In 2018 I posted on SO a review of Tim Crane’s book The Meaning of Belief: Religion from an Atheist’s Point of View:

Crane argues that atheists have largely misunderstood religion by regarding it as a sort of cosmological hypothesis, one that makes insupportable claims about the creation of the universe via the supernatural acts of a divine agent. By thus construing religion as a sort of spurious proto-scientific cosmology, atheists justify relegating it to the bin of irrelevance and irrationality. However, says Crane, religion should not be seen as any sort of hypothesis, but rather as consisting of the “religious impulse” and “identification.” The religious impulse is the drive to recognize a transcendent order that is both factual and normative. God is posited as real and his will is taken as defining right and wrong. “Identification” is the desire to belong to a community that defines itself in terms of a set of beliefs and practices and which understands the world in terms of those beliefs and practices. What unites these two elements is a shared experience of the sacred, which promotes a strong sense of identity. Atheists miss these points by dismissing religion as a crackpot cosmology and religious believers as superstitious.

In my comments on Crane’s claims, I note that if atheists are mistaken in regarding theism as a quasi-scientific hypothesis, this is not a gratuitous error, but is due to the fact that leading religious apologists defend theism as such a hypothesis. Defenders of “intelligent design” theory such as William Dembski and Michael Behe present their concepts of “specified complexity” and “irreducible complexity” as scientifically legitimate concepts. In The Existence of God, Richard Swinburne employs Bayesian confirmation theory in defense of his theistic hypothesis and appeals largely to the criterion of simplicity, which, of course, is a standard of theory choice in the natural sciences. William Lane Craig’s Kalaam cosmological argument is developed and defended in the context of physical cosmology. These considerations seem to justify the characterization of the theistic hypothesis as “proto” or “quasi” scientific.

However, such a designation is not really important. The important point is that theism is defended as a hypothesis. Whether that hypothesis is classified as “scientific,” “quasi-scientific,” or “metaphysical” is not of primary importance. In my review I make the point that, as John Hick argues in An Interpretation of Religion, the reasoning underlying religious  belief is primarily interpretive and not hypothetical. Hick says that the universe is religiously ambiguous in the sense that there are no facts that compel a religious or a naturalistic interpretation. The arguments for and against the existence of God are not compelling, and their conclusions may be reasonably rejected. Perfectly reasonable people may therefore disagree about the existence of God.

If Hick is right, what follows? Perhaps both atheists and religious apologists should cease their efforts to devise polemical weapons to bludgeon the other side into submission since we should know by now that this will not work. We should instead seek a more nuanced and informed view of belief and unbelief. We might actually learn something from each other!

In a 2018 podcast of “Reasonable Faith,” Kevin Harris interviews Professor Craig about Crane’s book and my review of it:

Jeff Lowder drew my attention to this just recently, and I would like to respond to it here.

Professor Craig argues that, while theistic hypotheses are explanatory, it is “tendentious and inaccurate” to characterize them in general as “semi-scientific” or “proto-scientific.” Craig does admit that the ID theorists regard their hypothesis as scientific. However, they claim that their arguments for intelligent design are religiously neutral, so I err in identifying this hypothesis as a specifically religious or theistic hypothesis.

ID theory is religiously neutral? How can that be when it was developed and promoted explicitly as part of an aggressive apologetic program? Well, to avoid church/state entanglements, ID theorists note that the designer could be something other than the God of Christian theism–something like Plato’s Demiurge, or the “Q” Continuum from Star Trek, maybe. This lawyerly ruse has no bearing on the philosophical issue, however. Could the designer be God? Of course. The most charitable reading of ID is therefore that it is an argument for a disjunction of mutually exclusive and exhaustive designer hypotheses, including the theistic hypothesis as one disjunct.

As for Swinburne’s and his own hypothesis, Craig says that they are not scientific or quasi-scientific because they posit a personal cause rather than a naturalistic one. Scientific explanations are in terms of natural laws and initial conditions, but theistic hypotheses posit a personal agent who creates by acts of volition. However, it certainly seems that, in principle, there could be scientific confirmation of a personal cause. Suppose, for instance, that the famous Hubble image of the Eagle Nebula—the “pillars of creation”—were accompanied by glowing gas in the form of Hebrew letters, light years wide, proclaiming “I, Yahweh, did this.” In this case, we would have outstanding scientific evidence of a personal cause. So, as a general demarcation criterion, the personal/impersonal distinction does not work.

Craig and Harris then have this exchange:

KEVIN HARRIS: Just to be more specific, when he [me] mentions you here, again, he says, “Craig’s Kalaam argument is specifically and explicitly a cosmological claim presented within the context of physical cosmology.”

CRAIG:Right. And it doesn’t appeal to a theistic cosmology or an alternative to contemporary cosmology. It appeals to the normal cosmological model that is affirmed by secular scientists. So it is not in any way positing God as a scientific or quasi-scientific hypothesis.

Craig’s statement here is a non sequitur. A scientific theory need not be an alternative to another theory, but could subsume it. Theory T₂ subsumes theory T₁ when T₂ provides a deeper and more inclusive explanatory framework that accounts for T₁’s empirical success within its domain while locating that domain within a larger one that T₂ covers. Advances in science often occur when a new theory does not just replace an old one, but places the old theory in a broader and deeper explanatory context. Thus, Carnot’s theories were subsumed by the thermodynamics of Kelvin and Clausius. Craig’s theistic hypothesis appears intended to provide a deeper and more inclusive explanation than physical cosmology. Physical cosmology is not falsified by Craig’s theistic hypothesis, but rather is subsumed by it. Craig’s theistic cosmology aims to go beyond physical cosmology and tell us why there is a universe at all. So, the fact that Craig does not present his hypothesis as an alternative to physical cosmology, but intends to provide a deeper context for it, does not disqualify it as “quasi-scientific.”

However, since nothing much really turns on it, let’s concede the point for the sake of argument and say that Craig’s hypothesis is a “metaphysical” hypothesis rather than a “scientific” or “quasi-scientific” one. The real problem identified by Crane is that religious belief is identified as any kind of hypothesis. Crane implies and Hick argues that the reasoning underlying religious belief is interpretive rather than hypothetical. That is, the reasoning supporting a religious worldview is more like understanding a text than confirming a hypothesis. We do not understand a text by confirming piecemeal hypotheses about its meaning. Rather, we seek a reading that will give us the most coherent understanding of the text as a whole. Likewise, for religious people, their faith is what, for them, makes the most coherent and comprehensive sense of their total experience. Nothing compels such a judgment; it is inevitably personal and subjective, but not unreasonable. Similarly for atheists. Nothing compelled me to become an atheist. Rather, a naturalistic worldview is the honest and authentic articulation of my total experience and knowledge.

Craig objects that if Crane is right, then he, Swinburne, Steve Meyer, William Dembski and other defenders of religious hypotheses must misunderstand religion, which he regards as implausible.

Craig does not reply to Hick’s view directly, but chiefly expresses surprise that I have supposedly so softened my view of theism that I am now willing to endorse Hick’s view that religious belief can be as rational as naturalism. (n.b., Actually, I have always regarded some religious belief as rational and some definitely not.) What, then, do I have against the apologetic enterprise that he represents? Why do I harshly characterize it as an attempt to “bludgeon” opponents into submission? After all, he is only trying to show that his belief is rational and not to show that atheists are irrational. Why do I persist in seeing the apologetic enterprise as coercive, i.e. as an effort to show not just that their belief is justified, but that mine is not? That is not his aim at all.

I honestly do not know what to make of Craig’s claim here. Does he regard his Kalaam argument as a refutation of atheism? I cannot read his presentation and defense of that argument in any other way. In this case, the argument is not a modest claim about what he is justified in believing, but the much stronger and more aggressive claim that atheism is demonstrably false and groundless. In other words, he seems to be arguing that he is right and that atheists are dead wrong. Atheists, of course, have often argued that they are right and that Craig is wrong. The debate between apologists and atheists therefore does appear to have an oppositional and aggressive character; it is not about what one may believe but what others must believe. However, if I have been misreading Craig all these years, and his aim all along has only been to affirm the rationality of his view and not to debunk mine, then I would suggest that Hick’s position provides a much better basis for such a softer and gentler apologetic.

Finally, Craig invites listeners to look at my debate with him on the existence of God to see if I did indeed effectively criticize his theistic arguments. I also would like to extend that invitation. (I think that Craig is referring to our debate at Indiana University in February 2002, not the earlier one at Prestonwood Baptist Church.).