An error theory of essentializing the human category and the diversity of elements of essentialism
In the following section, an error theory of psychological essentialism regarding the human category will be assumed. If we essentialize what it means to fall within the human category, we make an error, since, scientifically viewed, there is no such essence of what it means to be human. Within philosophy of science, Hull (1986) is considered as the starting point for the by-now broad consensus on such an error theory of essentializing the human category (see Kronfeldner 2018, for review and a contemporary defense of it). This also means that some humans (e.g., scientists) are cognitively able to categorize humans without using essences. This is compatible with experimental studies, which have been interpreted to show that many children and some adults are unable to reach a post-essentialist style of reasoning (see, for a canonical summary of that body of research, Gelman 2003). The respective studies thus neither show that essentializing is a necessary part of how humans categorize (since it applies to some adults but clearly not all since at least scientists have moved beyond it), nor is the essentialist interpretation of the respective studies uncontested. Strevens (2000), for instance, defends an alternative explanation of the accumulated data, an interpretation that does not attribute essentialist thinking to the study participants.
Worse even, there is no agreement on the definition of psychological essentialism, at least not if one takes into account the relevant literature across developmental psychology', cognitive science, social psychology, philosophy, history, and gender studies. Given this situation, I decided to specify and analyze the connection to dehumanization with respect to specific elements and dimensions of psychological essentialism rather than use one definition of essentialism (i.e., one combination of the elements only). We met some such elements already above (Section 24.2). Rothbart and Taylor (1992), in a paper that is widely acknowledged as an anchor for discussions about psychological essentialism regarding social categories, mention a further element — namely, informativeness (inductive potential), which is well known from natural kind thinking and reminds us that knowledge of group membership often comes with a potpourri of information that grounds inductive projection (inferences about further properties of the individual). For instance, learning that a piece of shining stuff in one’s hand is a piece of gold is quite informative. One canreliably infer from that alone some interesting additional facts about the very stuff in one’s hand: the weight of it, the hardness of it, when it melts, the value of it on the market, and so on.
With this, we have the following list of elements of essentialized thinking:
- • Homogeneity of group members with respect to salient properties;
- • Informativeness of group membership for inductive inferences about properties typical of the kind;
- • Inherence of essential properties;
- • Naturalness of essential properties, either via the concept of human nature as referring to the biological species Homo sapiens, or via the concept of natural kind;
- • Inalterability of essential properties, either developmentally and/or evohitionarily;
- • Non-observability of essential properties;
- • Distinctness of group boundary (i.e., group boundaries and membership are mutually exclusive);
- • Normality of properties and members, with a reference to normativity and thus to deficiency.
There are similar lists in the relevant literature that all differ slightly in the definition and number of elements.3 These differences do not matter for the purposes of this chapter. What does matter is that, according to Rothbart and Taylor (1992) and Haslam et al. (2000,2002), these elements can be aligned along two dimensions. In the following, I take it that only homogeneity and informativeness are necessarily part of the first dimension of essentializing social categories. This dimension is called entitativity since it gives these groups an entity-like coherence (an idea that goes back to Campbell 1958).4 The rest—in particular naturalness, inalterability, non-observability, and distinctness—belong to a second dimension of essentialism. It entails natural kind thinking, since it fits how 20th-century philosophy has characterized the latter (see Ereshefsky 2010, for review). Normality can attach to both dimensions. Since Smith and Leyens et al., who have made explicit claims about essentialism being necessary for dehumanization, used either natural kind thinking or entitativity, we will scrutinize these two kinds of essentialisms separately. 1 will start with the more demanding natural kind thinking.