In its lower orders at least, the intentional stance, like the design stance, saves time that might be vital to survival. Consequently, natural selection shaped brains to deploy the intentional stance as a short cut. We are biologically programmed to impute intentions to entities whose behaviour matters to us. Once again, Paul Bloom quotes experimental evidence that children are especially likely to adopt the intentional stance. When small babies see an object apparently following another object (for example, on a computer screen), they assume that they are witnessing an active chase by an intentional agent, and they demonstrate the fact by registering surprise when the putative agent fails to pursue the chase.

The design stance and the intentional stance are useful brain mechanisms, important for speeding up the second-guessing of entities that really matter for survival, such as predators or potential mates. But, like other brain mechanisms, these stances can misfire. Children, and primitive peoples, impute intentions to the weather, to waves and currents, to falling rocks. All of us are prone to do the same thing with machines, especially when they let us down. Many will remember with affection the day Basil Fawlty's car broke down during his vital mission to save Gourmet Night from disaster. He gave it fair warning, counted to three, then got out of the car, seized a tree branch and thrashed it to within an inch of its life. Most of us have been there, at least momentarily, with a computer if not with a car. Justin Barrett coined the acronym HADD, for hyperactive agent detection device. We hyperactively detect agents where there are none, and this makes us suspect malice or benignity where, in fact, nature is only indifferent. I catch myself momentarily harbouring savage resentment against some blameless inanimate such as my bicycle chain. There was a poignant recent report of a man who tripped over his untied shoelace in the Fitzwilliam Museum in Cambridge, fell down the stairs, and smashed three priceless Qing Dynasty vases: 'He landed in the middle of the vases and they splintered into a million pieces. He was still sitting there stunned when staff appeared. Everyone stood around in silence, as if in shock. The man kept pointing to his shoelace, saying, "There it is; that's the culprit." '⁸³

Other by-product explanations of religion have been proposed by Hinde, Shermer, Boyer, Atran, Bloom, Dennett, Keleman and others. One especially intriguing possibility mentioned by Dennett is that the irrationality of religion is a by-product of a particular built-in irrationality mechanism in the brain: our tendency, which presumably has genetic advantages, to fall in love.

The anthropologist Helen Fisher, in Why We Love, has beautifully expressed the insanity of romantic love, and how over-the-top it is compared with what might seem strictly necessary. Look at it this way. From the point of view of a man, say, it is unlikely that any one woman of his acquaintance is a hundred times more lovable than her nearest competitor, yet that is how he is likely to describe her when 'in love'. Rather than the fanatically monogamous devotion to which we are susceptible, some sort of 'polyamory' is on the face of it more rational. (Polyamory is the belief that one can simultaneously love several members of the opposite sex, just as one can love more than one wine, composer, book or sport.) We happily accept that we can love more than one child, parent, sibling, teacher, friend or pet. When you think of it like that, isn't the total exclusiveness that we expect of spousal love positively weird? Yet it is what we expect, and it is what we set out to achieve. There must be a reason.

Helen Fisher and others have shown that being in love is accompanied by unique brain states, including the presence of neurally active chemicals (in effect, natural drugs) that are highly specific and characteristic of the state. Evolutionary psychologists agree with her that the irrational coup de foudre could be a mechanism to ensure loyalty to one co-parent, lasting for long enough to rear a child together. From a Darwinian point of view it is, no doubt, important to choose a good partner, for all sorts of reasons. But, once having made a choice — even a poor one — and conceived a child, it is more important to stick with that one choice through thick and thin, at least until the child is weaned.

Could irrational religion be a by-product of the irrationality mechanisms that were originally built into the brain by selection for falling in love? Certainly, religious faith has something of the same character as falling in love (and both have many of the attributes of being high on an addictive drug*)). The neuropsychiatrist John Smythies cautions that there are significant differences between the brain areas activated by the two kinds of mania. Nevertheless, he notes some similarities too:

I made the comparison between falling in love and religion in 1993, when I noted that the symptoms of an individual infected by religion 'may be startlingly reminiscent of those more ordinarily associated with sexual love. This is an extremely potent force in the brain, and it is not surprising that some viruses have evolved to exploit it' ('viruses' here is a metaphor for religions: my article was called 'Viruses of the mind'). St Teresa of Avila's famously orgasmic vision is too notorious to need quoting again. More seriously, and on a less crudely sensual plane, the philosopher Anthony Kenny provides moving testimony to the pure delight that awaits those who manage to believe in the mystery of the transubstantiation. After describing his ordination as a Roman Catholic priest, empowered by laying on of hands to celebrate mass, he goes on that he vividly recalls

The equivalent of the moth's light-compass reaction is the apparently irrational but useful habit of falling in love with one, and only one, member of the opposite sex. The misfiring byproduct — equivalent to flying into the candle flame — is falling in love with Yahweh (or with the Virgin Mary, or with a wafer, or with Allah) and performing irrational acts motivated by such love.

The biologist Lewis Wolpert, in Six Impossible Things Before Breakfast, makes a suggestion that can be seen as a generalization of the idea of constructive irrationality. His point is that irrationally strong conviction is a guard against fickleness of mind: 'if beliefs that saved lives were not held strongly, it would have been disadvantageous in early human evolution. It would be a severe disadvantage, for example, when hunting or making tools, to keep changing one's mind.' The implication of Wolpert's argument is that, at least under some circumstances, it is better to persist in an irrational belief than to vacillate, even if new evidence or ratiocination favours a change. It is easy to see the 'falling in love' argument as a special case, and it is correspondingly easy to see Wolpert's 'irrational persistence' as yet another useful psychological predisposition that could explain important aspects of irrational religious behaviour: yet another by-product.