Affective Alternates: Comment on Aylett and Paiva

Typologies of culture, personality, and emotion have long been debated, at least since The Birth of Tragedy by Nietzsche in 1872. In the 20th century a “culture and personality” school of thought argued that cultures tend to have modal personalities (Benedict, 1934; Opler, 1959), and the “social relations” approach sought to unite cultural anthropology and sociology with personality and social psychology (Bainbridge, 2011). Since then, attempts at synthesis have come under increasing criticism, the hard-won consensus disintegrated, and today we face a bewildering array of alternatives. The article by Aylett and Paiva (2012) takes exactly the right approach under these circumstances, specifying which theory and method it uses, counting simplicity and utility high among the selection criteria, but then clearly stating that other choices might be better in future work.

A century after Nietzsche, Homans (1967) criticized the concept of culture and the associated notion that societies are integrated wholes, at the same time that psychologists were developing models of personality that implied great individual variation along measurable dimensions within any society (Cattell, 1949). When Aylett and Paiva distinguish two kinds of empathy, they describe a personality dimension. Cognitive empathy is like the rational choice approach, in which human behavior is always based on selfish calculation of personal gain, but all successful persons develop a set of rules governing their behavior, that makes them reliable exchange partners (Axelrod, 1984). This need not result in flattened affect, because emotions can be tactics for influencing people (Frank, 1988).

As Aylett and Paiva (2012) note, cognition and emotion are not easily disentangled, and one model of affective empathy also analyzes it in cognitive terms. Many neurobiologists believe that primates possess a brain structure of mirror neurons that automatically models the behavior of another person or animal, acting like a semiautonomous reflection of the action and motivation of that other entity (Rizzolatti & Craighero, 2004). If this structure is well developed in an individual, or perhaps connected in a particular way with other brain structures, that individual will be strong in empathy, without any elaborate cognitive effort other than that performed by the mirror neurons. There is much to debate about this perspective, but the result is not necessarily sympathy, because humans used the mirror neurons to model the motivations of predators and prey in ancient times in order to combat them effectively, as well as the motivations of other humans who might become friendly partners in joint activities. Religion may reflect hyperactivity of this specialized brain activity, and individuals who are extremely deficient in it are autistic (Boyer, 2001).

The fact that Aylett and Paiva have a specific application allows them to make reasonable technical decisions and produce clear results that explore scientific territory that might have a range of other applications in future work. In so doing, they achieve progress in parallel with other researchers who seek to give artificial characters more realistic behavioral repertoires, such as curiosity (Merrick & Maher, 2009) or the common Big Five dimensions of personality (Su, Pham, & Wardhani, 2007). The PSI drives in the current model could easily be replaced by a different set of categories, such as the four motivations for game players postulated by Bartle (2004), and the cultural categories could be replaced by the five pattern variables from Parsons and Shills (1951). To note these possibilities is to praise Aylett and Paiva, not to criticize them, because their systematic, modular approach is ideal for exploring a wide range of alternatives.

If distinct cultures actually exist, then the decision to teach tolerance is a political decision, rather than a scientific one, yet it can have scientific spin-offs. Currently, social science is connected to political liberalism, but this was not the case in the past and need not be in the future. Huntington’s (1996) clash of civilizations thesis, and Sorokin’s (1937–1941) model of civilization cycles, both imply that members of any one culture should oppose members of any other. The technology used to teach tolerance could readily be adapted to teach intolerance, merely adjusting the pattern of rewards and punishments.

If distinct cultures do not actually exist, then the methods developed by Aylett and Paiva can be shifted to model personality variations instead. The question of tolerance then becomes challenging, as some personality types are typically handled with less than perfect tolerance by psychiatrists and the criminal justice system. The main application of virtual characters in the world today is combat-oriented computer games, in which many of the nonplayer characters are designed to violate societal norms, so there is value in simulating the full range of human variation.