Toward a science of memetics consider a phenomenon I will call coercive perception. "That's not a vase, that's an old woman's profile." "That's not a sword, that's a phallic symbol." "She's cheating on you." "That cloud's shaped like a rabbit!" "That's not an idea, that's the false consciousness of the bourgeois." Or, of course, "That's not a belief, that's a meme."
A perception is coercive when simply understanding it reorients you. Understanding is sufficient; belief is irrelevant. The coercion is instantaneous, irreversible, and permanent; it is seen and cannot be unseen. You cannot be argued out of it, because you never believed it; but you may act as though you believe it, because you cannot forget it. This sounds terrible in the abstract; but in practice it is something we value and seek out. Reading horror stories will coerce your perception of small quiet backwoods towns, of quiet staring backwoods people, of blackletter books and remote silent wastes, to an atmospheric unity. Being coerced this way is pleasant, despite its unpleasant content. But then imagine a young girl or boy, and a sort of friend who says cruelly—"Did anyone ever tell you look like ———? Look at this picture—can't you see it?" And of course they see it, they have been coerced to see it, and they will always see it, no matter how absurd and wrong they know it is.
This kind of perception is unique to human beings. It is not the substance of human difference, but might make a good test for it—better than the silly Turing test, which even Eliza has passed. So it is supposed to be intelligent—can its perceptions be coerced? Does its intelligence close over its perceptions? Correct perception is no test of intelligence. A mirror perceives correctly; what only perceives correctly is no more than a mirror. What cannot misperceive cannot think; what cannot be coerced in its perceptions cannot communicate.