A novice, I do not feel prepared yet to, but shall in future, read Hume; please tell me if Hume's originals answer my question.
Source: pp 30-31, Philosophy ; A Very Short Introduction (2002) by Edward Craig.

We receive a report of something – for convenience call it The Event – supposed to be miraculous. So we are asked to believe that The Event occurred, and that this was contrary to a law of nature. For us to have good reason to believe that an event of that kind would have been contrary to a law of nature, it must be contrary to all our experience, and to our best theories of how nature works. But if that is so then we must have very strong reason to believe that The Event did NOT occur – in fact the strongest reason we ever do have for believing anything of that sort.

So what reason do we have on the other side – to believe that it DID occur? Answer: the report – in other words the fact that it is SAID to have occurred.

[1.] Could that possibly be so strong as to overpower the contrary reasons and win the day for The Event? No, says Hume, it could (in theory) be of equal strength, but never of greater.

There might be such a thing as testimony, given by sufficiently well-placed witnesses, of the right sort of character, under the right sort of circumstances, that as a matter of natural (psychological) law it was bound to be true. But that would only mean that we had our strongest kind of evidence both for The Event and against it, and the rational response would be not belief but bewilderment and indecision.

1 appears too extreme and unconditional; hypothetically, why can reports of an Event never overpower the contrary reasons and win the day for The Event? For example, imagine a island where all residents excel in philosophy, always are benevolent and tell the truth.
Then their reporting of a miracle CAN overpower?

I ask out of curiosity; realistically, I know that reports cannot be this strong and that my imagined island does not exist.

3 Answers 3


The Event can never overpower the contrary reasons because it came in the form of a report. It is always more believable that something statistically unlikely occurred than something "impossible" occurring. The logic would be that you could never have an island full of benevolent philosophers who always say the truth. The best they could do is give their best possible interpretation of what they perceived. If they were capable of always saying the truth, they would not be called philosophers. They would be called oracles, and themselves would be a miracle.

This would certainly hold true for any report received via natural means. A supernatural report could theoretically overpower the contrary reasons, because by definition it is not bound by the same rules as natural reports.


Since Hume says we must weigh the evidence between things based on the body of evidence, and miracles by definition happen rarely (and have little evidence), the body of evidence will always be against miracles.

Another view on the question, however, is provided by CS Lewis, who wrote a work titled Miracles. In it he argues

Those who assume miracles cannot happen are merely wasting their time by looking into the texts: we know in advance what results they will find for they have begun by begging the question.

Lewis's view is not that it's enough to weigh common experience (which does not occur as miracles) with reports of miracles. The question is whether a particular report is credible in its self. Assuming from the start that it's not credible because of its contents is a rhetorical fallacy (i.e., as he says, begging the question).

For another view that would disagree with Hume, one can look at a particular data point. Coincidentally, this question was posted a day after the anniversary of the (claimed) apparition of Jesus's mother to a young girl in France, and subsequent miraculous healings. This particular event:

  1. Happened not in antiquity, but in post-Enlightenment France
  2. Was initially rejected as a real thing by religious authorities
  3. Has since resulted in 68 healings that have been highly documented, and critically investigated by expert doctors of many different religious backgrounds.

Any theory concerning miracles needs to fit all the data appropriately, including the data that comes from Lourdes.


If I hear a report that a miracle happened, somewhere, some obvious possibilities are:

(1) A miracle happened. (2) Something happened that very much looks like a miracle, but a natural explanation can be found. (3) Someone very cleverly arranged events that looked like a miracle. (4) Someone with too much fantasy observed events that were a bit unusual and declared them to be a miracle. (5) Someone or some group got all excited about nothing and declared a miracle when nothing unusual actually happened. (6) Some people try to make themselves look important by falsifying claims of a miracle. (7) A newspaper made up a story from absolutely nothing.

From the report I heard, I have to decide which explanation I think most likely. Deciding that (1) is the most likely is very, very difficult.

What Hume says is different, of course: I have many years of life experience that tells me that miracles don't happen. Hume says that if something happens that is totally contrary to my life experience, and totally believable, just as my life experience to me is totally believable, then we have a stalemate between life experience and miracle. And that's if the report of the miracle is 100% believable.

I would add that the sum of my life experience has changed after the event. That report of a miracle is now part of my life experience, so my life experience isn't anymore that "miracles cannot happen". If there is another report that is 100% believable, and then another one, then eventually you would change your mind.

You must log in to answer this question.