Not our feminism

Witt further designates gender “the mega social role,” by which she means that one’s gender is a uniquely important social role, inasmuch as it takes priority over and systematically influences one’s other social roles (2011, chap. 4). She notes that some social roles, like being a mother or being a husband, are explicitly gendered and hence transparently influenced by gender norms. Other social roles, like being a doctor or being an academic, are not explicitly gendered, but Witt points out that gender norms still strongly influence how these roles are variably enacted by women and men, as reflected in differences in working hours, earnings, specializations, and clothing. Witt (2011, 76–77) contends that characterizing genders as mega social roles in this way explains how social individuals are synchronically and diachronically unified—why there is only one social individual with various social roles where each person is, rather than a multitude of distinct role occupiers: the mother, the doctor, the immigrant, and so on. This, she tells us, is why she takes genders to be the unifying essences of social individuals. Nonetheless, on Witt’s account, the role occupiers in her example can’t be unified into a woman in the same way that the bricks, boards, and so on are unified into a house. For clearly, the bricks, boards, and so on are many things that are brought together to form a house; whereas Witt pointedly takes the mother, the doctor, the immigrant, and so on to be just one social individual with various social roles. Presumably, then, she seeks to explain how these various roles can be combined, and this is why she designates genders as the uniessences of social individuals. A social individual’s various social roles are all organized according to that individual’s gender, on her view, and in this sense, they are unified by that individual’s gender. Given our preceding arguments against gender uniessentialism, however, it can be seen that this is a solution in need of a problem. A single person can occupy various social positions, both simultaneously and over time, in the same way that it is generally possible for a single entity to belong to various kinds or, equivalently, to instantiate various properties, both simultaneously and over time. So, for the social roles of mother, doctor, and immigrant to be combined, it suffices for them all to be occupied by a single person, as indeed often happens, regardless of whether those roles are unified in any stronger sense. Genders may still be uniquely important social roles, for roughly the reasons given by Witt, but again, this is compatible with straightforwardly identifying each woman and each man with the corresponding person or human organism. In response to Witt, some critics have questioned whether any specific social roles have such special importance. It might instead be thought that agents typically have many different social roles that interact with each other in complex ways, with none of those roles generally taking priority over the others (see Ásta 2012, 4; Cudd 2012, 3–4). Indeed, this point may be seen as a dictum of intersectionality studies (see Stoljar 2018, 130), notwithstanding Witt’s (2011, 101) assertion that “the intersectionality of multiple social roles with gender is compatible with gender being the mega social role.” Moreover, it might be thought that even if there are mega social roles in Witt’s sense, these needn’t be gender roles. Instead, one’s racial or ethnic identity, or one’s status as disabled or able-bodied, might appear to take priority over one’s other social roles, including one’s gender (see Cudd 2012, 5–6). Be that as it may, Witt’s characterization of genders as mega social roles essentially concerns how gender roles and gender norms significantly interact with other social roles and social norms, and all this is perfectly compatible with saying that it is people who occupy the relevant roles and are thus subject to the associated norms. Gender kind essentialism, as we indicated earlier, is thoroughly opposed by many feminist philosophers who, as Witt (2011, xii) reports, see it as “theoretically misguided and politically dangerous,” and as inimical to morally necessary social change. However, despite this prevailing hostility to gender essentialism, Witt argues that her gender uniessentialism is actually useful to feminism, and she further insists that, whatever the merits or demerits of gender kind essentialism, her gender uniessentialism does not entail it and hence does not require its defence (2011, 9–13, 15–16, 127–28). Yet, as we have seen, Witt’s theory actually includes a specific account of gender properties as social positions (as pointed out by Stoljar [2018]), and hence arguably entails a specific version of gender kind essentialism, given a minimal conception of kinds. Besides, as we have suggested, gender kind essentialism per se may appear fairly trivial and—as Witt herself indicates (1995; 2011, 7–9)—entirely harmless, once it is distinguished from stronger doctrines. In any case, Witt claims that gender uniessentialism, in combination with her normative account of gender kinds and her ascriptivist account of gender norms, has the virtue of drawing our attention to oppressive gender norms in a way that enables us to effectively oppose them, in line with standard feminist goals. As she sees things (Witt 2011, 47, 128–29), in order to end or reduce women’s oppression, it is better to concentrate primarily on how gendered social roles are systematically ascribed to individuals by others in society, and pay less attention to individual preferences and biases concerning these roles, including the preferences and biases of women themselves. She later clarifies (2012, 9), however, that she does not view the social position of being a woman as oppressive in itself, but instead takes the further social norms that are contingently associated with it to be frequently oppressive, and so apt for social change.

Here we see an instance of the “debunking” strategy of exposing a morally objectionable phenomenon as constitutively socially constructed, with the aim of showing that it is not inevitable but instead can be changed by social means (see Haslanger 2003; Witt 2011, 38). Notably, this strategy involves two ideas that are liable to be conflated: that of being socially constructed and that of being socially changeable, neither of which strictly implies the other (see Schaffer 2017, 2455; pace Díaz-León 2015, 1145; Griffith 2018, 395–96). For instance, the fact that so many people have high cholesterol is not socially constructed, but it is socially changeable. In contrast, the fact that public languages are widely used is socially constructed, but it is not socially changeable, at least not within any functioning society. Still, all this is compatible with acknowledging that the social roles and social norms with which Witt is concerned are both socially constructed and socially changeable, and that exposing their specific social character is useful for devising ways to alter or eliminate them, where some such outcome is deemed desirable.