John Derbyshire takes note of a study showing that large percentages of Americans believe in ghosts, angels, demons, and so forth, and writes:

Nothing very surprising there. It's interesting to note one's own reactions on reading a news item like that. The reaction must, I suppose, be personality-dependendent. It will fall somewhere in a range from:  "I am surrounded by idiots!"  to:  "What an oddball freak I must be!"  (I find myself closer to the latter end of that spectrum.)

News items like this also raise the cog-sci questions that our Mr. Hume is good at:  What do people actually mean by any of this?  Do they actually conduct their lives on the working assumption that the next stranger they meet may be an angel, a ghost, Satan, or a UFO crewman? (Ans: Obviously not.)  How many could give a coherent account of the theory they reject? (Ans: Vanishingly few.)  What does "believe" actually mean in this context? (Ans: Nothing very functional.)

Well, it depends on how you define "functional." I doubt that very many people who believe in angels and devils play an "is he a demon in disguise?" guessing game every time they meet a new co-worker. But having spent a fair bit of time around such people (and, well, being one myself), I'd argue that belief in the existence of supernatural beings does actually tend to exert a not-inconsiderable pull on the way that people interpret and respond to life - both to their ordinary experiences, psychological and otherwise, and to the supernatural/numinous experiences that are reasonably common features of human existence.

Now the relative importance of a belief in supernatural agents, and the extent which it intrudes on day-to-day affairs, depends, as you might expect, on the specifics of the belief system in question, and the particular emphases it places. For instance, Evangelicals and (especially) Pentecostals tend to invoke demonic agency much more frequently than Roman Catholics, at least in my experience (and in the American Catholic Church; things are rather different elsewhere). Thus while Hollywood's exorcists invariably come clothed in priestly habits, actual exorcism-prayer of various sorts - the kind of "spiritual warfare" that freaks Christopher Hitchens out so much - is a far more important part of religious practice among American Pentecostals than among American Papists, for whom the demonic is something you invoke to explain and understand extraordinary experiences, not more mundane body-and-soul crises like, say, alcoholism or drug addiction. (Though of course Catholics frequently invoke saintly agency, which is its own category of supernatural belief, in ways that members of other Christian traditions don't.)

But even extraordinary happenings aren't, well, all that extraordinary. Religious belief exists and persists in part because religious experiences exist and persist - even if they're far from universal, as Derb will be happy to inform you - and in existing and persisting seem to cry out for an explanation. And many of the numinous encounters that people seek to explain, both to others and to themselves, don't fall into the "oneness with the universe" category that gets the students of brain states and meditation so excited: They're often weirder than that, and often darker. Like, say, the childhood experience of the British writer Hilary Mantel:

When Hilary Mantel was seven, she met the devil. Well, not exactly: but she did encounter something so evil that even now she finds it hard to explain what she chanced on in the garden. "I couldn't say I saw it," she says slowly. "I'm talking about something at the very border of sensory experience. I could walk to where it was, could say how high it was and describe the speed at which it moved. But how I got the information, through which sense, I don't know." What Mantel does know is that she felt she had witnessed something she wasn't meant to. "The experience was absolutely destroying, as if my body was falling apart at a cellular level, which expressed itself in intense nausea. The way I rationalised it was that it was the devil. As a Catholic, that was the theology I had at my command." The family home itself was haunted and this presence seemed like "a concentration of things that were going on in the house - the unhappiness of our family and the pressure of secrets and lies."

As her language suggests, belief in the demonic, at the most basic level, is a way of explaining and rationalizing a certain category of human experience. There are other ways to rationalize such things, obviously: If you're a hard-bitten materialist who has a Mantel-style encounter ("It is as high as a child of two. Its depth is a foot, fifteen inches. The air stirs around it, invisibly . . . It has no edges, no mass, no dimension, no shape except the formless; it moves."), you may either find a non-Catholic, non-supernatural way of explaining what you've just experienced, or else simply profess agnosticism about what, if anything, it means. (For a fascinating example in this vein, see the staunch atheist A.J. Ayer on his own near-death experience.) But if you want to understand what, if anything, a person means when he says he believes in demons or angels or ghosts, the simplest baseline answer is this: He means that if confronted with an encounter or an experience that seems demonic or ghostly or angelic and asked to rationalize it, he will be inclined to give credence (like the seven-year-old Mantel) to the possibility that the encounter is, in fact, what it appears to be.