Since positive illusions are the one example McKay & Dennett (M&D) find of a bona fide contender for an adaptive evolutionary favored epistemic disability, my view (Flanagan Reference Flanagan1991; Reference Flanagan2007) that positive illusions may not be a class of well-behaved misbeliefs at all should matter.

Start with this conditional: If there are positive illusions and if they are, as the psychologists say they are, (a) common, plus correlated with (b) moral decency and (c) happiness, positive affect, optimism, as well as with (d) the capacity to engage in profitable, creative, productive work, then there is a problem with the view that flourishing requires, or demands, overcoming the tendency to harbor false beliefs. The reason is simple: The normative claim that contemporary people flourish truly only if they live in the light of the true, is in competition with the psychologist's claim that the capacities to love, work, and be happy are enhanced by false belief.

The conditional that causes the latter problem – competition between the ends of flourishing and fitness – and that also, albeit independently, warrants M&D's ingenious explanation for why there are positive illusions, involves accepting that there are positive illusions. But we can challenge sensibly the antecedent of the conditional.

Accept that “positive illusions” are states of mind that benefit the consumer, but reject the claim that they (all, most, many) are best interpreted as involving false beliefs, as opposed to having positive expectations and hopes – in other words, a positive attitude. Hopes and a “can do” attitude need not require false belief. “Exaggerated” and “unrealistic” are adjectives used to describe the whole set of allegedly questionable epistemic states of mind in Taylor and Brown's famous meta-analysis (Taylor & Brown Reference Taylor and Brown1988), which include unrealistic positive evaluations, exaggerated perceptions of control and mastery, and unrealistic optimism. One ought to worry about inferring false beliefs from (even) correct ascriptions of lack of realism or exaggerated views about one's powers and abilities.

When Muhammad Ali famously remarked before his final fight with Smokin' Joe Frazier – the “Thrilla in Manila” – that “It will be a killa … and a chilla … and a thrilla … when I get the gorilla in Manila,” did he believe that he would kick Smokin' Joe's ass? Or is he best understood as doing something, performing an action that was, in effect, part of the fight before the first bell sounded? Both boxers presumably believed that they could win and hoped that they would win. So far there is no epistemic mistake regardless of outcome. “Can” does not entail “will.” The epistemic standards governing hopes, desires, and the like, are different from those that govern beliefs. It would be very odd to say that the losers in zero-sum games always have false beliefs. In fact, Ali might have believed that he could not win unless he made Smokin' Joe worry that he might know how to beat him. Furthermore, Smokin' Joe need not have believed he would lose after Ali's provocation. The effect might work this way: Ali knows how to do things with words. He speaks with the intention of undermining Smokin' Joe's confidence, and does so. In this case, the mechanisms at work do not operate via beliefs at all, although they might commonly be assimilated to that class of states specified as folk psychologically. The point is that many things we do with words and thoughts can be viewed as strategic – engendering self-confidence, undermining the competitor's abilities – and not as straightforwardly epistemic.

Consider this from Aristotle:

Interpreted one way Aristotle can be read as encouraging two false beliefs; interpreted another way he can be read as encouraging an attitude that one can achieve something excellent if one sets one's eyes on the goal. Coaches often speak this way to their charges. A professional tennis match always produces one winner and one loser. Both players, if they are any good, go into the match believing that they can win, indeed that they will win. Believing one can win is a true belief. Hoping that one will win is a sensible expectation. In neither case is there a mistake.

A sensible counterfactual test for whether a person in fact holds a belief or is in some associated epistemic state in a strong and objectionable way would be: Does the state-in-question yield, and if so how quickly, easily, and so on, when there is strong countervailing evidence? If I get prostate cancer, or divorced, or in a motorcycle accident despite saying that I think I won't, I will quickly yield my initial thought or claim that these calamities will not befall me. Taylor and Brown (Reference Taylor and Brown1988, p. 197) write that “the extreme optimism individuals display [about such probabilities] appears to be illusory.” This is not obvious. Optimism can be unrealistic, perhaps – illusory is a different matter.

The overall point is that the positive illusion literature conflates and assimilates systematically such states as hopes, expectations, and positive attitudes with states of false belief, when the charitable analysis need not involve attributing any belief at all, let alone a false one. One final point: There is reason to believe that one class of alleged “positive illusions,” self-serving ones, is not common outside of the West (Flanagan Reference Flanagan1991; Reference Flanagan2007; Heine et al. Reference Heine, Lehman, Marcus and Kitayama1999). If so, this might cause trouble for M&D since there is no common phenotypic trait to explain it as an adaptation.