Hume's definition of ‘miracle’ we have:

Because prophecies are a species of miracle, Hume takes it to be the case that his argument against miracles ‘may be applied, without any variation, to prophecies' with the result that the two pillars undergirding reasonable religious belief are razed. As Hume explains in the final lines of Section 10, immediately following his brief remarks on prophecies,

On Hume's view, then, belief in God or something divine cannot be warranted by appeals to miracles or prophecies.

Prophecy and testimony

Let's begin by granting, for the sake of argument, that Hume's argument against miracles is sound. Does it undermine prophecies as he claims? In one respect, this question has a very straightforward answer: yes, given that Hume simply defines ‘prophecies' as a species of miracle, what he writes about miracles clearly applies to prophecies. But this answer will hardly satisfy even the most light-handed critic, as the worry naturally gets pushed back to Hume's definition of ‘prophecy’. The question becomes: is Hume, in claiming that his argument against miracles applies to prophecies, working with an accurate conception of ‘prophecy’?

Peter Harrison argues that Hume is not. Harrison writes: ‘Hume's chief objection to miracles – that one is never justified in crediting second-hand testimony to miracles – does not necessarily apply to the argument from fulfilled prophecies as it was understood in the eighteenth century.’Footnote ¹⁰ This ‘chief objection’ against miracles appeals to the infirmity of testimony of miracles. Indeed, Hume puts forth several considerations in favour of the view that reports of miracles are very unreliable. He points out that: (1) that most supposed miracles haven't been ‘attested by a sufficient number of men’;Footnote ¹¹ (2) ‘the passion of surprize and wonder, arising from miracles, being an agreeable emotion, gives a sensible tendency towards the belief of those events, from which it is derived ... But if the spirit of religion join itself to the love of wonder, there is an end of common sense; and human testimony, in these circumstances, loses all pretensions to authority’;Footnote ¹² (3) ‘It forms a strong presumption against all supernatural and miraculous relations, that they are observed chiefly to abound among ignorant and barbarous nations';Footnote ¹³ (4) and, finally, ‘there is no testimony for any, even those which have not been expressly detected, that is not opposed by an infinite number of witnesses.’Footnote ¹⁴ Hume argues that these considerations show that testimony of miracles is so unreliable that no one has ever been justified in believing reports of miracles.

But, according to Harrison, the fulfilment of prophecies, unlike reports of past miracles, need not rely on testimony. One may directly witness the fulfilment of a prophecy. So, since appeals to prophecies as support for religious claims need not rely on testimony, they are immune to Hume's ‘chief objection’: ‘The advantage of prophecies ... lies in the fact that the fulfillment of prophecy provides an “ocular demonstration” in a way that no historical report of a miracle can.’Footnote ¹⁵ Or so Harrison argues.

However, Harrison's argumentative strategy runs amiss in at least two ways. First, Harrison overlooks the fact that Hume was concerned to undermine not only historical reports but also direct observation of purported miracles as a source of religious justification. Consider the following passage from ‘Of miracles':

In this passage Hume makes his point not only in terms of those who hear reports of miracles, but also in terms of those who witness miracles. We can see, then, that Hume explicitly acknowledges that miracles, in functioning as evidence for religious claims, need not rely on testimony.

Second, even if we were to grant Harrison that prophecies have some sort of ‘advantage’ of ‘ocular demonstration’, Harrison ignores the fact that Hume's argument against miracles has two parts. In addition to appealing to the infirmities of testimony, Hume argues:

Here, rather than emphasizing the infirmities of testimony, Hume emphasizes the robust evidence that undergirds laws of nature: the ‘proof’ of laws of nature outweighs the weaker evidence underlying the ‘possibility’ of miracles. So, even if it turns out that some report of a miracle seems particularly reliable, it's less reliable than the regularities of nature which have been established by repeated experience.

While the argument in the passage quoted above is cashed out in terms of testimony, we have good reason to think that Hume took the central insight to apply to direct observation of miracles, as well as testimony of miracles, since, as I already pointed out, he was explicitly concerned with the fact that people cite direct observation of miracles as religious justification. Here is an example to illustrate how this parallel line of thinking might proceed: imagine that you observe what you take to be a violation of the laws of nature brought about by divine intervention. Given the regularity that characterizes the laws of nature, it is more likely you were mistaken or fooled about what you saw than it is likely that you indeed witnessed a violation of the laws of nature. But, even assuming that the circumstances of your observation were normal (there was good lighting and you were wearing your glasses, you weren't drunk, no one was pulling off an elaborate hoax, etc.), it seems that the apparent violation would be better explained by the fact that what we took to be a law of nature was not, in fact, a law. People can have mistaken beliefs about the laws of nature. This is not a surprising observation when we note that our evidence for many laws of nature is not remarkably robust; for example, today, we know that there are laws of nature that underlie the fact that smoking can cause cancer, but given our lack of scientific evidence, we have not fully specified these laws by making precise the connections between the chemical compounds in cigarettes and cellular changes. So it seems we can better explain an event that violates what some or even many believe to be a law of nature by changing our conception of the law to accommodate the event while still holding on to our other observations of constant conjunction – all without invoking divine intervention.

It is a short leap from this sort of story about the direct observation of supposed miracles to a story about first-hand experience of prophecies. Imagine that someone had a first-hand experience, such as confidently predicting some unlikely event that eventually came to pass, that made him believe that he is a prophesier. On Hume's view, this individual would be more justified in believing that he'd made a lucky guess, for example, than in believing that a divine being had empowered him in direct violation of the laws of nature. So it seems that Hume's argument against miracles, if sound, is also plausibly construed as an argument against prophecies.

Prophecy and divine intervention

But Harrison also claims that Hume overlooks the fact that prophecies, unlike miracles, need not entail violations of the laws of nature. As Harrison explains,

Harrison insists, here, that we can remain agnostic about the source of a prophecy. But recall Hume's definition of ‘prophecy’:

If Harrison is correct that we can remain agnostic about the source of prophecy, then Hume's definition of ‘prophecy’ is clearly undermined.

In one respect, I think Harrison is correct. He is correct that, contra Hume, a prophet's capacities to foretell future events need not be supernatural. That is, we have no reason to think, besides ad hoc stipulation, that the concept of ‘prophecy’ does not accommodate foretellings of future events brought about in accordance with the laws of nature by rather unextraordinary means such as vivid dreams, imaginings, or hallucinations. Surely, being a prophet involves making reliable predictions and not just lucky guesses and so it may be tempting, following Hume, to think that the reliability of these predictions can only be explained by appeal to the miraculous. However, as Harrison rightly argues, it's plausible to think that God could ensure the reliability of a prophet's predictions by acting within the laws of nature. Harrison and I are not alone in our agreement on this point. As Harrison points out, Augustine and Aquinas both argued that God's fulfilment of petitionary prayers, while properly thought of as divine intervention, need not be thought of as miraculous; according to these philosophers, ‘all potential petitions are foreseen by God and prior arrangements are made as it were to ensure their accomplishment’ in accordance with the laws of nature.Footnote ¹⁹ And Spinoza, for example, held that at least some prophets proceed from an exceptional imagination, rather like those of poets. He wrote that ‘revealed things were imagined by the prophets in a most vivid manner’,Footnote ²⁰ and that ‘the prophets almost always make their communications allegorically or enigmatically, and give bodily shape and form to spiritual things in general. The procedure is in entire conformity with the nature of the imaginative faculty.’Footnote ²¹

But in another, more important sense, Harrison is wrong. Even if we can remain ‘agnostic about the mechanisms involved in prognostication’, it does not follow that prophecies are left untouched by Hume's arguments. In fact, Hume's conception of ‘prophecy’ can (and should) be made to accommodate Harrison's critique while still maintaining the central aim of Hume's project: to undermine miracles and prophecies as a source of religious justification. This is my task in the next section.

A revised account of prophecy

Unsatisfied with Hume's definition of ‘prophecy’ Harrison offers us an account along the following lines:

However, this account of ‘prophecy’ is unacceptable. Even if we grant Harrison's point that a prophet may acquire knowledge of the future in accordance with the laws of nature, then that which a prophet prophesies about, the future event, must be miraculous.

So we should reject Harrison's account of prophecy, PH, in favour of a new, revised account of prophecy, PR:

Before defending PR a few clarificatory remarks are in order. First, the definition – for brevity's sake – builds upon PM. Second, the use of ‘or’ in the definition is exclusive. Third, and finally, I have used the term ‘mundane’ to mean ‘not miraculous', again for brevity's sake.

The first and third disjuncts of PR are consistent with P. What distinguishes PR, then, from Hume's definition of ‘prophecy’ is the second disjunct: the mundane act of foretelling, brought about by a ‘particular volition’ of God, a miraculous future event. For the reasons suggested by Harrison, and briefly spelled out above, our account of ‘prophecy’ should allow for the fact that a prophet's capacity to foretell future events need not be supernatural. However, in order for a foretelling to count as a religious prophecy it must, in Hume's words, be the result of a ‘particular volition of a Deity’. Why this qualification? Consider: the Apostles trailed around after Jesus as he cured the sick by laying his hands on them. Imagine that Barnabas, after seeing Jesus cure numerous people, predicted that the next sick person Jesus laid hands on would be cured. It would be a stretch to call this prediction ‘prophecy’.Footnote ²³ What's missing from this example, and what would preserve what's arguably our common-sense understanding and usage of ‘prophecy’, is divine intervention. But such intervention need not be miraculous because, again, God could intervene by operating within the laws of nature; although, in Hume's vernacular, such a foretelling may very well be ‘marvellous'. While miraculous events defy the laws of nature, marvellous events are consistent with the laws of nature (i.e. they're mundane) but are uncommon.Footnote ²⁴

So why should we reject Harrison's definition of ‘prophecy’, PH, in favour of PR? Because, following Hume, we're concerned to give an account of religious prophecy, and PH is too broad to do so. If neither the act of foretelling a future event nor the future event that is foretold is miraculous, then there is no evidence is to be adduced for the existence of God, or for the truth of a religious claim. This is so because if both prophecy and fulfilled prophecy can be entirely explained within a materialist framework governed by the laws of nature, then the supernatural is crowded out. Suppose a ‘prophet’ were to tell me that he'd had a lucid dream – one that he admittedly had no reason to think was brought about by miraculous divine intervention. Imagine now that this ‘prophet’ goes on to tell me that in his dream he'd had a vision that tomorrow the sun will rise and that, given his prediction, the coming to pass of this event will constitute evidence for the existence of God. When the sun rises tomorrow, it does not seem to me that in light of the supposed prophet's claims I (or he, for that matter) will have any more reason to believe in the existence of God than I have in this very moment. Since neither the ‘prophet's' foretelling nor what he foretells are miraculous, then, as Hume says, ‘it would be absurd to employ’ them ‘as an argument for a divine mission or authority from heaven’. Given Hume's arguments against miracles, either the act of foretelling a future event or the occurrence of that event must be miraculous in order to undergird religious belief.