The [obsessional search for meaning] has two main roots in the history of ideas. […] The first is the theological tradition and the problem of evil. Within Christian theology there emerged two main ways of justifying evil, pain and sin–they could be seen either as indispensable causal conditions for the optimality of the universe as a whole, or as inevitable by-products of an optimal package solution. The first was that of Leibniz, who suggested that monsters had the function of enabling us to perceive the beauty of the normal. The second was that of Malebranche, who poured scorn on the idea that God has created monstrous birth defects ‘pour le bénéfice des sages-femmes’, and argued instead that accidents and mishaps are the cost God hat to pay for the choice of simple and general laws of nature. In either case the argument was intended to show that the actual world was the best of all possible worlds, and that every feature of it was part and parcel of its optimality. Logically speaking, the theodicy cannot serve as a deductive basis for the sociodicy: there is no reason why the best of all possible worlds should also contain the best of all possible societies. The whole point of the theodicy is that suboptimality in the part may be a condition for the optimality of the whole, and this may be the case even when the part in question is the corner of the universe in which human history unfolds itself. If monsters are to be justified by their edifying effects on the midwives that receive them, could not the miseries of humanity have a similar function for creatures of other worlds or celestial spheres?

Jon Elster, Sour Grapes: Studies in the Subversion of Rationality, Cambridge, 1983, p. 102