Having defended Richard Dawkins, one of our greatest God-bashers, in the first of these columns, it seems only fair in the second to consider a flawed species of argument from the other side of the debate.
Johann Hari spoke for many when he wrote in the Independent in May that “faith is a dangerous form of bad thinking – it is believing something, without evidence or reason to back it up. Where does that end?” I have every sympathy with the rhetorical thrust of his questioning and its implicit critique of extremism and dogmatism. Yet Hari is guilty of a common epistemological error, or at least of an epistemological crudity. Which is to say that his argument is a misrepresentation of the nature of knowledge and of what it is possible for us to know.
It is true that many beliefs can and should be tested against what we know of the world. I do not, for example, consider it reasonable to claim that anyone has ever come back from the dead, or that the world is ringed by a giant serpent, or that smallpox represents a form of divine possession—yet people have at some stage in history had faith in all these things. But I also do not consider it reasonable to claim, as Hari does, that any line of thought not based in “evidence or reason” is a bad one. Where will this end? If every claim that cannot be verified against empirical evidence is to be dismissed, we will soon find ourselves having to do without much of morality, aesthetics and metaphysics—effectively taking the position that there is nothing of importance to be said about many of the questions that most people spend much of their lives worrying about. This, surely, is a rather limited definition of importance.
As Hume argued, induction—the basis of the scientific method—is a matter o…