Definitely read Hume himself! I've tried to read many old philosophy texts and mostly failed (unless they're modern translations), but Hume was the exception. He writes exceptionally clearly (and surprisingly modern) such that even I, a non-native speaker, could read it. It also helps that his arguments are brilliant (one of my all time favorite philosophers) which motivates you to read more.
I'm also not sure Hume was an atheist like people believe. We will never know for sure since you couldn't be one publicly at that time, but he at least said he wasn't and from what I remember from reading "The Natural History of Religion" and "Dialogues Concerning Natural Religion" some years ago, his view seemed to be neither typically atheist nor typically theist.
He does attack many facets of religion (people focus on his argument against miracles but I think that e.g. his attacks on the afterlife and the charge that nobody actually believes in it, is much more biting). However, what I took away from his dialogue between Philo and Cleanthes (Demea is clearly not Hume's real position, which I think is mostly Philo with a bit of Cleanthes) is a kind of naturalistic pantheism. Of course, if he was an atheist this would be in line with expectations since this was about as much as you could get away with at the time (so who knows), but at least the text itself seemed to point towards pantheistic naturalism (within a broader framework of agnosticism).
As a modern communication guy, I think he’s definitely a better communicator than his contemporaries, but his style is still a little hard for me. Though reading analyses of essays like Of Miracles definitely made it more clear as I was reading it.
I actually agree with you wrt his religious beliefs: I think he was so skeptical that he was an agnostic. Your reading of the Dialogues is in line with Simon Blackburns and I think at least one other person I’ve read. I think I’ll write a post about it at some point, because I think people reading him as an atheist really, truly lose out on deeper points he’s making.
With regard to the specifics of his dialogue, I think his point is that we can’t really prove what the theist is trying to say, even if we can agree with it (I’m so murky on it because I’m going from memory of what blackburn described it, but will obviously get back to it sooner or later). But yeah, I generally agree that his views were different and worth exploring. Which is why I’m doing it!
"With all of that context out of the way, I want it clear that people getting Hume wrong is not a mark of unintelligence or dishonesty, so much as it’s just going with the flow or META of philosophy posting in 2025. It’s a meme at this point."
Unfortunately, this is not a new thing restricted to 2025. People have been doing it for about 250 years now.
I find the discourse around Of Miracles to be one of the most frustrating parts of religion debates. Not even for the reasons brought up here (though strawmanning Hume also sucks), but because the arguments surrounding it always start up when apologists bring him up for no reason other than that it's easier to dunk on their interpretation of his argument than on the actual arguments being put before them. Every time you make an argument that even vaguely resembles Hume's case in Of Miracles (and the resemblance can be really, really vague), there are even odds or better that an apologist will step in and say, "Oh, that's just Hume's argument from Of Miracles. Don't you know that argument has been completely debunked and is stupid for all these reasons I'm about to list?" They won't even address your version of the argument because they assume that if they can refute Hume's version, that automatically refutes any argument that even slightly resembles it. This then leads to endless debates over whether they've really refuted Hume's case, but it's all irrelevant to the original discussion. Even if the strawman interpretation of Hume was more accurate to what Hume originally meant than the steelman interpretation, it wouldn't matter because apologists still have to address the steelman interpretation (or provide some evidence that isn't based on testimony) if they want to defend miracle claims. And even if both the strawman and steelman interpretations of Hume are bad arguments, they still have to address other arguments against trusting the evidence for miracle claims, rather than deflect by talking about Hume every time someone makes one.
Also, I liked the point about how everyone already agrees with Hume's actual argument, despite the endless attempts to dismiss him by debunking a strawman. It reminds me of the apologetic attempts to show that the saying, "Extraordinary claims require extraordinary evidence," is wrong, despite the fact that everyone, including religious believers, implicitly understands it to be true when evaluating evidence for anything else. In this case, rather than strawmanning a famous philosopher, these arguments usually just use incredibly poor logic - typically, the apologist points to some claim with an extremely low prior probability, like, "The winning numbers for the lottery will be [some exact sequence]," says that the claim is "extraordinary" on the basis of its low prior, then points out a piece of evidence with an extremely high likelihood ratio that would get you to believe the extraordinary claim (e.g., someone reporting the winning numbers and reading out the exact sequence), and somehow not realizing that by their own definition of "extraordinary," the evidence presented would count as extraordinary evidence. These arguments are typically followed up with all the same condescension and overconfidence (acting as though their terrible arguments have completely destroyed the common saying, and that anyone who still uses it is therefore an unsophisticated rube), all while still using the actual principle the saying points out in their evaluations of evidence not in support of their religion, since the actual principle is obviously correct.
Nice piece. Poor Hume, the greatest philosopher to write in the English language, and persistently misunderstood since 1739. He famously complained that _A Treatise of Human Nature_ “Fell dead-born from the press, without reaching such distinction as even to excite a murmur among the zealots.” So he wrote his Enquiries to explain things all over again. I’ve published on Nietzsche, and that guy is misunderstood even worse!
You’re 100% right about this being about the epistemic issue of warrant for beliefs in miracles and not the metaphysical issue of whether reality has any miracles in it. However, I think you err in saying Hume’s only concern here is to argue against the warrant of testimony-based miracle beliefs. Non-negligible portions of the essay are directly about warrant for other, non-testimonial, beliefs, including those based on experience and experiment. “On Testimony of Miracles” wouldn’t be a more informative title.
I think your assessment is 100% accurate! As a matter of fact, I almost wrote something like you say in the last sentence. The first section is basically an outline of the probability assessment through an empirical lens, culminating (in what I would argue) as a pragmatic summary of that argument (which is confused as his argument). The bulk of Hume's essay is really about what religious beliefs do to (see: corrupts) testimony, and the criteria by which we evaluate reliable testimony, especially as it pertains to novel events.
The feedback I'm getting on this inspires me to dive deeper. I think my post here is a good run down about how people get Hume wrong, but an essay on the actual in length, in detail argument of Of Miracles would be valuable
The tendency to misunderstand Hume is like the tendency to misunderstand his fellow Skeptic, Sextus Empiricus. Not only is there a language problem (archaic English/translating from ancient Greek), but there's also a worldview problem. Dogmatism entails a certain "black and white" way of seeing the world: the world *is* this way or it *is not* this way. The Skeptic instead talks about the world *appearing* in a certain way, and is parsimonious with declaring is and is not. This way of thinking about and talking about things is confusing to the Dogmatist, who instinctively thinks the Skeptic is making kinds of claims different from the ones the Skeptic is actually making.
If the argument is as you have described it here, I honestly don’t think the difference between testimony and other forms of evidence is that important. The basic argument appears to be that given the broad experience behind the laws of nature, we would need extraordinary evidence to believe in miracle and as a factual matter while such evidence could be provided in theory, it has not actually been provided so far. In fact, all such evidence would have to be strong enough to more than cancel out the evidence against miracles and even so the resulting update would be only as strong as the update once it has been cancelled against the update from the evidence against miracles. Even if the original essay restricts the argument to testimony, this seems equally applicable to other forms of evidence. Assuming I have understood your understanding of the argument correctly. I think the bigger misunderstanding if you’re understanding is correct. Is that most people think he’s arguing that it is impossible even in principal to be persuaded of a miracle when he is only arguing that existing evidence is not sufficient to believe in a miracle.
So I think you're a little wrong here. Namely, he seems completely agnostic or silent on whether it's reasonable to believe a miracle if you observe it yourself! I said this in a footnote, but people sometimes I think equate a Christopher Hitchens point about thinking "maybe I'm mistaken" is more likely than you directly observing a miracle, but as far as I know, Hume doesn't argue this!
In this way, the primary thrust of Hume's argument is empirical epistemology: always trust the evidence you directly observe over what you are just told via testimony. One can expand upon the argument in the way you have, or have a separate conversation about whether empiricism is true, but I think the narrow claim is correct. At the very least descriptively, we don't find evidence or data compelling unless we experience it someway.
I do not think we actually disagree because I was arguing that the distinction is not one that is actually important. Not that the actual historical Hume thought that the distinction was not important. I find it entirely possible that he would respond as you responded, although I don’t personally find that convincing since for almost all the knowledge I actually possess I purely rely on the testimony of other individuals and yet I am much more confident that the testimony that Physicists have provided regarding the fact that the universe is expanding then I am regarding my memory of what I eat last Monday. Obviously, the reliability of testimony depends on the context like any other method of knowledge acquisition, but I just think that there is no good argument for categorically distinguishing between testimony and other types of evidence. This is not an argument about what a historical person believed, but rather my view of what the argument actually implies if taken to its logical conclusion.
I don’t think Hume’s argument that testimonies cannot justify beliefs about miracles is successful , I think it largely depends on one’s prior commitments, for example if you are a theist, then it won’t be unreasonable for you to believe that a miracle occurred based on testimonial evidence. A person who is a naturalist will not trust such testimonies.
Yes, I think I talk/mention this in the footnotes, but Hume definitely equate a “reasonable person” to mean someone who agrees with him wrt empiricism (i.e those who believe the most rational interpretation of evidence is the kinds we can directly observe). I think people are better suited to attacking that aspect of the argument, which is somewhat implicit, but that’s different from most arguments against Hume that I’m talking about.
But it's not applied consistently and thus not logical. The Christian evangelical will believe a testimonial claim about divine intervention by a fellow Christian but not fron a Muslim or Janseian.
As an addendum to this, it's worth mentioning that Hume is taking aim at natural religion and the idea that one can come to believe religions established on miracle claims coming from an unbiased perspective that doesn't presuppose its truth or reliability.
I think that testimony, paired with a rigorous systematic elimination of conceivable natural explanations, is very powerful evidence in favor of supernaturalism. I think most skeptics are way too dismissive of anomalous experiences of ordinary people, which are copious.
Yea, I don't mean to assail Hume's point. I also find miracle claims from ancient manuscripts (such as that of the Bible) very weak evidence by themselves. (That being said, I find 400-year old extensive reports of miracles by authorities hostile to the one performing said miracles, such as the case of the Inquisition examining the levitation of St. Joseph of Cupertino, quite strong. For commentary on the documentation of this case, I recommend They Flew: A History of the Impossible by Carlos Eire.) I just wanted to express my opinion on something tangentially related to your post, but still important overall, in my opinion.
No worries! My bluntness is in being defensive against being dragged into an argument that isn't relevant to the post. Having said that, though I think Hume's point is rather solid, I think there are possibly ways around it, to justify "rationally" believing in religious miracles. I just don't think people attack him in the correct way to do it (i.e. they should attack his empirical assumptions, not against a straw man belief that he believes miracles can't happen)
The trouble with this is that a "rigorous systematic elimination of conceivable natural explanations" is easier said than done, and I don't think it ever has been done despite innumerable attempts. People always try to do it in arguments about miracle claims, but this typically takes the form of someone listing all the obvious naturalistic explanations they could think of and coming up with some arguments against each one. The arguments generally don't show each option to be impossible, but only give some reasons that each would be unlikely to happen. This is not very convincing to skeptics because it's unsurprising that in the entire course of human history, there would be some unlikely events that get misinterpreted as miracles due to the unlikeliness of their true causes (while it's unlikely for any individual event to be like this, over a long enough stretch of history, it's nearly certain that some events like this will happen). Furthermore, these types of argument frequently leave out plausible options and rarely provide any reason to think that the list of possible natural explanations is exhaustive.
You did it mate, a post on miracles I actually found interesting!
Definitely read Hume himself! I've tried to read many old philosophy texts and mostly failed (unless they're modern translations), but Hume was the exception. He writes exceptionally clearly (and surprisingly modern) such that even I, a non-native speaker, could read it. It also helps that his arguments are brilliant (one of my all time favorite philosophers) which motivates you to read more.
I'm also not sure Hume was an atheist like people believe. We will never know for sure since you couldn't be one publicly at that time, but he at least said he wasn't and from what I remember from reading "The Natural History of Religion" and "Dialogues Concerning Natural Religion" some years ago, his view seemed to be neither typically atheist nor typically theist.
He does attack many facets of religion (people focus on his argument against miracles but I think that e.g. his attacks on the afterlife and the charge that nobody actually believes in it, is much more biting). However, what I took away from his dialogue between Philo and Cleanthes (Demea is clearly not Hume's real position, which I think is mostly Philo with a bit of Cleanthes) is a kind of naturalistic pantheism. Of course, if he was an atheist this would be in line with expectations since this was about as much as you could get away with at the time (so who knows), but at least the text itself seemed to point towards pantheistic naturalism (within a broader framework of agnosticism).
As a modern communication guy, I think he’s definitely a better communicator than his contemporaries, but his style is still a little hard for me. Though reading analyses of essays like Of Miracles definitely made it more clear as I was reading it.
I actually agree with you wrt his religious beliefs: I think he was so skeptical that he was an agnostic. Your reading of the Dialogues is in line with Simon Blackburns and I think at least one other person I’ve read. I think I’ll write a post about it at some point, because I think people reading him as an atheist really, truly lose out on deeper points he’s making.
With regard to the specifics of his dialogue, I think his point is that we can’t really prove what the theist is trying to say, even if we can agree with it (I’m so murky on it because I’m going from memory of what blackburn described it, but will obviously get back to it sooner or later). But yeah, I generally agree that his views were different and worth exploring. Which is why I’m doing it!
"With all of that context out of the way, I want it clear that people getting Hume wrong is not a mark of unintelligence or dishonesty, so much as it’s just going with the flow or META of philosophy posting in 2025. It’s a meme at this point."
Unfortunately, this is not a new thing restricted to 2025. People have been doing it for about 250 years now.
I find the discourse around Of Miracles to be one of the most frustrating parts of religion debates. Not even for the reasons brought up here (though strawmanning Hume also sucks), but because the arguments surrounding it always start up when apologists bring him up for no reason other than that it's easier to dunk on their interpretation of his argument than on the actual arguments being put before them. Every time you make an argument that even vaguely resembles Hume's case in Of Miracles (and the resemblance can be really, really vague), there are even odds or better that an apologist will step in and say, "Oh, that's just Hume's argument from Of Miracles. Don't you know that argument has been completely debunked and is stupid for all these reasons I'm about to list?" They won't even address your version of the argument because they assume that if they can refute Hume's version, that automatically refutes any argument that even slightly resembles it. This then leads to endless debates over whether they've really refuted Hume's case, but it's all irrelevant to the original discussion. Even if the strawman interpretation of Hume was more accurate to what Hume originally meant than the steelman interpretation, it wouldn't matter because apologists still have to address the steelman interpretation (or provide some evidence that isn't based on testimony) if they want to defend miracle claims. And even if both the strawman and steelman interpretations of Hume are bad arguments, they still have to address other arguments against trusting the evidence for miracle claims, rather than deflect by talking about Hume every time someone makes one.
Also, I liked the point about how everyone already agrees with Hume's actual argument, despite the endless attempts to dismiss him by debunking a strawman. It reminds me of the apologetic attempts to show that the saying, "Extraordinary claims require extraordinary evidence," is wrong, despite the fact that everyone, including religious believers, implicitly understands it to be true when evaluating evidence for anything else. In this case, rather than strawmanning a famous philosopher, these arguments usually just use incredibly poor logic - typically, the apologist points to some claim with an extremely low prior probability, like, "The winning numbers for the lottery will be [some exact sequence]," says that the claim is "extraordinary" on the basis of its low prior, then points out a piece of evidence with an extremely high likelihood ratio that would get you to believe the extraordinary claim (e.g., someone reporting the winning numbers and reading out the exact sequence), and somehow not realizing that by their own definition of "extraordinary," the evidence presented would count as extraordinary evidence. These arguments are typically followed up with all the same condescension and overconfidence (acting as though their terrible arguments have completely destroyed the common saying, and that anyone who still uses it is therefore an unsophisticated rube), all while still using the actual principle the saying points out in their evaluations of evidence not in support of their religion, since the actual principle is obviously correct.
Nice piece. Poor Hume, the greatest philosopher to write in the English language, and persistently misunderstood since 1739. He famously complained that _A Treatise of Human Nature_ “Fell dead-born from the press, without reaching such distinction as even to excite a murmur among the zealots.” So he wrote his Enquiries to explain things all over again. I’ve published on Nietzsche, and that guy is misunderstood even worse!
You’re 100% right about this being about the epistemic issue of warrant for beliefs in miracles and not the metaphysical issue of whether reality has any miracles in it. However, I think you err in saying Hume’s only concern here is to argue against the warrant of testimony-based miracle beliefs. Non-negligible portions of the essay are directly about warrant for other, non-testimonial, beliefs, including those based on experience and experiment. “On Testimony of Miracles” wouldn’t be a more informative title.
I think your assessment is 100% accurate! As a matter of fact, I almost wrote something like you say in the last sentence. The first section is basically an outline of the probability assessment through an empirical lens, culminating (in what I would argue) as a pragmatic summary of that argument (which is confused as his argument). The bulk of Hume's essay is really about what religious beliefs do to (see: corrupts) testimony, and the criteria by which we evaluate reliable testimony, especially as it pertains to novel events.
The feedback I'm getting on this inspires me to dive deeper. I think my post here is a good run down about how people get Hume wrong, but an essay on the actual in length, in detail argument of Of Miracles would be valuable
Of possible interest: https://youtu.be/K1pq1-DA1Bc?si=ieytYLpwRw_3HnrI
This is good!
thank you!
will give it a look later
The tendency to misunderstand Hume is like the tendency to misunderstand his fellow Skeptic, Sextus Empiricus. Not only is there a language problem (archaic English/translating from ancient Greek), but there's also a worldview problem. Dogmatism entails a certain "black and white" way of seeing the world: the world *is* this way or it *is not* this way. The Skeptic instead talks about the world *appearing* in a certain way, and is parsimonious with declaring is and is not. This way of thinking about and talking about things is confusing to the Dogmatist, who instinctively thinks the Skeptic is making kinds of claims different from the ones the Skeptic is actually making.
If the argument is as you have described it here, I honestly don’t think the difference between testimony and other forms of evidence is that important. The basic argument appears to be that given the broad experience behind the laws of nature, we would need extraordinary evidence to believe in miracle and as a factual matter while such evidence could be provided in theory, it has not actually been provided so far. In fact, all such evidence would have to be strong enough to more than cancel out the evidence against miracles and even so the resulting update would be only as strong as the update once it has been cancelled against the update from the evidence against miracles. Even if the original essay restricts the argument to testimony, this seems equally applicable to other forms of evidence. Assuming I have understood your understanding of the argument correctly. I think the bigger misunderstanding if you’re understanding is correct. Is that most people think he’s arguing that it is impossible even in principal to be persuaded of a miracle when he is only arguing that existing evidence is not sufficient to believe in a miracle.
So I think you're a little wrong here. Namely, he seems completely agnostic or silent on whether it's reasonable to believe a miracle if you observe it yourself! I said this in a footnote, but people sometimes I think equate a Christopher Hitchens point about thinking "maybe I'm mistaken" is more likely than you directly observing a miracle, but as far as I know, Hume doesn't argue this!
In this way, the primary thrust of Hume's argument is empirical epistemology: always trust the evidence you directly observe over what you are just told via testimony. One can expand upon the argument in the way you have, or have a separate conversation about whether empiricism is true, but I think the narrow claim is correct. At the very least descriptively, we don't find evidence or data compelling unless we experience it someway.
I do not think we actually disagree because I was arguing that the distinction is not one that is actually important. Not that the actual historical Hume thought that the distinction was not important. I find it entirely possible that he would respond as you responded, although I don’t personally find that convincing since for almost all the knowledge I actually possess I purely rely on the testimony of other individuals and yet I am much more confident that the testimony that Physicists have provided regarding the fact that the universe is expanding then I am regarding my memory of what I eat last Monday. Obviously, the reliability of testimony depends on the context like any other method of knowledge acquisition, but I just think that there is no good argument for categorically distinguishing between testimony and other types of evidence. This is not an argument about what a historical person believed, but rather my view of what the argument actually implies if taken to its logical conclusion.
Nice!
Interesting post. You definitely helped me understand Hume's arguments more clearly.
It doesn't have Hume in the title, but he is central to the book :)
https://www.amazon.co.uk/Opinion-Mankind-Sociability-Theory-Hobbes/dp/0691178887
DM me and i'll happily send you a PDF
I may do that once I'm through all these other books
This is one of my new favorite posts of yours!
Thanks! I’ve only begun!
I don’t think Hume’s argument that testimonies cannot justify beliefs about miracles is successful , I think it largely depends on one’s prior commitments, for example if you are a theist, then it won’t be unreasonable for you to believe that a miracle occurred based on testimonial evidence. A person who is a naturalist will not trust such testimonies.
Yes, I think I talk/mention this in the footnotes, but Hume definitely equate a “reasonable person” to mean someone who agrees with him wrt empiricism (i.e those who believe the most rational interpretation of evidence is the kinds we can directly observe). I think people are better suited to attacking that aspect of the argument, which is somewhat implicit, but that’s different from most arguments against Hume that I’m talking about.
But it's not applied consistently and thus not logical. The Christian evangelical will believe a testimonial claim about divine intervention by a fellow Christian but not fron a Muslim or Janseian.
As an addendum to this, it's worth mentioning that Hume is taking aim at natural religion and the idea that one can come to believe religions established on miracle claims coming from an unbiased perspective that doesn't presuppose its truth or reliability.
Thanks for mentioning my comment! Will take the time to read this post more thoroughly later
I think that testimony, paired with a rigorous systematic elimination of conceivable natural explanations, is very powerful evidence in favor of supernaturalism. I think most skeptics are way too dismissive of anomalous experiences of ordinary people, which are copious.
That's fine. That's not what Hume argued against, or relevant here, but that's fine.
Yea, I don't mean to assail Hume's point. I also find miracle claims from ancient manuscripts (such as that of the Bible) very weak evidence by themselves. (That being said, I find 400-year old extensive reports of miracles by authorities hostile to the one performing said miracles, such as the case of the Inquisition examining the levitation of St. Joseph of Cupertino, quite strong. For commentary on the documentation of this case, I recommend They Flew: A History of the Impossible by Carlos Eire.) I just wanted to express my opinion on something tangentially related to your post, but still important overall, in my opinion.
No worries! My bluntness is in being defensive against being dragged into an argument that isn't relevant to the post. Having said that, though I think Hume's point is rather solid, I think there are possibly ways around it, to justify "rationally" believing in religious miracles. I just don't think people attack him in the correct way to do it (i.e. they should attack his empirical assumptions, not against a straw man belief that he believes miracles can't happen)
The trouble with this is that a "rigorous systematic elimination of conceivable natural explanations" is easier said than done, and I don't think it ever has been done despite innumerable attempts. People always try to do it in arguments about miracle claims, but this typically takes the form of someone listing all the obvious naturalistic explanations they could think of and coming up with some arguments against each one. The arguments generally don't show each option to be impossible, but only give some reasons that each would be unlikely to happen. This is not very convincing to skeptics because it's unsurprising that in the entire course of human history, there would be some unlikely events that get misinterpreted as miracles due to the unlikeliness of their true causes (while it's unlikely for any individual event to be like this, over a long enough stretch of history, it's nearly certain that some events like this will happen). Furthermore, these types of argument frequently leave out plausible options and rarely provide any reason to think that the list of possible natural explanations is exhaustive.