A key issue in the debate around atheism concerns what happens at a cognitive level when we say we believe something. Sam Harris, one of the main voices of the "new atheism" and a researcher in the field of cognitive studies of religion, based his earlier diatribes against religious belief on the explicit and unsupported claim that there is only one cognitive structure to all beliefs: that of a proposition. As he puts it, "Believing a given proposition is a matter of believing that it faithfully represents some state of the world." In other words, all beliefs correspond to the form "I believe that S is P," and thus can be evaluated as truth claims based on the availability of evidence for the claim that, indeed, S is P. In light of this position, Harris has consistently defined "faith" as belief without evidence. When societies are dominated by people of faith, people who are compelled to believe unreasonable propositions about the world without evidence, the result is a recipe for intolerance, injustice, and violence.
I should quickly interject here that this impulse to violence is, at the very least, shared by Harris. As he writes, "There is, in fact, no talking to some people. If they cannot be captured, and they often cannot, otherwise tolerant people may be justified in killing them in self-defense." The justification for such intolerance from otherwise tolerant people is that, as people of faith, by definition they fall outside of a sphere of civil conversation that other, normal people, inhabit.
Now, in Harris's estimation some 240 million of the currently 300 million inhabitants of the United States fall into the category of people with whom there is just no talking, by virtue of their responses to surveys asking about their religious beliefs. He is also less willing to grant what fellow atheist Richard Dawkins suggests, that a good portion of those responding to such surveys don’t really mean it.
While I am not proposing to undermine such survey responses in the same way Dawkins does, I do in fact suggest that there is a considerable spectrum of possible meanings to the claim to believe that is attributable to so many US citizens. The mere fact that a majority of US Catholics fail to follow the Church's teaching on contraception deeply undermines this conception of implicit unanimity in the way believers believe. More recent research in the neurology of religious belief, including articles co-authored by Harris, has demonstrated perceptible differences in brain activity between religious and non-religious test subjects, and research conducted by Andrew Newberg and others has revealed different brain activity within the same subject when that subject focuses on religious as opposed to non-religious thought contents.
The conclusion for my work is quite straightforward: we have no justification at all, and indeed much evidence to the contrary, that all beliefs share the same cognitive structure. But if this is true, then the entire justification for the attack on all religious beliefs as if they were co-responsible for the evils of fundamentalism (a core point of agreement among the new atheists) simply vanishes. As I argue in An Uncertain Faith, it may well be that various forms of non-fundamentalist or moderate religious belief are far more antagonistic to fundamentalist thinking than all the bombast and aggression the new atheists can muster.