Read Believing Bullshit: How Not to Get Sucked into an Intellectual Black Hole for Free Online Page A

—those of triangularity and four-sidedness—is sufficient to refute her claim.
    This raises the possibility that various claims about the supernatural might also be refuted without any appeal to science. Indeed, my introduction (appendix B) provided a possible example. If we understand God to be, literally, an agent —a person who acts in a rational way on the basis of his beliefs and desires, but God is (or was?) also a nontemporal being, capable of existing outside of a temporal setting, then we run into similar conceptual obstacles. The concepts of agency, action, belief, desire, and so on, are, it seems, essentially temporal concepts. Talk about a nontemporal agent or person seems, on closer inspection, to make little more sense than talk of a four-sided triangle. But if that is true, then we can refute the claim that the time is itself the creation of such an agent without any appeal to empirical science. A simple conceptual argument does the trick.
    So here's one way in which a rational refutation of a claim need not be a scientific refutation. It might be a conceptual refutation. These are, of course, the kind of arguments in which philosophers specialize.
    Empirical but Non-“Scientific” Refutation
    It also seems to me that even an empirically based refutation—that is to say, a refutation based, at least in part, on observation of the world around us—need not be “scientific.” Suppose Jim claims to have a cat stuffed inside his shirt. We carefully go around him, visually inspecting and patting every part of his shirt. We hear no “meows” and find no out-of-place bumps that might be a cat. So we conclude, reasonably, despite not having actually looked inside Jim's shirt, that there's no cat there. Wehave pretty clearly refuted Jim's claim, and have done so on the basis of empirical observation. Were there really a cat inside Jim's shirt, we would surely expect to detect some signs of its presence. If, even after careful checking, we find no such signs, we are justified in supposing there's no cat there.
    I extract two morals from this example. The first is that, while this refutation is empirically based, it would surely be odd to class it as a scientific refutation. Were we really doing science when we noted the absence of bumps and “meows” and concluded there was no cat present? Surely, this is an example of the common or garden-variety empirical refutation people have been conducting for millennia, long before the development of the rather refined and specialized tool known as the scientific method. If we conclude that it's not raining because the ground outside is not wet, or that the chicken is not cooked because the juices are not running clear, or that it can't be 8 p.m. yet because the sun is still up, those are perfectly acceptable empirically based inferences to draw, despite the fact that these inferences are not ordinarily classed as scientific. Indeed, such common or gardenvariety, everyday refutations can be just as devastatingly effective as their laboratory-based counterparts. Call them “scientific” if you like, but, given that such refutations aren't typically performed by scientists and don't involve the “scientific method,” it seems to me less misleading to describe them as empirical but nonscientific.
    The second moral I draw is that the effectiveness of such everyday refutations is not threatened by the fact that we could yet turn out to be mistaken about there being a cat in Jim's shirt. This refutation, like any empirical refutation (even of the properly scientific variety), is open to the possibility of error. It is possible, for example, that Jim has secretly been producing mute microcats. Perhaps, by a program of selective breeding, Jim has managed to get them down to just an inch or two in size, and he has one of these microcats hidden under his left armpit, where we have failed to detect it. This is a possibility. But the mere fact thatwe might be mistaken doesn't