Three of the biggest points atheists claim in the cumulative case for their worldview are religious pluralism and non-belief, previously failed supernatural explanations, and the ruthless progress of science.
Let’s look first at religious pluralism. How can belief in God be justified (or rational) when people believe different things about God, and some people don’t believe in God at all? Is not diversity of religious belief (and non-belief) better expected on atheism? As it turns out, no.
Many forms of confusion exist about this. The first arises from people conflating theism with Christianity. For even if religious pluralism were evidence against Christianity—and superficially, it might seem to be—how is it even superficially evidence against the existence of God? Answers to this question are hard to find, especially since religious pluralists—who believe in God—frequently use pluralistic belief in their favor. Perhaps, they suggest, God wants us all to have our own unique path to finding him. In other words, perhaps all religions are false (in a sense) but useful in our groping toward the transcendent or developing morally. Obviously, Catholics do not believe this, but we are evaluating theism versus atheism, not Catholicism versus atheism. One step at a time. (For the cumulative case for Catholicism, see here.)
However, Catholics and Protestants have resources to explain pluralism and non-belief as well—namely, the fall, the noetic effects of sin, demonic activity, and so forth. And surely if one is going to see whether religious pluralism can be explained in relation to Christianity, he must consider what explanatory resources Christianity has to offer and cannot ignore important doctrines such as those.
But what of the more substantial analysis? I argue that what is more surprising is not a plurality of religious beliefs given the existence of God, but the existence of any beliefs given the non-existence of God. Belief requires mental activity (particularly conceptual ideas), and mental states are scandalously resistant to physicalist explanations. Arguments against physicalist accounts of consciousness (and rationality) are beyond the scope of what can be defended here (see here, here, and here), but it is surely far too quick for atheists to claim a data point in their favor (a particular belief state, say) when a strong case can be made that only the existence of God can make possible such a data point (any belief state) in the first place. The more fundamental analysis invites reversal. Theism better predicts—that is, would lead us to expect—conscious, belief-holding agents.
What about failed supernatural explanations in the past and the ruthless progress of science? A brief response must suffice. First, just because some or even many prior supernatural explanations have failed, that does not mean that all supernatural explanations have failed. From near death experiences to Fatima, there is no short number of phenomena that are, again, highly resistant to any plausible naturalistic hypothesis and far better expected if theism were true than atheism. One only needs to take an honest look. Plus, abuse does not negate proper use, and if people were hasty to attribute to the supernatural what could be adequately explained by the natural, this doesn’t mean supernatural explanations are illegitimate; the lesson is simply to be more patient in exhausting natural causes. Finally, the progress of science is not of itself a point for atheism because science itself presupposes many data points that are better explained by theism (conscious embodied rational agents, cognitive faculties oriented toward truth, the principle of sufficient reason, natural substances that display regular causal patterns, a stable and intelligible universe, etc.).
Let us take just two points mentioned above: 1) that we are in possession of cognitive faculties naturally oriented toward truth (and generally good—even if not infallible—at getting it) and 2) the principle of sufficient reason. I say these are a requirement of science because science depends upon our ability to know, and our ability to come to know depends upon these prior conditions, which themselves are better expected—if not only possibly explained—if God exists, in which case scientific knowledge is better expected and more probable on theism than atheism. As Alvin Plantinga and others have argued, naturalism gives us reason not to automatically assume that our cognitive abilities would be naturally inclined toward truth or generally good at getting it, since falsehood may prove just as useful for survival—hence, the probabilities surrounding the reliability of our belief-forming mechanisms are either low or inscrutable (this is further supported by computational evolutionary psychology), which causes an epistemological nightmare for naturalists—namely, even if naturalism were true, one could never be justified in saying he knows naturalism is true. However, if God exists, then there is good reason to believe he made us to come to know the truth about things (even if it is not always easy), which ultimately means coming to know God. Thus, if we better expect to be scientific knowers if theism were true, and if we are convinced we are scientific knowers, this favors theism.
Next, the principle of sufficient reason (PSR) states that contingent beings (anything that exists but could have not existed—for example, electrons, cats, Donovan McNabb, etc.) have a cause distinct from themselves, in which case contingent beings can be made intelligible via a causal explanation. This is required for knowledge because if the PSR were false, then something having “no explanation” becomes just as competitive as any causal explanation. Maybe human evolution is true, or maybe (if the PSR is false) something lizard-like walked into a swamp and—for no reason whatsoever—two human beings popped out and began to procreate. Obviously, this “no explanation” possibility would be ruinous to human inquiry—not to mention fundamentally absurd. However, once the PSR is admitted, one is ineluctably led to theism because in order to explain why any collection of contingent facts obtain, one must admit (to avoid circularity) at least one necessary fact, and it is only a few short moves from necessary fact to God.
In short, science is impossible if we are not knowers, and the condition for being knowers is far better expected and explained on theism than atheism. Thus, given that we are knowers who engage in science—fruitfully, indeed—this fact favors theism, significantly, if not decisively, over atheism.
It strikes me that most if not every point advanced by atheists in their cumulative case is only superficially favorable for atheism but actually favorable for theism. Deeper analysis of each consideration invites reversal. The cumulative case for atheism fails, and this is not surprising; in fact, it is just what we should expect if atheism is false and theism is true.