A Systematic Look at the “Witness Argument”: Hume’s Irony (Column 673)

With God’s help

Disclaimer: This post was translated from Hebrew using AI (ChatGPT 5 Thinking), so there may be inaccuracies or nuances lost. If something seems unclear, please refer to the Hebrew original or contact us for clarification.

In the last two columns I analyzed Hume’s critique of the “Witness Argument” and explained why, in my view, he erred in that critique. Here I wish to return to the question of what exactly he meant—namely, what his own conclusion was. Was he an atheist, as many think, or not?

Was Hume Ironic? And Was He an Atheist?

In my correspondence with Jeremy Fogel after our podcast, an interesting interpretive question arose. When he sent me a photo of the chapter, he added a note: “Beware of the concluding paragraph; it is, of course, entirely ironic…”

He apparently meant the paragraph at the end of the chapter:

Thus, in general, we may conclude that not only was the Christian religion accompanied by miracles at its inception, but that even today a prudent person could not believe in it without the aid of some miracle. Reason alone is not sufficient to convince us of its truth; and anyone who is moved to embrace it by faith is, in effect, conscious within himself of a continuing miracle which undermines all the principles of understanding and causes him to believe resolutely in what stands opposed to habit and experience.

At first glance David Hume is mocking believers here, and is ironically asserting that belief in miracles is itself a miracle, since it runs contrary to reason and experience (it has no basis). That is, in light of what is said in the chapter, we should reject that belief out of hand and adopt an atheistic stance.

But when I read the chapter, I concluded that this paragraph was not written ironically. As I understand it, Hume intends to say that one may (and perhaps ought to) believe, but this cannot be done by reason; rather, it must occur by “miraculous” means, i.e., in ways that do not comport with logic. If that is indeed his intent, then he aligns himself with an approach common among many believers according to which faith is above reason; therefore, although rational considerations cannot lead to it, it should nonetheless be adopted. On this reading, Hume himself remains a believer; he is not a scoffing, ironic unbeliever. In the chapter I analyzed, he merely wishes to sharpen the proper (miraculous) basis for faith and to reject the rational basis proposed for it.

I would note that I myself do not accept this thesis, since in my view faith is nothing but knowledge (see the introduction to The First Being and throughout). To my mind, talk about considerations “above reason,” and the like, is empty prattle. Still, I think this is what Hume set out to demonstrate throughout the chapter, and it is what he also says in that concluding paragraph, as we shall soon see. He apparently holds that faith is an alternative instrument to reason, observation, and intellect.

As noted, this is an interpretive dispute (about what Hume meant) and not a substantive one (about what is true). One may ask: why should I care what Hume meant? The interesting discussion is the substantive one—whether these arguments or others are true—regardless of who said them. But as I will show below, this historical discussion connects to my claims in the column and sheds light on many points regarding faith and its complex relation with reason.

The Opening of the Chapter: Dr. Tillotson’s Argument

The first indication for the interpretation I am proposing is the paragraph that opens the chapter, which presents the words of the clergyman Dr. Tillotson:

In the writings of Dr. Tillotson, one finds an argument against the real presence that is as concise, elegant, and sharp as can be advanced against a doctrine hardly worth the trouble of refuting. Everyone agrees, that learned clergyman argues, that the authority of Scripture, or of tradition, rests only upon the testimony of the apostles, who were eyewitnesses to the miracles wrought by our Saviour and thereby proved his divine mission. Consequently, the evidence we possess for the truth of the Christian religion is weaker than the evidence we possess for the truth of our senses; for even in the case of those earliest authors of our religion, their evidence was no stronger than the testimony of their senses, and it is clear that its force must necessarily have diminished when transmitted to their disciples; nor can anyone trust testimonies to the same degree that he can trust the immediate objects of his senses. But a weak proof can never override a strong one; therefore, even if the doctrine of the real presence were to appear explicitly in Scripture, its acceptance would stand in direct contradiction to the proper rules of reasoning. It contradicts the senses, while both Scripture and tradition, on which this doctrine is supposed to rest, are not supported by evidence as strong as that of the senses; all this when they are viewed as merely external evidences, such as do not penetrate the human heart by the immediate operation of the Holy Spirit.

Here, in brief, Hume’s own argument is presented (indeed, it is not clear to me what the entire chapter adds to Tillotson’s argument, though Hume presents it as a new argument of his). I would also note that Tillotson distinguishes between the dilemma regarding the transmitter himself (the one who initiated the tradition and passed it on) and the dilemma for his listeners—the recipients of his report or of the tradition. That is, Hume’s argument indeed attacks these two components of tradition, exactly as I argued against Jeremy Fogel (see the end of the first column).

In any case, note that Dr. Tillotson was a clergyman. He certainly did not seek to undermine faith and present knock-down arguments against it. He merely argues that faith should not be grounded on rational and empirical evidence (“external evidences… that do not penetrate the human heart”). In his view, “faith is the result of the immediate operation of the Holy Spirit,” meaning it is not the product of reason but of a religious-mystical mechanism above reason, precisely as I suggested for Hume’s own words. Hume says he is out to prove the same thing in a slightly different way (see above why, in my eyes, it is unclear what and how it differs). That is, he too does not come to reject faith but only to refute its rational-empirical basis.

Back to Hume Himself

So far we have seen Tillotson’s words, which are quoted at the beginning of the chapter. But the very same point appears in Hume himself (pp. 150–151), where, as he approaches the conclusion of the chapter, he writes:

But the method of reasoning here delivered will satisfy me the more, because I think it may serve to confound those dangerous friends or disguised enemies of the Christian religion who have undertaken to defend it by the principles of human reason. Our holy religion is founded on faith, not on reason; and to put it to the trial of reason, which it is by no means capable of standing, is the surest method of exposing it to danger. To make this point clearer, let us examine the miracles related in Scripture; and, that we may not lose ourselves in too wide a field, let us confine ourselves to those contained in the Pentateuch; and let us consider them, according to the principles of these pseudo-Christians, not as the word or testimony of God himself, but as the production of a mere human writer and historian…

It is evident that he himself believes in the sanctity of the books and the religious tradition. His aim is to undermine the attempt to ground them on empirical rationality. One who tries to do so is only ostensibly a friend of faith but is in fact an enemy in disguise. Note: the “enemy” here is not the believer but the one who tries to place faith on rational and intellectual foundations. He calls them “pseudo-Christians” (as opposed to himself, who is presumably a “true Christian”). To make the point, he examines the credibility of the book were we to assume it was written by a human author, and shows that there is no rational basis for accepting it. The conclusion, in my view, is not that the book is unreliable, but that its authority derives from its divine source (the Holy Spirit, not a human author) rather than from rational arguments. That, according to Hume, is what the “true Christian” believes, as opposed to the pseudo-Christians.

The Possibility of Esoteric Writing

Of course, one could claim that this is esoteric writing born of fear of the religious establishment. On this suggestion, Hume actually seeks to undermine faith and preach atheism, but he does so in an esoteric manner, as if under the guise of defending faith, so that he will not be persecuted. Indeed, Hume was a librarian and did not receive an academic appointment due to his views (which is to say that even if there was esotericism here, it did not exactly succeed).

Whether he means what he says in earnest or wrote it merely as a way to escape the religious establishment’s wrath cannot be decided from the text. But in any case, we are certainly not dealing with irony—at most with esotericism (had it been irony, he would not have escaped the religious establishment; on the contrary, they would have persecuted him all the more as one who mocks faith and believers).

Parallel Issues in Hume’s Thought

In my subsequent correspondence with Jeremy Fogel, I raised further examples that support the interpretation I offer here, drawn from other issues in Hume’s thought. More than once I have pointed out that Hume showed how consistent empiricism must lead to deep skepticism. One who accepts only the objects of experience as valid facts about the world will have to make do with very few facts and doubt the rest (anything not derived from direct observation). This immediately raises the question: does such doubt mean that one should not accept such facts, or may one accept them while being aware that they have no empirical basis?

For example, the question of the existence of an external world. Hume thought there is tension between our grasp of the world and the conclusion that it really exists “out there.” This conclusion cannot be inferred from observations alone. And yet, at the end of the section in Wikipedia dealing with Hume’s view on this matter, we find the following:

Hume’s position in his examination of the external world remains a disputed issue among his interpreters. Some argue that the inquiry shows that Hume, like Berkeley, was an idealist.[1] Others detect in the inquiry a skeptical spirit. In any case, few attribute to Hume the position that the objects of our cognition cease to exist when we are not perceiving them—particularly in light of the fact that Hume himself relies on the persistence of the objects of cognition in most of his investigations, and, some would say, even in the investigation of the external world itself.

So after he raises doubt and says that the matter lacks a solid empirical and rational basis, it is still possible that he accepted the existence of the external world. His claim is only that it lacks an empirical basis. He accepts it as a presupposition of reason, or as an intuitive grasp, and the like.

The same holds for causality and induction. Both were attacked by Hume, as I have noted more than once. His main claim is that these two notions lack an empirical basis. Now the question arises: as a result of these attacks, did Hume indeed not accept them as valid principles, or did he merely point out that they lack an empirical basis while still accepting them? It is hard for me to accept that Hume denied causal relations between events or the principle of causality and claimed there is only a temporal correlation between them. As I understand him, his aim was only to argue that they have no empirical basis and therefore we must be aware that they are a priori principles of reason. And perhaps we also ought not to use them in the scientific context—at least not as tools for discovering truth (but perhaps as tools for organizing information whose aim is not truth but the efficiency and effectiveness of our thinking about reality). The same goes for induction. There is no doubt that Hume himself uses the principle of induction. Some explain that for him this is only a habit of thought, not something he truly believes in. That is implausible, since habits of thought are not a basis for knowledge of the world—certainly not for an empiricist like Hume. Moreover, in the previous column I showed that he is so convinced by inductive inferences that it is obvious to him that any report of a miracle cannot be true, because it contradicts the laws of nature (which are, of course, the products of induction).

If so, both causality and induction—attacked by him as principles lacking an empirical basis—are nonetheless adopted by him as valid principles. It is quite possible that the very same stance characterizes Hume’s attitude toward miracles, revelation, and religious traditions. His claim is that we have no way to reach these by empirical or rational means; but that does not mean that miracles and tradition are false. We must be persuaded of them by way of faith and not by reason and observation. He opposes arguments that seek to give them a “scientific” rationalization, but not the beliefs themselves.

Here is another passage I found in that same Wikipedia entry:

There are two principal perspectives through which most interpreters analyze Hume’s thought within the political context of his time. On the one hand, one can see Hume as part of the Scottish Enlightenment and view his inquiries as an attack on theology and religion. Thus, some argue that one of Hume’s central aims was to show that we cannot use the inferential mechanisms by which we conduct ourselves in the world in order to learn about things beyond it, such as God or life after death. On the other hand, some interpret Hume as an opponent of the Enlightenment. They claim that Hume’s inquiries were intended to show that our knowledge of the world is not based on reason, as Enlightenment thinkers held, but on sentiment.

I would phrase it differently: his aim was to limit the validity of the tools of observation, logic, and reason—but not to say that there is necessarily nothing beyond them (somewhat like the early Wittgenstein, who ends the Tractatus with the immortal line: “Whereof one cannot speak, thereof one must be silent”). Our knowledge of the world may also be based on intuition; we should simply not pretend to root it in observation and science. This is precisely the interpretation I offered above. When he says that induction and causality are not products of observation, he means that we have no scientific basis for the supposition that the sun will also rise tomorrow—but not necessarily that we should doubt that it will rise.

It may also be that he was troubled by religious certainty, which can lead to zealotry and fundamentalism; thus, it was important to him to show that these are not scientific theses and that we must not grant them the certainty that science enjoys (Jeremy Fogel noted in our conversation that Edward Craig, the editor of the Routledge Encyclopedia of Philosophy, with whom he studied in his youth, understood it that way).

Two of My Remarks

This entire “historical” discussion has dealt with Hume’s thought and with clarifying his intentions. To conclude, I must add two remarks regarding my own view of this discussion.

First, in column 653, which I devoted to the topic of intuition, I explained that I do not accept the distinctions between feeling and reason. In my eyes, intuition is part of reason, and in fact every rational argument rests on it. Science too is replete with assumptions whose source is our intuition. What is opposed to recursive reason is not feeling but intuition. Feeling expresses subjective inner states and cannot serve as a basis for knowledge about the world, whereas intuition can. I have also argued in several places that intuition is a cognitive tool, not merely a thinking one. Therefore, the products of intuition can be acceptable even to empiricists, if they are willing to broaden the cognitive-observational canvas a bit.

Second, I do not accept the distinction between faith/the Holy Spirit and reason. Faith is a synonym for knowledge, since, in my view, we have no other instrument for reaching conclusions but reason. All the talk about instruments “above reason,” so common in religious discourse, is nonsense. Hume indeed intends this, in my estimation, but only because he misses the fact that even if some result does not arise from observation, that does not mean there is no way to justify it rationally and logically. Such extreme empiricism is naïve. As I noted in the previous column, inference by way of induction is itself nothing but the application of an a priori presupposition of reason (intuition). To regard the conclusions of induction (such as the laws of nature) as empirical results is naïve at best. Yet on the other hand, they are not subjective feelings either. They are products of intuition, which we use to process and analyze observational facts and to infer general conclusions from them.

Therefore, when I suggested that Hume’s intent is to say there is no rational way to arrive at faith in traditions and miracles, for my part this should be translated thus: there is no empirical and scientific way to arrive at it. But intuition too is part of reason, and it is in no way related to the Holy Spirit, prophecy, or mysticism.

In other words, I personally identify with Jeremy Fogel’s Hume—the one who sees no alternative to rational inferences (even if, in my view, these are not necessarily empirical inferences). Only I am not at all sure that this is indeed what Hume meant in that chapter. People who hold various beliefs and discover that they cannot anchor them in observations or scientific inferences from observations feel a need to retreat into subjectivity (Rabbi Shagar’s “circle of differences”). But that is a mistake. Intuition is an entirely rational tool that we use in every domain, including science, and it can also ground faith. In short, Hume thought that to be a believer one must depart from rational thinking, whereas in my view one can certainly be a believer and remain rational.

[1] I would only note that Berkeley did not accept the existence of the external world even at the very moments when we are perceiving it, not merely afterward as with Hume.