Uncanny sociocultural categories.

Abstract

Humans are well adapted to their social environments. Experimental evidence suggests that humans are either born with, or quickly learn, the necessary affective and cognitive processes that allow them to recognize others, and to understand their mental states and social behavior. Mori's (1970) proposal of an uncanny valley, which describes affective response as a function of distance from a human category defined by morphological and behavioral features (i.e., human likeness), appears to be a sensible extension of these ideas. Following Mori's initial proposal, the uncanny valley has largely been considered in the context of cultural artifacts such as robotics, prosthetics, toys, and puppets. He associated “healthy people” with the greatest level of familiarity and positive affect, prosthetic hands and corpses with a global negative affective maximum, and bunraku puppets and humanoid robots with intermediate levels of familiarity and positive affect. It is important to note that these cultural artifacts represent the most contemporary features of human societies. The uncanny valley likely depends on extensions of prepotent responses to stimuli via general learning mechanisms (e.g., face recognition; Haxby et al., 2002; Sperber and Hirschfeld, 2004). Empirical studies of the uncanny valley have just begun to explore the authenticity of Mori's proposal. Contemporary studies examining the uncanny valley hypothesis have drawn heavily on the psychological literature to explain these phenomena. The shift from an account of the uncanny valley based on a dimension of human-likeness to that of categorization and frequency-based exposure (Burleigh and Schoenherr, 2014) suggests that classes of cultural artifacts might provide evidence for the ubiquity of phenomena across cultures and within human history. These social representations can become external representations available to all members of a human group and can thereby increase familiarity and anchor human judgments (Moscovici, 1981). Supporting Mori's initial claim, negative responses are the result of a lack of familiarity (e.g., Burleigh and Schoenherr, 2014) that emerge over the course of development (Lewkowicz and Ghazanfar, 2012) as humans' affective systems have yet to adapt to these artifacts. If the uncanny valley does have a general cognitive basis, then evidence from affective, behavioral, and cognitive paradigms should exist both across cultures as well as within human history. These social representations will consequently affect observers' judgments. Human and non-human classification Folktaxonomies and covert categories A particularly compelling source of evidence for the uncanny valley comes from research into folktaxonomies. When we encounter an organism, our knowledge of folkbiological categories can cause us to classify stimuli in terms of a species (e.g., “fish”) or an ecological niche (e.g., “aquatic habitat”) that is available within a folktaxonomic structure. While the preferred level of categorization within these taxonomies differs between cultures (e.g., Rosch et al., 1976; Medin et al., 1997) and expertise (Tanaka and Taylor, 1991; Medin et al., 1997), such taxonomies form the basis for all judgements of category membership. In the context of this work, we suggest that cognitive anthropological research on folktaxonomies has revealed uncanny valley-like phenomena in the form of “covert categories”–categories that cannot be readily placed into a taxonomical structure (e.g., octopus). Covert categories are cognitively isolated from other ontological categories (Berlin, 1974; Atran, 1983). For instance, informants might be able to identify a number of basic-level properties of an octopus, and yet be unable to associate it with a superordinate category (e.g., “fish”). In Berlin et al.'s (1968) study of Tzetal Mayans' folktaxonomies, they found numerous covert groups that did not fall into any of the major lifeform categories (this term is used in the anthropological literature but is equivalent to the superordinate level in the psychological literature; c.f., Brown, 1974). These categories are also associated with aversive responses, such as food prohibitions (Douglas, 1966/2002; Sperber, 1996). For example, Henrich and Henrich (2010) observed that the ambiguity in classifying an octopus as a “fish” or “non-fish” was associated with a food taboo. Douglas (1957) also observed similar outcomes with other ambiguous animals, like the flying squirrel. Such responses to categorically ambiguous stimuli are consistent with the uncanny valley hypothesis.