However, this is not a historical claim about what has happened, it is a philosophical claim about what can happen and what can be allowed into our theories. Unfortunately, most scholars in this field have little to no philosophical training, so this assumption against miracles is more of a reflection of modern academic culture than the result of philosophical analysis. It is, essentially, a bias, dating back to Hume, that any naturalistic explanation is more likely than a supernatural one. A few points should be made about this.
First, why are miracles so improbable? As Kreeft and Tacelli (two philosophers) put it in Handbook of Christian Apologetics, "They are certainly unusual, but how do we know whether they are likely to happen or not? Only if we have already decided whether or not it is likely that God exists -- or that he would ever work a miracle. In that case, calling miracles ‘maximally improbable’ is not a neutral description: it stacks the deck against them." No one has ever produced a reason why, if God exists, he would be unwilling or unable to perform a miracle. So unless we can be absolutely certain that God does not exist, miracles are possible. And if miracles are possible it’s neither valid nor honest to blindly assume a priori that they’ve never occurred.
Moreover, a miracle can be perceived by the senses like any other event and thus can be supported by historical testimony. Because of this, testimony to a miracle is not nullified by the regular order of events. As William Lane Craig, another philosopher, writes in Reasonable Faith,
Of course, a historian could be so deeply prejudiced in favor of naturalism that he resolutely refuses to accept any miraculous hypothesis. But that is just a fact of psychology, which does not undermine the objectivity of history, any more than does the case of a Marxist historian who shuts his eyes to un-economic causes of historical development or a Confederate historian who refuses to acknowledge any responsibility of the South in bringing on the Civil War.
As a result of this bias, when certain scholars examine the historical evidence with regard to the resurrection of Jesus, some of them start by presupposing it couldn’t possibly have happened. As such, their conclusions that "therefore, it didn’t happen" are hopelessly circular. In Jesus Under Siege philosopher Gregory Boyd writes,
The obvious question that we must be asking ourselves through all of this is, why do these scholars assume that God could not have become a human being and that divinely inspired miracles cannot occur? On what basis can they be so confident about what can and cannot happen in history? By what means do they come to know so much about God and the nature of the world that they can confidently pronounce, prior to any investigation of the evidence, that God has never intervened into the world!? Wouldn’t you have to be God Himself to know this?
I think that part of the problem is that any miraculous explanation of an event is automatically considered to be ad hoc. The more a theory goes beyond what is known, the more ad hoc, or contrived, it is. But "miraculous" and "ad hoc" are not synonyms.
This was demonstrated to me in a debate between Craig and Robert Greg Cavin on the historicity of Jesus’ resurrection. Cavin acknowledges that the historical evidence proves that Jesus was killed and buried, that the tomb was found empty a few days later, and that many people experienced what they understood to be appearances of Jesus alive from the dead. However, Cavin doesn’t accept the resurrection of Jesus as the best explanation of these facts: rather, he believes it’s more plausible that Jesus had an evil twin (no joke). They were separated at birth when one of the identical twins was switched with the baby to which Mary had given birth. Immediately after Jesus was killed, the twin just happened to arrive in Jerusalem and, learning of the events, decided to steal Jesus’ body and impersonate his long lost brother. This theory is ridiculously ad hoc in that there is virtually no evidence in support of it, and Cavin has to suggest more and more outrageous tidbits in order to plug all of the holes in it. There is no reason to think that Jesus had an identical twin brother; there is no reason to think that Mary wasn’t Jesus’ biological mother; there is no reason to think that Jesus and his twin would have grown up and lived geographically separated from each other; there is no reason to think that Jesus’ twin would have arrived in Jerusalem at exactly the time when Jesus was crucified; etc. There’s also blatant inconsistencies in this scenario: the twin would have to be ethical enough to want to continue Jesus’ work and teaching, but unethical enough to want to hoax the resurrection. He would have to be brilliant enough to have the idea occur to him (Jesus’ resurrection contradicted fundamental Jewish categories of thought) and moronic enough to want to impersonate someone who had just been condemned by the Jews as a blasphemer, and executed by the Romans for treason. Obviously, Cavin’s theory is completely contrived and ad hoc. By way of contrast, the only extra supposition we have to make to accept the resurrection is that a God capable of doing it exists -- and since most people already believe in God, this wouldn’t be an additional supposition.
Some also get nervous that any miraculous event would demonstrate that we live in a capricious universe, since any miracle would supposedly be an entirely arbitrary act of God. But, of course, this is false. With regards to Jesus’ resurrection, for example, the Messianic expectation of the Jewish people, the uniqueness of Jesus’ life, and the relevance that it has for one of the most fundamental aspects of human experience (death) makes it about as far removed from caprice and arbitrariness as anything could be.
Update (15 Feb 2010): See also part 2, part 3, part 4, and part 5.