Miracle Discourse never ends on substack. This post had been stirring in my brain for the last few months, but it wasn’t until Joe Schmid announced his conversion to Catholicism that I was able to find a uniting theme.
Schmid’s conversion is relevant here because arguments for miracles were decisive for his conversion. That was disappointing to me because I consider many of the arguments and evidence for popular Catholic miracles really really bad.
Seeing an agnostic convert for those reasons is kind of like watching your political party lose a wave election for reasons outside of their control (like the predecessor nuked the economy or something before the term started). It’s a gut-punch and a wakeup call, but it’s not the kind of wakeup call where you throw your entire worldview out the window, so much as you change your communication strategy.
I will not be addressing anything else Schmid-related. I’m happy for him so long as he’s happy. At the risk of sounding like a hippy dippy liberal, I don’t care what he believes so long as he’s kind to other people.
My primary motivation for this post was, as usual, education. I wanted to give a clear and account of the skeptical position, especially given that there’s so much misinformation and misunderstanding on the internet about what educated skeptics, Humeans, and non-theists believe about miracles.1
Though I’m agnostic on whether miracles can happen,2 for reasons I’ll get into in this post, I’m confident they don’t. At the very least, if God exists and works miracles, I would hope he wants the faithful to believe in miracles for good reasons, and I don’t think we have any at the moment.
A Recap Of Humean Philosophy
I’m going to start this post off by referencing and summarizing some of my previous writing. It’s important table setting.
One of my more recent posts I published before going on a semi-hiatus outlined why testimony was practically not evidence for low-probability events. I still think it’s a banger, and haven’t heard any substantive negative feedback about it.3 Below that post are some additional posts on Hume’s essay Of Miracles
Testimonial Claims Are Practically Not Evidence For Extremely Improbable Events
In my last post, I argued that claims, properly understood, are not evidence. Admittedly, my definition of “claim” was very narrow. What I really wanted to demonstrate in that post was a few things:
The Internet Is Wrong About David Hume
This post is too long for email, so please click on the title to read the whole thing!
How To Understand David Hume's Argument On Miracles
This post is too long for email. Please click on the title to read the entire thing in browser!
"Hume Defines Miracles Out of Existence” and “It’s a Circular Argument”
For the first part of this series, I’m lumping these two criticisms of Hume together because they’re basically arguing the same thing: Hume is committing a logical fallacy, begging the question, and therefore his argument is invalid. Specifically, Hume either defines miracles as impossible (and therefore they cannot happen) or that he presupposes (someh…
C.S. Lewis Misunderstood and Misrepresented David Hume
In my previous post, I dispelled the argument that Hume’s argument in Of Miracles is circular. You can find a summary in this footnote. C.S. Lewis likely popularized that argument in chapter 13 of his book Miracles.
Below is a rough recap of the main ideas of these posts. I’ve put them in numbered format because this post is already pretty long and I don’t think we need to overcomplicate things further. If you’ve already read/remember/understand my previous posts, you can skip to the next section.
Key Ideas
Testimonial evidence is by its nature inferior evidence to non-testimonial evidence because testimonial evidence requires additional inferences about witnesses on top of the data they report.
For Hume (and me) natural laws are observed patterns in the world that have not been observed to deviate. When I or Hume say a law of nature hasn’t been violated, we mean no one has credibly seen it deviate from the observed pattern.
People confuse this to think Hume is defining a miracle out of existence or proposing a circular argument, but that’s not what he’s doing. Rather his model of probability is implicit in his writing (and really only comes to full light when someone like Bill Vanderburgh explains it to you separately).
In this way, a miracle being “by definition” the most improbable event isn’t Hume just describing the definition (though I think reasonable people would agree with his definition), but saying “If a law of nature is the most universally attested and frequent pattern with no contravening observation, and someone claims to observe a contravening event, that claim would consequentially be one of the most unlikely observations imaginable to be true.”
I’ll admit that #4 is confusing and nuanced, but to get at the core of this you have to understand Hume’s formulation of probability, how it contrasts contemporary probability, and how it plays into his vocabulary. Namely:
Probability is a weight, not a calculation, where one event is tallied and compared against another also-tallied event.
The events that have no counterweight and are observed in extremely high frequencies are natural laws.
Those that have counterweights are considered mere probabilities,
Probabilities that have no observation in favor but with great observation against are considered impossibilities.
Hence, the contravening evidence against natural laws is considered a posteriori impossible (as a matter of fact, it could otherwise be different and we can revise this assessment with new evidence).
Laws are not rendered useless or meaningless with single observations contravening them. They’re just downgraded to very high and reliable probabilities. Similarly, impossibilities are not upgraded to likely with single observations, but just mere (very low) probabilities.
Hume believed that religious miracle testimony do not meet the standard of probabilities because of the universally low quality.
In all, this means that a wise person is always justified in rejecting that a miracle happened when only presented with eye-witness testimony. Even if a witness is highly credible (or if there are multiple credible witnesses), the probability of error by the testifiers is higher than the probability of the laws of nature being violated.4
Now, that doesn’t mean we a priori object to novel observations of spectacular events merely because all the evidence we have is testimony. If many people say they saw something weird, something weird may well have happened. It’s just unlikely that the weird thing is a religious miracle because of the nature of religious testimony. We’ll talk more about this in this post!
Keep in mind that Hume’s argument is an epistemological one, not an ontological one. If your ontology permits the existence of miracles, then it can be rational to believe in them. But for Hume, this is a trivial assertion. His argument is examining whether neutral wise observers should believe in miracles based on testimony (and thus whether it’s wise to believe in religion, given the only evidence we have for the foundations of modern religion is testimony). Saying “well if you presume the religion is true and that miracles do happen then it works out” is question-begging and circular logic.
As this relates to my posts on testimony not being evidence for improbable events, I’m basically echoing Hume’s belief (though I don’t think he says it directly or clearly) that we should proportion our credence in eye witness testimony to the baseline probability of an event.
In this way, eye witness testimony for everyday events is credible because those events are probable (or at least not improbable). At the same time, that testimony is useless when it comes to near impossible events or even rare ones.
If you audit your own behavior when it comes to improbable or rare events with severe life circumstances (not just miracles) you’d find that you also don’t trust mere testimony. That’s because we structure our lives over what is true and what is probable, and learning of improbable consequential truths usually entails restructuring our life or making radical financial decisions. Before making these changes, we demand non-testimonial evidence because we intuitively understand the high probability of human error and deception.
And so with miracles, we demand more than testimony to be convinced.
Miracle Explanations Aren’t Like Other Explanations
For non-believers, skeptics, naturalists, Humeans, and other miracle-skeptics, miracle claims are inherently suspicious because they assert causality where we have little evidence to support it, often without investigation.
Let’s explain this with an illustration. Think of two people performing a spectacular event: a real miracle worker (Jesus Christ himself perhaps) and the world’s best magician. If both of them performed a feat where they similarly appeared to walk on water and you didn’t know which person was which, both of their walks would look identical. You couldn’t tell them apart!
In this way, when we experience a novel spectacular event, by definition, we do not know what we are observing. For these events, whether they be miracles or magic tricks, we do not see the underlying causal forces creating the appearances we are observing. They are hidden from us, as we see the consequence of the miracle/magic trick, but not the cause. That’s what makes it a miracle or magic trick!
I call this problem the problem of auxiliary or secondary causality. I don’t think this is something Hume talks about directly (outside of general problems of inferring causality) but Hume scholars like Vanderburgh allude to it in their work.
The problem of secondary causality is this: When we see an extraordinary event, whether it be the sun moving weirdly or a diseased person healing unexpectedly, we are seeing the effects of that event, not the cause. We are completely unfamiliar with the cause.
This, in contrast with ordinary events. When you see a car crash, you see a car going too fast and hitting another car. You have an understanding for the plausible causes for this event, from human carelessness and intoxication to mechanical failure. You also have an understanding of how to investigate or eliminate alternative candidate hypotheses. As a result of abundant experience with cars, drivers, and the causal forces that influence them, we know how to evaluate those candidate hypotheses, affirming or eliminating them.
You cannot say the same of supernatural causation.
The problem for miracles and supernatural events is that they resist the same level of explanation as natural events. If our car example were judged miraculous, people would recall and explain the event as: “We saw a car crash, God caused it, and no further explanation is needed.” This is an odd way to draw a conclusion about anything, but because it’s judged miraculous, it’s considered permissible. At the risk of sounding like a cringe 2010s New Atheist, “God did it” is not much of an explanation!
In this way, causal explanations need to be plausible, well explained, and better fit the data than alternative hypotheses. Merely reporting the event as observed and implying “no additional explanation needed” eliminates all aspects of explanation.
To use a marginally more concrete example, there’s a difference between saying “God made the host bleed” explaining how God made the host bleed, and explaining how God made the host bleed using high quality evidence, methodologies, and measurement tools, while also eliminating alternative hypotheses with that same data.
In my experience, miracle proponents almost always lack high quality evidence or tools of measurement, and are generally unconvincing when eliminating naturalistic explanations.5 They may have a quasi-plausible sounding story about how God did something, but no good evidence to support it.
Instead of finding this high quality evidence, miracle proponents will use the sophisticated-surrounding rhetoric of hypothesis testing. As I’ve written before, this rhetoric is misleading because, at best, it’s inferior to other evidentiary methodologies. At worst, it’s outright deceptive and abused. After all, when we use hypothesis testing methodologies in science or our everyday life, we’re applying them to data with which we have abundant experience of its causal relationships. What’s more, we’re also using it for events that are highly probable or not improbable. The same cannot be said of extremely improbable miracles and supernatural causation, where we have scarce familiarity.
Miracle proponents seem to think hypothesis testing and conditional probabilities are incompatible with Hume’s argument, but they’re wrong.6 If anything, Hume’s argument accurately frames these methods, as most non-philosophy-of-religion uses of them are done in a context of familiar causal relationships and non-improbable events. In this way, miracle proponents are misusing these methods because they’re taking them outside of the context of highly probable and familiar cases.
And so, the miracle proponent will use tactics like asking a question around the lines of “Does this evidence best fit the supernatural or natural hypothesis?” pretending that baseline probability of either hypothesis is similar.
Such questions also presume that supernatural hypotheses aren’t catastrophically under-defined. They may sound legitimate, but compared to naturalistic hypotheses, they are much less detailed. It’s an impressive rhetorical move (and I want to reiterate that I don’t believe it’s intentional or bad faith) by believers, but it doesn’t get us closer to the truth.
This note below does a good job of explaining this asymmetry with magic as a stand-in for supernatural explanations. (To be clear, I don’t think Charlie is a skeptic, he just knocked this explanation out of the park)
For the skeptic, it appears as if miracle proponents are playing word games. Naturalistic explanations are held to higher standards, while supernatural explanations seem to have no standards. It’s easy to call a competing hypothesis ad hoc, unparsimonious, and all these others bad things when your hypothesis doesn’t say anything at all.
Correcting a Humean Confusion
Here I want to clear up a misconception about the Humean position on novel spectacular events (that is, amazing events you’ve never seen before).
Non-skeptics (and unfortunately, some skeptics) erroneously believe that the Humean wouldn’t believe anything spectacular based on testimonial experience. Or, if a credible novel observation undermined the Humean’s understanding of the laws of nature, they would disbelieve it. Both of these beliefs are wrong.
For a Humean, eye witness testimony for something spectacular is good evidence for something spectacular, it’s just not evidence for something miraculous in origin because we do not observe miraculous causal origins. 7
We can formulate reliable natural probabilities because we have abundant evidence of conjunctive natural events. At the same time, we cannot formulate supernatural or miraculous probabilities because we don’t have sufficient evidence or experience for those events.
Even if you believe miracles happen, they are exceptionally rare, non-replicable, and their causes (beyond “God did it”) are mysterious. Consequently, it’s impossible to identify them with high confidence if and when they do happen, especially because miracles rarely happen in the same way (a healing miracle is different from a sun miracle).
Put another way: I’ve seen plenty of billiard balls move, make contact with other balls that conjunctively move, and so I can formulate reliable probabilities and expectations about those events and similar events. In contrast, I’ve never seen someone miraculously move a billiard ball, so I can’t formulate any reliable probabilities or expectations of what that looks like.
And so, the Humean or skeptical position is that these spectacular observations are not evidence of supernatural causation. Hume did not deny that spectacular or novel events happen, just that it’s unjustified to infer that their cause is supernatural or religious instead of natural.
It’s thus no major concession to say extraordinary things happen when presented with testimonial evidence for something extraordinary. It’s also not a major concession to observe a purported law of nature being undermined. Rather, at worst, the Humean merely downgrades the law of nature to a very strong probability. No harm, no foul.8
In all, saying something spectacular happened is different from saying something supernatural happened; downgrading a law of nature to merely an extremely strong probability is much different from saying a supernatural force interceded and overruled the law. On those two points, the Humean will always object because there is not sufficient evidence that something supernatural happened (or that something natural yet currently misunderstood didn’t happen) or that something supernatural “interceded” or “overruled” the law.
We just don’t know what those causal forces (supernaturalism, a force overruling the laws of nature) look like.
Hume’s Maxim Is Just Correct
As a consequence of all of this, I instinctually shrug off pretty much every miracle claim. I think purported miracle evidence is usually just evidence for the spectacular, and upon further examination, that spectacular event is either clearly natural or inconclusive in origin.9
The primary reason I don’t argue about specific miracles, though, is that they typically only have testimonial evidence supporting them, and I think Hume’s Maxim10 is simply correct.
That is, when witnesses claim to observe a miracle, the two explanations (human error and the uniformity of nature vs. human correctness and the violation of the uniformity of nature) are mutually exclusive, and so we should weigh the probabilities. From there, it’s exponentially more likely that humans have made a mistake in their perception than it is that the laws of nature no longer function as universally observed.11
What’s more, even if you assume a spectacular observation is accurate, it’s unlikely to be supernatural in origin, for at least two reasons that even supernatural believers should affirm:
Natural causes are exponentially more likely than supernatural causes (I don’t feel like I have to elaborate on this?).
In the history of unexplained events that have been investigated, the overwhelming majority (being generous to supernaturalists) have turned out to have a natural origin; why think the next one would be different?
I understand that some may object to my general dismissal, believing that a sufficient number of witnesses could make a miracle claim probable. Still, I disagree with this for reasons I expressed in a note while restacking a great post by Analytic/synthetic skeptic. I borrowed from an argument advanced by the Humean philosopher Arif Ahmed:
Long story short, we can identify the circumstances when otherwise reliable tools of measurement and “attestation” are clearly concurrently failing (even if the concurrent failure is itself improbable). I apply that same reasoning to otherwise reliable eyewitness accounts.12 Humans in groups are much more likely to concurrently err than the laws of nature are to deviate.13
I understand if theists and general miracle believers think I’m stubborn and pig-headed for thinking this way, but at the end of the day, I’m just acting the most wise/rationally I can.
Wrapping Up
Originally, this post continued on for another few thousand words, diving into the psychology of miracle believers and skeptics. I was going to talk about how some believers insistence on the necessity of believing in bad miracle claims created an interesting problem of evil. But upon uploading this into substack, I realized that it read like a completely different post, with a different overall tone, and so this was a good time to pause. I’ll upload that at a later date.
For now, the key takeaway from this post is that skeptics disbelieve miracles because they believe that:
Supernatural explanations are unsatisfactory, as they lack explanatory power relative to natural ones,
Hume’s maxim is correct,
We don’t know what supernatural causation looks like, so claiming it as an explanation is unjustified,
Eyewitness testimony for the miraculous is really just evidence for the spectacular,
Sufficient investigation into the spectacular is exponentially more likely to reveal natural causes.
It’s not an exaggeration to say that, if you search on YouTube “Hume’s Argument For Miracles,” quite literally 60-80% of the videos are Christian Apologists getting Hume wrong and spreading misinformation. Seriously.
I’m a skeptic, I’m agnostic on many questions.
Though it did activate some ad hominem attacks from detractors, which warranted swift blocks.
If you express this formulation in Bayesian terms, it’s easy to get it technically wrong, for nuanced reasons I will eventually write about. Keep in mind, Hume is not a Bayesian. But with some slight alterations “Hume’s maxim” does work in Bayesian terms, and stays in line with the spirit of Hume’s original formulation
Curiously, they try to eliminate improbable naturalistic causal explanations, when the effect itself was improbable. We would expect improbable causes to improbable events!
And yes that includes John Earman, who was wrong about Hume.
In fact, you should believe something happened and investigate accordingly! Hume said as much in Of Miracles:
I beg the limitations here made may be remarked, when I say, that a miracle can never be proved, so as to be the foundation of a system of religion. For I own, that otherwise, there may possibly be miracles, or violations of the usual course of nature, of such a kind as to admit of proof from human testimony; though, perhaps, it will be impossible to find any such in all the records of history. Thus, suppose, all authors, in all languages, agree, that, from the first of January 1600, there was a total darkness over the whole earth for eight days: Suppose that the tradition of this extraordinary event is still strong and lively among the people: That all travellers, who return from foreign countries, bring us accounts of the same tradition, without the least variation or contradiction: It is evident, that our present philosophers, instead of doubting the fact, ought to receive it as certain, and ought to search for the causes whence it might be derived. The decay, corruption, and dissolution of nature, is an event rendered probable by so many analogies, that any phænomenon, which seems to have a tendency towards that catastrophe, comes within the reach of human testimony, if that testimony be very extensive and uniform.
In fact, I’d argue this happens all the time, as the nature of Humean natural laws are just very strong probabilities. For example, it’s impossible for a human to fly. But a human can fly in an airplane or on a glider. The reality of the latter somewhat invalidates the lawhood of the former, but at the same time it doesn’t invalidate the law that men cannot fly naturally or unaided. This is why all purported natural laws need to be extremely specific. For the most part, they are in science, which is why science literate people are extremely skeptical of outright violations of the laws of physics without significant caveat.
And even then, I’m being overly generous to supernaturalist explanations, because as I said above, I don’t believe there are high quality, satisfactory supernatural explanations.
“That no testimony is sufficient to establish a miracle, unless the testimony be of such a kind, that its falsehood would be more miraculous, than the fact, which it endeavours to establish; and even in that case there is a mutual destruction of arguments, and the superior only gives us an assurance suitable to that degree of force, which remains, after deducting the inferior.”
Some may say that this is a false dichotomy, that it could be the case that God overruled the laws of nature. I would reiterate that as an observer with no a priori commitment to religious belief, this argument just doesn’t work. After all, we’re trying to use miracles to prove God, so using God to prove miracles would be circular. Once again, we are having an epistemological debate, not an ontological one.
Especially considering human memory at its best is much less reliable than a digital thermometer at taking temperature
It’s worth mentioning that when miracles supposedly happen, more than one “law” is probably being violated. Take Moses parting the red sea so the Israelites could cross it. He’s not just violating gravity with the sea parting, but there’s probably something else going on with the sea floor so the people could cross, as it’s not naturally a hiking trail. You can come up with other examples for other miracles if you think about it long enough.
Interesting post; I'll be interested to see if any of the miracle testifiers respond. I do have one thought that I'm interested to see what you think. So, basically, Hume's thought is that natural laws are just events that have never been reliably reported to deviate. Now, for the sake of the thought, let's presume a group of witnesses did reliably report a deviation in the natural law (they all caught the event on separate cameras or something without collaboration). Wouldn't a reliable deviation from the natural law make it likely that something supernatural happened? After all, if what is natural is what has always happened, then would it not be the case that an observation consistent with the data would very likely be natural, whereas an observation that deviates from the data would very likely be something not natural?
In other words, it's not just the effect that has never deviated but also the cause, so if the effect is something that deviates, the cause is something that deviates, and because what is natural has always acted the same way, when something acts differently it suggests that the cause is different, and this is unlikely to be in a natural manner because all our evidence of natural causes act a certain way, so by virtue of the event having an unlikely effect its improbable that its cause would be natural since natural causes have never acted this way before.
I could not figure out how to say what I was trying to say, so I hope that makes sense.
“In all, this means that a wise person is always justified in rejecting that a miracle happened when only presented with eye-witness testimony.”
I think you clarify this later, but you would not hold that it’s always justified in rejecting that a *spectacular* event happened (e.g., someone levitating unaided) when only presented with eye-witness testimony. Right?
Good post!