Do ‘extraordinary claims require extraordinary evidence’?

NB: This post is several years old now. I will update it at some point, and cannot confirm that I agree with everything in it. Nevertheless, I am leaving it here on the website to provoke some thought.

This was copied almost verbatim from a response I made on a forum a couple of minutes ago, and so I have not spent a huge amount of time on it. As such, it is rather unfurnished and may not read particularly well out of context, but I have tried to tidy it up. Some ideas have inevitably come from other sources and minds, to which due credit will be given in the footnotes, but the articulation, presentation, structure, some sub-arguments, and the general arguments are my own work.

Do ‘extraordinary claims require extraordinary evidence’?

This essay is a general response to the argument appropriated by sceptics in reference to the reliability of the New Testament (and other theistic claims)- not in terms of general history or any general theology, but specifically deliberating over the philosophical and historical nature of the resurrection of Jesus of Nazareth. The argument may come in the very vague form, ‘extraordinary claims require extraordinary evidence'[1], or they may be sentiments that have been elaborated on slightly more, such as David Hume’s idea ‘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 endeavors to establish.'[2] I commend Hume’s articulation, as it seems to give the admission that some supernatural miracles could hypothetically be more probable than certain events occurring given their falsehood. However, most arguments of this kind are not as clear as Hume’s (whose own still gives some ambiguity), and rely on a vague description involving probability to the general notion that “Miracles are exceptionally improbable, and it is therefore almost certain that Jesus did not rise from the dead”. This is admittedly a weak form of the argument, but I hope that my essay will clarify and illumine some of the flaws so that a better articulation is constructed. I may give a suggestion on how this may be done at the end.

The argument put to me

The original argument I was given then, was this:

“- When judging evidence, we should consider probabilities. Is it more unlikely that the event happened, or that the source of the claim is incorrect? Whichever is more unlikely is the one we reject.
– For most of us, all our sensory experience points towards nature. Furthermore, claims of contemporary supernatural events are, in almost all cases, rejected by everyone as being mistaken. Supernatural events are extremely unlikely.
– The evidence for the supernatural events in Jesus’s life are of a good standard for the time they were written, but they are not of an exceptional standard. The gospels are written perhaps 50 years after Jesus’s death, and are secondary sources that are drawn from two or three (possibly) primary sources. They seem to be written from the perspective of followers of Jesus (rather than people who observed the miracles and were convinced), and were written in a superstitious time. We can say that the evidence is good for the time, but not of an extremely good standard compared to the evidence for other historical events. It is not extremely unlikely that the gospels are mistaken.
– It is more unlikely that the events happened than the gospel writers made a mistake.
– We do not have good reason to accept the historical evidence for the supernatural acts of Jesus.”

This argument in particular was a general response to an argument concerning the unlikelihood of Christianity starting, unless the resurrection were true, which will be referenced in my rebuttal. Let us then, begin.

While, of course, I wouldn’t build a case for the existence of God solely on the historical evidence, I do think there are certain problems with the ‘extraordinary claims’ argument. They all link together, but I will try to separate them to an extent.

The vague and subjective nature of ‘extraordinary’

Firstly, the Saganised form (ie the title) relies on a very ambiguous idea of ‘extraordinary’, which is likely to involve biased, subjective judgements on how ‘extraordinary’ a particular event is. Of course, ‘extraordinary’ may simply mean anything out of the ordinary, in which case we most certainly have enough evidence for Christianity, on the basis of the New Testament. True, resurrection is an unusual event, but the series of events surrounding the supposed resurrection are, to my mind, unique among antiquity. However, it is clear that ‘extraordinary’ in this context is being used relatively- and it is this aspect I am criticising. It is, in fact, an appropriate criticism to any argument of this nature, as all will invoke some subjective, personal, biased judgements on how ‘extraordinary’ or ‘improbable’ something is. It is therefore immediately clear that we should not give the argument any more credibility than is due, as all approaching the issue will have some sort of bias- an ontological naturalist will have a bias that excludes any supernatural action, a priori attaching an infinitessimal likelihood to ‘miracles’ (I will elaborate on why this is unfair in another post), whereas a devout theist will have a natural tendency to explain phenomena in terms of a personal God, perhaps giving the likelihood of a supernatural occurrence more probability than is appropriate.

The misunderstanding of the Christian claim

Secondly, it seems to rely on a very ambiguous conception of probability, especially with regard to establishing the ‘extreme improbability’ of a supernatural event. It is difficult to verify exactly what we mean by this. Of course, it is (for the sake of this argument, though this can be disputed) extremely improbable that the laws of nature we suppose (and have supposed since the dawn of modern science- ironically the child of a ‘religious’ worldview) should, at any given moment, cease to function with no warning, rationale, explanation, or purpose. When we see a feather floating in the air, we are justifiably inclined to suppose that it is a windy day, as opposed to the spontaneous cessation of gravity. When something contrary to our immediate sensory experiences happens without a rationale, it is fair to assume that there is another explanation behind it. However, this is not the Christian claim. It should be remarked that the Christian claim is based on the observations we do have, not the observations we should expect to make at any arbitrary moment in history. Certainly, if no explanation, expectation or reasoning was given for Jesus’ apparent resurrection, it would not be such a forceful or important claim. If he was supposed to have naturalistically risen from the dead, we might consider that of such an extreme probability as to dismiss it. This, however, is not the proposition. The proposition is that Jesus was the anticipated Messiah[3], sent by God to bring about a new covenant with his people- and not only that, but it was God who (by fair reason) raised Jesus from the dead. The converse of this is not that people generally tend to stay in their graves, but that Jesus did not, and it is this evidence we should be assessing. Christianity does not contend that the laws of nature are inconsistent, and are prone to random distortion. After all, they too live on the basic assumption that nature is fairly consistent (it would have to be for the Christian doctrine of free will [4]), they buy a house on the basic assumption that it should not burst into flame and crumble every so often. Indeed, it follows this general assumption with regard to most other corpses (contemporarily, at least- this ignores the Jewish notion of the final resurrection of the dead at the end of time), but what is instead being proposed is that Jesus rose from the dead. In order to refute this, then, we must look at the evidence for Jesus’ resurrection, not the lack of resurrection of most other people in history. Now, certainly, we should not be led into some sort of naivety to suppose that everything that seems prima facie contrary to nature is some sort of supernatural act- but when there is an expectation (Old Testament prophecy), a fair explanation (that, if it were true, should increase the probability enormously, ie the action of YHWH), a reasoning given (to initiate a new covenant for human salvation[5]), and plenty of other reason to believe in a quasi-Christian God[6], the proposition is not so absurd. Again, the Christian argument is not that Jesus rose from the dead naturalistically and arbitrarily (as that is, indeed, improbable), but that Jesus conquered death after prophecy, expectation, and for a purpose, through power that was appropriated to him through He who supposedly created the (general) laws of nature to be contravened.

The unqualified use of ‘probability’

The third point is related to the second- on the general notion of the vagueness of ‘probability’. As demonstrated from the second point, we are not concerning ourselves with a simple, arbitrary, statistical or physical probability- one that is concerned with probabilities of the nature, “that one person arbitrarily picked from history has come back from the dead”, or, “that any given person in the future should spontaneously and naturalistically be resurrected”. The probability we are assessing is of a different nature- using ‘abductive’ reasoning, or ‘inference to the best explanation’. Not only do we assess the probability of explanation A (in this case the resurrection of Jesus), which I have explained is a flawed premise, but we also assess the likelihood of the observations had not-A occurred. I commend Hume for noticing this idea in his articulation of the argument, although with my final point I will contend that his conclusion was unjustified. The idea that we not necessarily need extraordinary evidence for all unlikely events is portrayed reductio ad absurdum by philosopher William Lane Craig: “…this standard would prevent you from believing in all sorts of events that we do rationally embrace. For example, you would not believe the reports on the evening news that the numbers chosen in last night’s lottery were 4, 2, 9, 7, 8, and 3, because that would be an event of extraordinary improbability. The odds against that are millions and millions to one, and therefore you should not believe it when the news reports it. Yet we obviously believe we’re rational in concluding it’s true. How is that possible?…if the evening news has a very high probability of being accurate, then it’s highly improbable that they would inaccurately report the numbers chosen in the lottery. That counterbalances any improbability in the choosing of those numbers, so you’re quite rational to believe in this highly improbable event.”[7] This principle can be applied throughout history- it may be considered extremely improbable for someone to have conquered the known world by the age of 33. Yet it is reasonably accepted that Alexander the Great had done so, and there is plenty of evidence to verify it[8]. One can strengthen this point even further by illustrating it through the example of the existence of homo sapiens. It is considered fantastically improbable for, not only the universe to provide the physical laws necessary for life to arise, but combined with the unlikelihood of the human race being exactly as it is now, we are faced with an extraordinarily improbable event. However, it is abundantly clear that the human race is as it is (this is no more than mere tautology), and so we can see that what may originally seem improbable is in fact the best explanation of a given circumstance. And so, ironically, the more emphasis that is placed on the original argument, the less we can accept the theory of evolution, or even our own existence! The weaker the argument from the New Testament becomes, the stronger the teleological argument becomes- and so to use the “too extraordinary” argument as an argument against the existence of God is literally self-refuting![9] Note that this is not a defamation of evolutionary theory; I believe the evidence is perfectly sufficient to conclude that evolution by natural selection has occurred, this merely illustrates the implausibility of the original argument. Now, when discussing the historicity of the New Testament, we might point out the masses of manuscript numbers (over 5,000 in Greek) in comparison to other works of antiquity, we might point to the early fragments, we might point to the corroboration from several different sources[8] and conclude that, if it were a secular pronouncement, it would be the most well attested event in all history; certainly a vast multitude of historians would share this conclusion. However, I have included that all important clause- “if it were a secular pronouncement”. This, says the sceptic, is the clause that ruins the argument from the New Testament. For, though we have strong historical evidence, we are trying to verify an immense claim, one that requires a huge amount of evidence. This is a fair criticism, and will be dealt with in the following points, but the point to be remembered is that ‘probability’ has a wide range of uses, from the statistical, to the physical, to the inductive, to the abductive; the sceptic argument generally invoking a very vague description.

The unfair demand for evidence

The fourth point is related to each of the previous points, and may be posed by a question, “What would count as ‘extraordinary evidence?’ Most of the replies I generally receive are not particularly sensible, and generally resemble the clichéd refrain, “I’d believe in God if he appeared out of nowhere and moved mountains for me when I said so,” with various paraphrases. Now, ignoring the suspicion that this may not be true in many cases, the lack of a sensible answer in most cases highlights the implausibility of the sceptic’s demands. True, there are some fair remarks (eg, if the resurrection were true, we may expect to find accounts of appearances at the end of Mark’s Gospel), but generally they involve an implausible demand. “Why did hardly any secular historians mention Jesus’ resurrection?” is a typical example. To demonstrate, I will suggest why this is an unfair requirement, and why we would not expect there to be a multitude of secular historians attesting to Jesus’ resurrection. Firstly, the question presupposes that there were many 1st century historians who could look at the evidence and remain a sceptic. This is a simple point- some of the first century historians (eg Luke) converted to Christianity because of the strength of the evidence, and so it is unfair to ignore those historians who became Christian. Second, Jesus died a third of the way through the first century, and because oral tradition would successfully retain the story early on, there would be no need to write events down until a few centuries later. This greatly reduces the amount of time possible for historians to mention Jesus’ ministry. Third, Jesus was not of any great concern to secular historians around the time of his ministry. Of course, he is rightly worshipped and of utmost importance now, but to most secular historians in the 1st century he was merely the crucified leader of a Jewish cult. Historians were far too busy writing accounts of wars and the like. Fourth, we would not expect a secular historian to be strengthening the case of Christianity if he did not believe it. Again, this is an elementary point. We have no expectation of someone who wasn’t a Christian to start writing accounts confirming the central Christian message! Fifth, there are secular accounts of Jesus’ ministry. True, they do not attest to the resurrection, but we have people like Josephus and Tacitus verifying important aspects of Jesus’ ministry fairly soon afterwards. From these points, we can see that the sceptic’s claims are often implausible with regard to expecting ‘extraordinary evidence’. The reason we may not have it is because we would not expect it, and that is a perfectly fair defence. This only highlights a problem with evidentialism (something the objector often treasures more than life itself), as opposed to a deficiency in Christianity.

The undermining of the strength of the Christian case

Finally, I think we DO have very strong evidence for the resurrection of Jesus; as demonstrated from point three, it is inference to the best explanation that is appropriate in this context, as opposed to a weighing up of each side while trying to keep them distinct. The original article in question[9] is the main ‘advocate’ of those things observed that we should not expect if Jesus had not been resurrected, and I feel they do add up to an incredibly strong case- certainly invoking a large improbability. Because that article and the general claim of the predisposition of early Christians to adopting Christianity are well known (I have referenced the link), I do not feel it necessary to repeat these arguments in detail here- we are talking generally of the likelihood of a movement starting around a crucified Messiah, an egalitarian social philosophy, a divine figure who displayed a lack of omniscience, omnipotence, and who displayed typically ‘weak’ emotions like weeping; the list goes on. On top of this, of course, are the empty tomb and resurrection appearances, but again, these are well known[10].
A further point on this: it does seem to me that the ‘extraordinary claims’ line crops up when a sceptic has no other defence for an atheist position. The likelihood of Jesus’ resurrection does not prove conclusively that God exists, only that it is more probable. It is therefore perfectly within the capability of an intelligent atheist to give other grounds for supposing that the Christian God does not exist- why not just accept that the 1st century phenomena increase the probability that the Christian God exists, and then give further evidence to suggest that the Christian God does not exist? It does often tend to be a strong reliance on the “onus of evidence is on the theist” claim (one which I disagree with, and will expand on in other posts). To take a simple example of how one phenomena increases the probability of Jesus’ resurrection (and hence the Christian God’s existence), we can look at the empty tomb. If we accept the historicity of the empty tomb, we can establish a Bayesian confirmation of Jesus’ resurrection. If A is Jesus’ resurrection, and B is the empty tomb, Bayes’ Theorem states that P(A|B) = [P(A)P(B|A)] / P(B). That is to say, the probability of Jesus’ resurrection given the empty tomb, is equal to the prior probability of Jesus’ resurrection (P(A)), multiplied by the probability of the empty tomb given Jesus’ resurrection (P(B|A)), divided by the prior probability of the empty tomb (P(B)). If P(A) (that is, the probability of Jesus’ resurrection without considering the empty tomb) remains constant (which it shall do in this case, since it is unaffected by whether there is an empty tomb or not), it follows that, if P(B|A) is larger than P(B), it is more likely that Jesus was resurrected than is otherwise so. This seems well established: if Jesus rose from the dead, it is very very likely that there would be an empty tomb, and therefore P(B|A) is high. Without taking into account whether Jesus rose from the dead or not, there *are* still reasons why an empty tomb would be present, but the probability is not particularly high. It then follows that P(B|A) > P(B), and therefore the probability of Jesus’ resurrection is made higher by the empty tomb. I realise this is perhaps confusing, and I did not have time to elaborate in this post on Bayesian proof. I will therefore elaborate on this in another post, to which a link will be given in time. My general point, however, is this: there are evidences that make Jesus’ resurrection more probable than it would have been without them, and the sceptic often tends to ignore this, and the fact that they could criticise Jesus’ resurrection on other grounds.


It is therefore my conviction that this argument fails on several grounds:
1. The ambiguous and subjective nature of “extraordinary”.
2. The lack of clarification of the Christian proposition, and what it means to say it is ‘unlikely’.
3. The ambiguous nature of “probability” and the absurdities it can lead to.
4. The unfair demand for evidence which we would not expect to have if the resurrection had occurred.
5. The incredibly strong affirmative evidence we do have for the resurrection of Jesus Christ.

I have shown my respect for Hume’s formulation of the argument, and I think his argument fails instead, on the first and fifth points. What must be done to refute the New Testament is to clarify the nature of probability as Hume (and others) have done, and then refute my fifth point, perhaps showing that the evidence we do have is not as strong as it seems. As I am personally convinced that the evidence for the New Testament is strong enough to suggest that Jesus did rise from the dead, my formulation of a sceptic’s argument would be to suggest that perhaps it is a matter that cannot be resolved historically. True, we have as much evidence as we would expect, but perhaps this is not enough evidence to base a whole life upon. This, again, invokes subjectivity, but I would point out as I did at the start, that there are other evidences (and that some degree of subjectivity is necessary to retain our free will- this then moves onto the argument from hiddenness which is not appropriate to discuss here). To refute the resurrection, either we conclude that we cannot make a decision of such magnitude on the New Testament, which would be criticising the domain of competency (of evidentialism) as opposed to the actual argument (and as I have demonstrated in this essay, is a particularly fallacious attitude), or we try to refute the evidence for the resurrection itself- the resurrection of Jesus that is, not everyone else in history.

Thank you for reading, feel free to leave a comment or mail me for more information!

NB: I hope to improve and expand on this to deal with how improbable miracles really are, and whether ‘extraordinary claims’ really do require ‘extraordinary evidence’ (further). It is probable that I shall use CS Lewis, CB McCullogh, and John Earman (Author of ‘Hume’s Abject Failure’) when doing so, but revision is a bit heavy at the moment.


1. Generally attributed to Carl Sagan with no one particular reference.
2. David Hume, An Enquiry Concerning Human Understanding.
3. Some even claim the Old Testament prophecies pinpoint a specific time for the arrival of the Messiah, cf. the Book of Daniel predicting Him a certain time after the reign of Artaxerxes of Persia.
4. See CS Lewis, The Problem of Pain, 20-25.
5. See Anthony Phillips, God B.C. (Oxford: Oxford University Press, 1977), 36-39 for a brief discussion of the understanding of this new covenant prior to Jesus.
6. As given in, for example, Richard Swinburne, The Existence of God 2nd ed. (Oxford: Oxford University Press, 2004).
7. Lee Strobel quoting William Lane Craig, The Case for Faith, 88-89.
8. See CARM, “Extraordinary claims require extraordinary evidence”,
9. That is, if you take evolution to count against the teleological argument (many don’t, perhaps myself included). A more appropriate example to use in place of evolution would be our own ability to reason. It seems to me greatly improbable that, if naturalism were true, we should have true beliefs about the world. For more on this argument, see CS Lewis, Miracles, Chapter 3, ‘The Cardinal Difficulty of Naturalism”, 17-36. Alvin Plantinga and Victor Reppert are contemporary advocates of this argument.
10. For further reading on these, see FF Bruce, The New Testament Documents: Are They Reliable?
9. See JP Holding, “The Impossible Faith”,
10. See Lee Strobel, The Case for Christ. There are better books out there, but that’s a good one for beginners. Ask me for more if you want.

Further Reading
Keith Ward, The Big Questions in Science and Religion (Pennsylvania: Templeton Press, 2008), chapter 4.
CS Lewis, Miracles (London: Harper Collins, 2002)
Lee Strobel, The Case for Faith (Michigan: Zondervan, 2000), ‘Objection #2: Since Miracles Contradict Science, They Cannot Be True’ pp. 77-117

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s