One way, and I would argue a very Humean way, to look at the actual state of evidence regarding miraculous testimony is this:
Miracle stories are like anomalous data, very rare compared with testimony to the mundane. If there were no miracles in the world, we would expect testimony to them on rare occasions simply because of known fallibilities in human testimony (people lie, hallucinate, mistake what they see, etc.). Likewise, what is typically known to be uniform about genetics (that animals with a certain amount of genetic variation cannot interbreed) has extremely rare counter-testimony, like Herodotus who wrote about a horse giving birth to rabbit; there are fringe sources claiming human-pig offspring and other dubious hybrid claims like the woman who ‘gave birth’ to a cat and rabbits. We don’t accept those fringe claims for reasons exactly as strong as we do not believe miracles, and should not believe them unless they meet a high bar of testimonial evidence (ie a team of scientists whose findings can be replicated, perhaps even multiple times with photographic and video evidence, and examination of the offspring by independent experts). I take it as obvious no known miracles meet an evidential burden like that, which is exactly Hume’s point.
Related: Of Miracles (my version); The Probability of a Miracle; Of Miracles (Hume’s Original); Questioning Miracles (John Loftus); Hume, Licona, and Resurrections;
