After a couple weeks hiatus I find myself picking up my review of The End of Christianity once again. Although I have published about fifteen posts in critique of The End of Christianity over the last couple months, I have not received a single response from the essayists. Well okay, there was one response: John W. Loftus summarily dismissed everything I say because I’m a Christian. But beyond that … silence.
I feel like the wolf killing one sheep after another. I can see the shepherds over there huddled by the fire. When are they going to intervene? Never mind. I turn my attention to another plump sheep…
Today the focus will be Matt McCormick’s essay with the intriguing title: “The Salem Witch Trials and the Evidence for the Resurrection.” You can pretty much guess where he’s going with that. McCormick’s argument is aimed at the rationality of belief in the resurrection:
“If we do not have enough evidence to justify believing that Jesus came back from the dead, then the central tenet of Christianity, that the son of God sacrificed himself to provide moral and spiritual salvation for humanity, isn’t reasonable to believe.” (195)
Note that McCormick is not arguing that the resurrection didn’t happen. Rather, he is arguing that we lack sufficient evidence to believe rationally that it did happen. And where do the Salem Witch Trials fit into all this? Allow me to summarize the argument in four propositions:
(1) The evidence for the claim that there were witches at Salem is better than the evidence for the claim that Jesus was resurrected.
(2) The evidence for the claim that there were witches at Salem is not sufficient to believe reasonably that there were witches at Salem.
(3) If your evidence for claim (a) is evidentially better supported than your evidence for claim (b) and you do not have sufficient evidence to believe (a) reasonably then neither do you have sufficient evidence to believe (b) reasonably.
(4) Therefore, the evidence is not sufficient to believe that Jesus was resurrected.
So what can be said for this argument? In my critical evaluation of McCormick’s argument (or, more correctly, my reconstruction of his argument; if McCormick thinks I misrepresented him he is free to point it out) I’m going to proceed in three steps. I’ll begin by making a critical point about his treatment of confirmation bias. Next, I’ll focus on a critical engagement with (1). I’ll conclude by critically engaging with (3). But as we will see, the confirmation bias will emerge as the unintended star.
1 McCormick’s confirmation bias
McCormick points out that “humans have a very strong tendency to find evidence for the conclusions they favor.” (201) I’d be more convinced of that thesis if McCormick wasn’t so selective in the evidence he presented for it. Just joking. I agree, which is a good thing because the evidence for the confirmation bias is overwhelming. In fact, McCormick evinces it in this very essay.
But wait, I’m getting ahead of myself.
McCormick’s focus is not on his own confirmation bias (unfortunately) but rather on that of Christians. He avers that Christians are predisposed toward evidence for the resurrection of Jesus:
“People do not, by and large, become Christians on the basis of a strict reflection on the historical evidence. By contrast, how often does someone adopt the Christan views held by his parents from his childhood and then conclude that there is a compelling case for the existence and resurrection of Jesus?” (201)
No doubt this is true. For many Christians historical study is not the primary catalyst of faith. (Though it is for some, including folks like Frank Morison and Lee Strobel.)
Nonetheless, McCormick is actually conflating two different issues here: (A) the extent to which an individual adopts certain beliefs through a non-evidentialist process; (B) the extent to which an individual exercises a confirmation bias in their evaluation of evidence. I will defer discussion of (A) to the third section where I will critique (3).
For now let’s keep focused on (B). As I said, I agree that Christians frequently exercise a confirmation bias in the way they evaluate evidence. Just like everybody else, including (ahem) Matt McCormick. It is pretty clear that if some Christians go a little easy on the historical evidence for Jesus, Professor McCormick goes way harder on that evidence than do most historians. This is evident when McCormick is critiquing Tom Wright:
“What’s puzzling about Wright’s argument here is the conflation of the attributions of Jesus we have in the Gospels with everything that Jesus would have ever said to the disciples (assuming that Jesus existed at all).” (205, emphasis added)
Wait a minute, what’s with that “assuming that Jesus existed at all” quip? McCormick is clearly trying to convey the impression that there is some reasonable doubt that Jesus existed. He reminds me of the Exxon scientist being interviewed on Fox News who can’t help but quip “assuming that climate change exists at all”. Hey Mr. Exxon, there is no doubt among scientists that climate change exists. And Professor McCormick there is no doubt among historians that Jesus existed.
But all is not lost. By expressing a lingering skepticism about the existence of Jesus McCormick handily demonstrates the confirmation bias at work. Lessons like this are pedagogically effective, for few students will forget the irony of a professor who rails against the confirmation bias while exercising that very bias. Well done Professor McCormick!
Unfortunately there is a downside. Once we realize that McCormick’s engagement with the evidence is skewed against Jesus it immediately calls into question his assessment of the data and thus his support for (1). A quick perusal of some elements of McCormick’s argument in the next section will confirm those worries.
2- The evidence for the resurrection
Let’s recap (1):
(1) The evidence for the claim that there were witches at Salem is better than the evidence for the claim that Jesus was resurrected.
I am not going to bother with Salem here for a simple reason: it is a distraction. Whatever we say about Salem, our question here is whether McCormick has provided adequate justification in his essay to conclude that there is insufficient evidence to believe rationally in the resurrection of Jesus.
On that count McCormick’s essay is a great failure. It is a sign of problems when he can dismiss world renown New Testament scholar NT Wright’s critically acclaimed 800 page tome The Resurrection of the Son of God (Augsburg Fortress, 2003) in a couple pages.
The problem can be illustrated by pointing out three inter-related problems: McCormick tries to set the evidential burden for rational belief in the resurrection at an inordinately high level, he fails to recognize the role of supporting webs of belief and he engages with Christian historians like Wright in a way that is tendentious and uncharitable.
Let’s begin with the first point: the unreasonably high evidential burden. McCormick writes:
“let us consider a rough summary of the sorts of reasons that are often given for accepting the historical evidence that allegedly proves the resurrection of Jesus.” (202)
The problem centers on the loaded choice of the word “prove”. Proof can mean all sorts of things including logical proof, proof beyond a reasonable doubt, and proof by a preponderance of the evidence. McCormick never specifies what he means or precisely how much evidence is required to believe rationally a particular reconstruction of a historical event. This is unacceptable for it is essential to know just how much evidence McCormick demands for rational belief. Without such guidance the reader will likely imbue his or her own sense of proof. So then the question becomes: how much proof will the reader assume is required?
At this point I am forced to offer my own speculation. The average reader will not assume we’re looking for a logical proof. Nor, I suspect, will that average reader go for the lower proof by a preponderance of the evidence. Instead, they will likely go in the middle for proof beyond a reasonable doubt. And surely McCormick knows this.
This brings us to the problem: to demand proof beyond a reasonable doubt as the standard before you can rationally assent to the truth of a historical reconstruction of a past event is absurdly high. And if we did accept this as a standard it would mean most historians were irrational. Consider:
Historian A believes the A reconstruction of event Q. Historian B believes the B reconstruction of event Q. If the rational stipulation for believing a reconstruction of a past event is that the reconstruction demand the universal assent of rational people then either Historian A is irrational or Historian B is irrational or both Historian A and Historian B are irrational
Of course, given that historians disagree about all sorts of things, this entails that most if not all historians are being irrational much (or most) of the time. But rather than conclude that historians are generally irrational, we surely ought to reject the “proof beyond a reasonable doubt” criterion as too high.
This means that we ought to favor the weakest sense of proof. It is sufficient to believe a particular historical reconstruction if one believes that the evidence, on balance, supports that reconstruction rather than another one. But this immediately forces a rather uncomfortable question: does McCormick seriously believe that he has shown the evidence does not on balance support the resurrection hypothesis?
This brings me to the second point. McCormick’s appraisal of rational belief has all the nuance of a student who just read Descartes’ “Meditations” for the first time. Now I admit that sounds harsh, but it is true nonetheless. Here’s why. McCormick writes as if there is one monolithic way to approach the evidence: either it is, or is not rational to believe p. But that’s all wrong. All sorts of other beliefs are relevant when we are assessing evidence. Consider Matt and Randal as two examples:
Matt believes God does not exist.
Randal believes God does exist.
Do you think the fact that Randal, but not Matt, believes in the existence of God will affect how each assesses the historical evidence? Of course it will. If Matt comes to the evidence believing that there are substantial defeaters for the existence of God then those defeaters will provide substantial obstacles to Matt’s coming to the conclusion that God raised Jesus from the dead. By contrast, if Randal has defeaters for those defeaters then he will assess the quality of the evidence differently.
This means that simplistic statements about what it is rational to believe which make no reference to the other beliefs a person holds should raise red flags. Belief doesn’t work like that.
Now to the third point. Here I’ll focus on the way that McCormick interprets N.T. Wright. To do that we’ll go back to the passage I quoted above:
“What’s puzzling about Wright’s argument here is the conflation of the attributions of Jesus we have in the Gospels with everything that Jesus would have ever said to the disciples (assuming that Jesus existed at all).” (205)
Here our attention is on the first part of the sentence. Wright makes no such conflation. Consider the passage from Wright that McCormick cites in support of this claim:
“It is out of the question, for a start, that the disciples were simply extrapolating from the teaching of Jesus himself. One of the many curious things about Jesus’ teaching is that though resurrection was a well-known topic of debate at the time, we only have one short comment of his on the subject.” (Cited in 205)
So McCormick assumes that this passage implies that Wright believes the gospels record everything Jesus taught to his disciples. But the passage reflects no such assumption. What then is Wright assuming? Allow me to offer a much more plausible suggestion:
Wright’s Assumption: if belief in the resurrection is to be plausibly traced geneologically to the teaching of Jesus then teaching about the resurrection would figure more prominently into that teaching as recorded in the gospels.
Surely this is a plausible, even persuasive assumption. So why does McCormick adopt such an implausible, tendentious reading of Wright?
The most plausible answer to that question is also the most plausible answer to the question of why McCormick adopts a monolithic concept of reason and an inordinately high evidential demand for belief in Jesus.
Like termites eating out a foundation that darned confirmation bias has eaten through the foundation of this essay. Although this is fatal to McCormick’s atheological apologetic it is a memorable lesson for the rest of us as we see first-hand the terrible havoc that unchecked confirmation bias can wreak on one’s scholarly work.
Do not be a McCormick.
3-The whole evidence thing
We now turn to the final issue: the evidentialism that lies at the heart of McCormick’s essay. Here is the relevant proposition in the argument:
(3) If your evidence for claim (a) is evidentially better supported than your evidence for claim (b) and you do not have sufficient evidence to believe (a) reasonably then neither do you have sufficient evidence to believe (b) reasonably.
My problem is not with (3) per se (although if I wanted to spend more time here I would point out some potential problems). Rather, my problem is with the assumption that rational belief about the truth of a historical reconstruction is always evidentially based. This is false.
First, let me explain something. There are two ways you can have a rational belief: non-basic and basic. The non-basic way is the evidentialist way. In this case you reason to the truth of a claim based on evidence that supports that claim. The basic way is non-evidentialist but that doesn’t mean non-rational. On the contrary, all our reasonable beliefs trace to properly basic, rational starting points.
So what kinds of beliefs are non-evidentialist starting points? Well some obvious ones include beliefs from rational intuition and sense perception. But there are many others too. For example, memory beliefs. I have a belief about what I ate for breakfast. Assuming there are no defeaters to that belief I can believe it rationally. It is important that we understand how things work here. The occasion in which I find myself holding that belief is not evidence for the truth of the claim. It is simply the reasonable occasion in which I come to hold the claim.
Another source of properly basic rational belief is testimony. For example, my history teacher tells me “The last spike in the Canadian railway was driven at Craigellachie in 1885.” Assuming I have no defeaters to that claim, I can accept that testimony in a properly basic way.
Let’s say that N.T. Wright tells me “God raised Jesus from the dead.” Is it possible that this testimony might also be properly basic, and thus fully rational to accept? Of course it is. McCormick provides absolutely no reason to think otherwise.
Well perhaps McCormick provides one reason: his own testimony to the contrary. After all, he dismisses all attempts to ground historical belief about the resurrection in a properly basic way as “arational and nonevidential”. (198) And if you think that McCormick is a credible authority on these matters this might provide you with a properly basic reason to hold the same conclusion. The only problem is that I’ve provided ample evidence that McCormick is not a credible authority in these matters.
But let me just note one more problem. Immediately following that “arational” quip he declares
“There are Wittgensteinian, Fideistic, Kierkegaardian, Tillichian, and Plantinga style approaches, among many others where, in one form or another, a straightforward appeal to the facts is not considered necessary or even important to the grounding of religious belief.” (198)
First, you have to wonder about anybody who construes Kierkegaard and Tillich as sharing an epistemological agenda. And how could anybody put anti-realist Wittgensteinian views of doctrine (e.g. those of D.Z. Phillips) alongside Plantinga’s doctrinal realism?
But this conflation of highly distinct positions is merely a warning flag. The real problem lies in McCormick’s final claim that on a properly basic view “a straightforward appeal to the facts is not considered necessary or even important to the grounding of religious belief.” Nothing could be further from the truth. The proposition “God raised Jesus from the dead” is a truth claim. If you come to believe it based on testimony then you come to believe a putative fact based on testimony. Facts are at the very foundation of proper basicality. McCormick actually should have written this: “an appeal to external evidences is not considered essential to the rationality of certain truth claims.” Yeah, so?
Let us close this review not with a sense of smugness but rather with the healthy recognition that we all struggle with the confirmation bias. Let us recognize the plank in our own eye before picking out the speck in someone else.
