See the first post in this series here.

In his well known but deeply incoherent letter to his nephew, Peter Carr, Thomas Jefferson recommends to Carr a strong foundationalist approach to the examination of purportedly miraculous events.  Given Jefferson’s commendation of the work of David Hume on matters religious and the resemblance of Jefferson’s program for assessing the purportedly supernatural, I think we are safe to infer that the Jeffersonian program simply is the Humean program.  But this means that Jefferson’s procedure for deciding whether or not a miraculous event did in fact occur is fatally flawed on two counts.  The first fatal flaw is the commitment of the Humean (and Jeffersonian) program to strong foundationalism, given the self-referential incoherency of strong foundationalism.  The second fatal flaw has to do with the corrosive effects of Humean epistemology upon knowledge of any kind, including knowledge arrived at by way of inductive generalizations from experience.  Moreover, Hume may be aware of the ironic implications of his skepticisms for his treatment of rational belief in the miraculous.  Jefferson clearly is not.

But before unpacking Hume’s skeptical epistemology and its implications for dismissals of miraculous stories premised upon invocations of the (physical) laws of nature, we ought to recall just what Jefferson tells Carr concerning stories of miracles.  Just what is the Jeffersonian program here?  Well consider the story recounted in the book of Joshua in which the sun is said to have stood still.  When considering the veracity of a story such as this, Jefferson tells Carr that he should do two things.  First, according to Jefferson, the writer of the Book of Joshua claims to have been inspired in his recounting of events.  Consequently, Carr must “examine upon what evidence” such “pretensions” are founded.  More generously, Jefferson exhorts, “Examine therefore candidly what evidence there is of [the writer] having been inspired.”  Why should Carr do this?  “Because millions believe it.”  Now, from the standpoint of Jefferson’s modernist epistemology “Because millions believe it” can hardly, for him, provide epistemic warrant or rational evidence.  “Because millions believe it” cannot for Jefferson constitute a rational motivation for the inquiry.  This, again, holds for Jeffersonian and modernist rationality.  Other accounts of rationality might assign different epistemic weight to the testimony of millions.  But to return to Jefferson . . . Given that the belief of millions carries no epistemic or rational weight for the strong foundationalist, the actual motivation for a rational inquiry into a story of the miraculous can only be for one reason: so that Carr’s beliefs concerning such stories derive from the evidence—that is, so that Carr’s beliefs concerning the miraculous derive from duly supported prior beliefs and/or from beliefs that are properly basic. 

Second, when evaluating a story of a purportedly miraculous event, Carr is to consider other things he knows, such as the orbit of the earth around the sun or the result, in accordance of the laws of nature, of the earth stopping its rotation so that the sun could remain in its place in the sky.  Carr is to evaluate the story in Joshua based on these other things he knows.  The reasoning goes like this: If the sun really stood still in the sky, then, given the laws of planetary motion, the earth must have stopped its rotation upon its axis; if the earth stopped rotating upon its axis, then persons and animals and trees and buildings should been thrown to the ground.  Indeed, they should have been “prostrated” a second time when the earth started moving again.  Jefferson is implying that we have no evidence (and, in particular, no written account) of one much less two general prostrations, as a result of the earth stopping and then resuming its rotation upon its axis.  Consequently, Jefferson poses what presents itself as a rather rhetorical question: “Is this arrest of the earth’s motion, or the evidence which affirms it, most within the law of probabilities?”  Jefferson quite clearly takes the latter to be the case and so rather clearly insinuates that Carr should therefore reject the story as being true.  The rational person will therefore treat such stories as legendary developments such as those that enter the writings of Livy or Tacitus.

Jefferson goes on to apply this way of reasoning to the accounts of Jesus contained in what Christian’s call the New Testament.  Thus, Jefferson tells Carr to “Keep in your eye the opposite pretensions 1. Of those who say he was begotten by god, born of a virgin, suspended & reversed the laws of nature at will, & ascended bodily into heaven: and 2. Of those who say he was a man of illegitimate birth, of a benevolent heart, enthusiastic in mind, who set out without pretensions to divinity, ended up believing them, & was punished capitally by the Roman law which punished the first commission of that offense by whipping, & the second by exile or death in furca.”  Jefferson’s skepticism of the first “pretension” clearly results directly from applying the sort of program commended to Carr above to claims about Jesus of Nazareth contained in Christian Scripture.  What should be abundantly clear, therefore, is that if such a procedure of analysis is logically flawed, then stories of the miraculous cannot be discounted as irrational on account of their conflict with said procedure.  And, of course, the Humean procedure for analyzing stories of the miraculous is a procedure premised on strong foundationalism.  And strong foundationalism is self-referentially incoherent.  But, if we take Hume’s work on the understanding as a whole, Hume’s claim that it is irrational to believe in the miraculous is self-defeating in another way.

If we turn to Hume’s claims concerning the nature of human understanding, the problem for Hume’s claim concerning the irrationality of belief in the miraculous begins to emerge.  According to Hume, there are two propositions that we can know to be the case: analytic and synthetic.  Analytic propositions are known a priori.  They are known because they are undeniable (i.e., to deny the truth of such a proposition involves one in a self contradiction).  But Hume held that all analytic propositions are tautological and so philosophically trivial.  Synthetic propositions, by way of contrast, are those that we know based on observation.  All propositions other than those that are analytic or synthetic Hume referred to as nonsense.  As the historian of philosophy, Donald Palmer points out, this suggests a certain program for the assessment of any proposition to see whether it makes sense.  With regard to any proposition, we ask first whether the proposition is analytic.  If it is analytic, then the proposition is true just because the denial would entail a contradiction.  However, the proposition is also (according to Hume) tautological and therefore philosophically meaningless.  Thus, it is true that all bachelors are unmarried men. This claim is true given the definition of men and bachelors.  But the proposition is only true by definition and so does not amount to a substantial claim.  Moving forward with the Humean program, if the proposition is not analytic, then we must ask whether or not it is synthetic.  But for a proposition to be synthetic, it must be based on observation.  The only philosophical claims of substance, for Hume, are those based on (or perhaps simply descriptive of) experience.  If a proposition is neither analytic (undeniable and tautological) nor synthetic, then Hume says it is nonsense.  But just here we must note the number of propositions that are neither true by definition nor empirically observable—and that are, therefore, nonsensical.  It seems Hume would place into this category God or the cosmos (that is, of an external world) or material substance or even the self.  And given Hume’s epistemology, other minds must be considered nonsense as well.  For none of these things (God, the external world, material substance, the self, and other minds) are undeniable.  Consequently, none of these things are analytic.  Nor are they synthetic.  For none of these things are observable.  From the Humean standpoint, they must all therefore be nonsense.  But the Humean list does not stop here.  According to Hume, the law of causality is also neither undeniable (and so not analytic) nor observable (and so not synthetic) and must, therefore, also be considered nonsense.  To be a bit more precise, that x causes y is neither undeniable (even if occurrence of x always follows upon the occurrence of y) nor observable.  Hume’s argument may seem counterintuitive.  But what we observe is the coincidence of x and y, the contiguity of x and y, and the priority of one to the other (say of x to y).  From these three observations, we infer causation.  But Hume is entirely right to claim that we never empirically observe causation.  And here is the lynchpin of the conflict between Humean epistemology and Hume’s argument against rational belief in miracles.  For if we never observe causation, then causal claims (which are quite obviously not analytic in nature) cannot possess the status of being true.  From this follows the impossibility of knowledge of causal laws based upon inductive generalizations.  That is, Humean epistemology prevents us from being able predict any future occurrence based on the regular occurrence of past events.  Given Hume’s epistemology, it is impossible to know even that the sun will rise tomorrow. 

Let’s apply Hume’s epistemology to Jefferson’s letter to Carr.  The Humean criteria for knowledge entails that we cannot know causality.  But if we cannot know the principle of causality, then we cannot have knowledge of the laws of nature, for such laws are causal laws (or causal relations).  But in that case, we cannot come to know that miracles do not occur OR that they are improbable such that we can never have knowledge of their occurrence on the basis of the laws of nature.  That is, if the principle of causality is nonsense, then so must be claims to know the laws of nature.  And, in that case, any claim that rational knowledge of those laws should discount our belief in stories of the miraculous is also nonsense. 

The irony is delicious.  Hume’s analysis of belief in the miraculous cannot hold given Hume’s analysis of the understanding—the latter defeats the former.  Indeed, Hume’s analysis of the understanding, given empiricism, may undermine the possibility of rational knowledge of any sort.  But Hume’s analysis of the understanding is an analysis given empiricism.  One might argue that Hume’s work therefore constitutes a reductio ad absurdum refutation of empiricism.  And there is at least some indication that Hume thought of his work in this way.  In 1754 he wrote the following to one John Stewart: “But allow me to tell you that I never asserted so absurd a Proposition as that anything might arise without a cause: I only maintained that our Certainty of that proposition proceeded neither from Intuition nor Demonstration, but from another source.”  It is possible to read Hume as here asserting that we have certain knowledge of the law of causality and that such knowledge is neither analytic nor synthetic.  Likewise, we could read him as saying that knowledge of the principle of causality is neither innate (as for Descartes) nor achieve by observation and inductive generalization from observation.  But I will not claim too much in the way of interpreting Hume.  I claim only that his epistemology, given empiricism, shoots in the foot the claim that belief in accounts of the miraculous is irrational because improbable given what we know.  For what we know (or the relevant probabilities) are supposed to be the laws of nature.  But Hume’s epistemology entails that such laws (whether as deterministic or even as probabilistic laws) cannot be known.  Of this, Jefferson seems blissfully unaware.