In a classic scene from the Woody Allen movie, Everything You Wanted to Know About Sex (But Were Afraid To Ask), sperm congregate in a holding area awaiting ejaculation. One sperm, played by Allen, is gripped by existential doubt as he contemplates his impending odyssey into the great unknown. As humorous as Allen's dilemma is, imagine his character's distress if he had paused to consider his odds of successfully fertilizing an egg (roughly 1 in 40,000,000, assuming ovulation). These odds would surely shake the confidence of even the most Panglossian spermatozoon, let alone Allen's anxiety-ridden schlemiel.

Of course, sperm do not calculate probabilities. But had nature endowed them with the ability to do so, she would have needed to similarly endow them with the ability to inflate their own likelihood of success. Otherwise, they, like Allen, would be paralyzed by the reality they were about to confront.

Extrapolating from Hollywood cinema is obviously hazardous, and spermatozoa are not people (although both function to pass their genetic material to the next generation), but Allen's scene touches on two important questions: Are people positively biased in their beliefs, and are these beliefs ultimately beneficial? In 1988, Shelley Taylor and I examined research relevant to these questions and offered two conclusions. First, when it comes to self-relevant beliefs and appraisals (e.g., “How kind am I?” “How capable am I?” “How bright is my future likely to be?”), people are positively biased (Taylor & Brown Reference Taylor and Brown1988; see also Brown Reference Brown1986; Reference Brown, Snyder and Forsyth1991; Reference Brown, Baumeister and Vohs2007; Taylor Reference Taylor1989; Taylor & Brown Reference Taylor and Brown1994a; Reference Taylor and Brown1994b). On virtually all positively valued attributes, most people view themselves in unrealistically (though not excessively) positive terms. Second, we argued that these positive illusions are ordinarily beneficial. Under normal circumstances, people who entertain moderately (though not excessively) positive self-beliefs fare better on measures of achievement, adjustment, and physical well-being than those who are less positively biased. Certainly, there are limits to the benefits positive illusions provide (Baumeister Reference Baumeister1989; Dunning et al. Reference Dunning, Heath and Suls2004), and we never claimed that the more biased one is, the better off one is going to be (see also, Marshall & Brown Reference Marshall, Brown and Chang2007). Instead, our claim was simply that (a) most people view themselves in overly positive terms and (b) under many – if not most – conditions, these beliefs are beneficial.

In their target article, McKay & Dennett (M&D) echo these arguments, concluding that positive illusions provide the firmest evidence for evolved misbelief. From this perspective, natural selection favored those whose self-perceptions were positively biased. In sympathy with this conclusion, my colleagues and I have found consistent evidence that positively biased self-perceptions are a pervasive, cross-cultural phenomena (Brown Reference Brown2003; Brown et al. Reference Brown, Cai, Oakes and Deng2009; Brown & Kobayashi Reference Brown and Kobayashi2002; Reference Brown and Kobayashi2003; Cai et al. Reference Cai, Brown, Deng and Oakes2007; Reference Cai, Wu and Brown2009; Kobayashi & Brown Reference Kobayashi and Brown2003).

At the same time, I think the target article would have benefited by taking a broader view of positive illusions. An exclusive focus on people's self-enhancing beliefs (e.g., “My commentary is more insightful than most other commentaries”) ignores the myriad processes that conspire to produce and sustain them (Brown Reference Brown, Snyder and Forsyth1991). In most instances, positive illusions are the downstream product of an extensive system of information-processing biases and selective affiliation patterns. Insofar as these biases and patterns generate and perpetuate adaptive illusions, they may also be products of natural selection.

Numerous cognitive processes, such as self-serving attributions, idiosyncratic trait definitions, and biased judgments of a trait's importance sustain positive illusions (for a review, see Brown Reference Brown1998), but interpersonal processes ordinarily produce them. For the most part, people believe positive things about themselves because they receive mostly positive feedback from the people they spend most of their lives with (Murray et al. Reference Murray, Holmes and Griffin1996). In this sense, positive collusions produce positive illusions.

Positive collusions rely on two interrelated processes. First, people's self-enhancing biases include aspects of what William James (Reference James1890) called the “extracorporeal material self.” The extracorporeal material self refers to everyone and everything we call “mine” or “my.” With respect to positive illusions, we exaggerate not only our own virtues, but also those of our friends, neighbors, colleagues, family members, and loved ones (Brown Reference Brown1986; Reference Brown, Snyder and Forsyth1991; Brown & Kobayashi Reference Brown and Kobayashi2002). Positive collusions begin the moment we are born. Most (though certainly not all) parents view their infants in overly positive terms, believing their offspring are cuter, smarter, and more socially advanced than are most other infants. As children grow, they internalize these biased evaluations, producing the well-known “better than average” effect (Alicke Reference Alicke1985; Brown Reference Brown1986).

It is hardly surprising that parents view their infants through rose-colored glasses; what is surprising, however, is just how tenuous the self-other connection can be in order for this positivity bias to emerge. Research on in-group favoritism in the minimal group paradigm (Billig & Tajfel Reference Billig and Tajfel1973; Tajfel et al. Reference Tajfel, Billig, Bundy and Flament1971) makes this point most graphically. In these studies, people are arbitrarily divided into groups on some patently trivial basis (e.g., they drew a blue marble from a bag instead of a red one). Despite the meaninglessness and obviously arbitrary nature of this designation, people view their fellow in-group members in more positive terms than out-group members (Brewer Reference Brewer1979). In short, anything or anyone that is part of “me” is viewed in more positive terms than anything or anyone that is “not me.”

How do these effects sustain people's beliefs in their own capacities? Once we forge an association with someone (e.g., make a friend; join a club; select a mate), we become part of that person's extracorporeal self and reap the self-enhancing benefits the association provides (i.e., we receive feedback that we are more likable, capable, and charming than are most other people). In this fashion, mutual admiration begets mutual benefits.