Duties of social identity? Intersectional objections to Sen’s identity politics

ABSTRACTAmartya Sen argues that sectarian discord and violence are fueled by confusion about the nature of identity, including the pervasive tendency to see ourselves as members of singular social groups standing in opposition to other groups (e.g. Democrat vs. Republican, Muslim vs. Christian, etc.). Sen defends an alternative model of identity, according to which we all inevitably belong to a plurality of discrete identity groups (including ethnicities, classes, genders, races, religions, careers, hobbies, etc.) and are obligated to choose, in any given context, which among our multiple affiliations to prioritize. While Sen’s model of discrete identity prioritization is a clear advance over single-factor accounts, it overlooks significant lessons about identity from over 150 years of scholarship by feminists of color. In ignoring the experiences of women of color, Sen’s model falsely assumes that identities are in-principle separable for the purposes of practical deliberation; and, in obligating individuals to make such identity-based ‘reasoned choices,’ Sen’s model forces those with multiply marginalized identities to choose from a set of externally defined identity options, none of which sufficiently captures their experiences.KEYWORDS: Social IdentityAmartya Senintersectionalityfeminism AcknowledgmentsThis paper has benefited greatly from audiences at the Mind, Language, and Social Change Conference in Knoll Farm, VT in August 2021, especially from Gabbrielle Johnson, Kate Ritchie, Umrao Sethi, Rima Basu, Jessica Moss, Michael Brownstein, Daniel Kelly, Ellen Fridland, Daniel Harris, Jonathan Phillips, and Peter Epstein; at the Philosophy of Social Science Roundtable in Vancouver in May 2017, especially from Fiona Jenkins, Georgie Warnke, and Derick Hughes; and at the Pacific Division meeting of the American Philosophical Association in April 2015, where Ranjoo Herr provided extremely helpful and incisive commentary.Disclosure statementNo potential conflict of interest was reported by the author(s).Notes1 Among work contemporaneous with Sen’s, and as we will gesture toward in §6, we are most sympathetic with, and influenced by, Alcoff (Citation2006; but see also Afshar, Aitken, and Franks Citation2005; Appiah Citation2007; Barvosa Citation2008; Phillips Citation2009; Warnke Citation2008).2 Sen’s Identity and Violence is cited over 5,550 times on Google Scholar but typically just in passing or sloganized form. We cite theorists who engage more carefully with his work on identity in what follows.3 We find this example phenomenologically implausible (must a person choose rather than discover their rooting preference?) and return to related concerns in §2.2.4 See Wills (Citation2018) for discussion of the shortcomings of class-reductionist readings of Marx. Wills (Citation2018, 236, original emphasis) also draws a key distinction between class exploitation and ‘classism’: ‘Workers do experience ‘classism’ – oppression on the basis of their working-class identity. They experience this for example when their specifically class-marked ways of speaking, eating, and dressing are marginalized and treated as inherently inferior to the habits of elites. But this oppression is importantly analytically distinct from class exploitation, which is an economic relation in which the value of their labor is systematically extracted from them.’5 See also the putative ‘reasoning’ and ‘reification’ problems for identity politics in Alcoff (Citation2006, chap. 2).6 Note that Sen’s contrast between ‘chosen’ and ‘discovered’ identities does not track the relevant distinction here. Sexuality, for example, can be a ‘discovered’ identity on Sen’s framework, but is not an essentially visible identify in Alcoff’s sense.7 For example, Appiah (Citation2008, 481) argues that Sen ignores a key difference between identities that are merely descriptive (e.g., to be a theist, you simply have to believe in God) and identities that carry normative prescriptions (e.g., to truly count as a member of a certain group (religion, gender, etc.), you are expected to act according to a rich set of norms of ‘conduct and feeling’). For related criticisms of Warnke’s (and Appiah’s) accounts of social identity, see Alcoff (Citation2012). See also Deaux (Citation1993) and references therein regarding attempts to distinguish social identity from ‘personal identity,’ by which social scientists mean the self-defining traits and values that distinguish oneself as an individual from other individuals.8 For example, eliding the distinction between identity and value may render practical deliberation inappropriately self-regarding. To return to one of Sen’s examples, suppose I’m a proud Londoner who values the development of the city I love, as well as a fiscal conservative concerned to limit public spending. Should the decision to support city development be framed in terms of identity, about ‘who I really am?’ Or should it be framed in terms of the relative merits of the two options (the things themselves rather than the deliberator’s identity)? This is not to imply that policy deliberation should never involve identity, e.g., individuals may have good identity-based reasons for supporting a community development plan.9 As Sen (Citation2008, 14) himself considers, perhaps it is ‘nothing other than an artificially fostered avowal … ’10 Sen claims that the singular model of identity is implied by certain communitarian ethical positions (Sen Citation2007, 32–36). We agree with Ranjoo Herr that few if any communitarians genuinely endorse such a view. In comments on an earlier version of this paper, Herr suggested that the real problem with Sen’s account is that he slips into lazy criticisms of identity politics and communitarianism that straw-person his opponents.11 For a thorough engagement with Huntington’s work, see Barvosa (Citation2008, chap. 1).12 Sen (Citation2009a, 142) offers a nearly identical example: a feminist activist in America who wants to do something to remedy particular features of women’s disadvantage in, say, Sudan would tend to draw on a sense of affinity that need not work through the sympathies of the American nation for the predicament of the Sudanese nation. Her identity as a fellow woman, or as a person (male or female) moved by feminist concerns, may be more important in a particular context than her citizenship … 13 For more on the difficulty of separating out identities, see, e.g., Lugones (Citation2003) and Spelman (Citation1988). For insightful empirical work on the contrast between identity prioritization and what social scientists call ‘identity content’ (i.e., an identity’s meaning or definition), see, e.g., van Zomeren, Kutlaca, and Turner-Zwinkels (Citation2018) and Mikołajczak, Becker, and Iyer (Citation2022). See also §5–6.14 An important question is whether there are correctness, authenticity, or fittingness conditions on some identities. Do some interpretations veridically track (among other things) the ongoing historical, political, and material situatedness of the identity better than other interpretations? Cf. the distinction between operative, manifest, and target concepts in Haslanger (Citation2012). For the purposes of this paper, we remain agnostic about these questions. However, our intuition is that, for some identities, it is possible for people to interpret their identities wrongly. For example, Black feminists have argued that the interpretations of Blackness and womanhood offered by some antiracist men and feminist white women are not just politically exclusionary but also false and misleading.Given how identities shift and emerge over time, new correctness conditions for a given identity may be brought into existence through social and material processes, including historical shifts in laws and institutions, as well as through collective and individual processes of identity exploration and experimentation (§6). New identities can be forged, and existing identities can change, through collective, dynamic, and creative processes (e.g., Anzaldúa Citation2012; Moraga and Anzaldúa Citation2015). Which if any identities have correctness conditions, how to determine them, and how they might change, are questions beyond this paper’s scope. One option for balancing the concerns to respect creative agency and to track social reality accurately might be to grant that there is a plurality of different ‘fitting’ interpretations but some incorrect ones (e.g., Alcoff Citation2006; Madva Citation2019). Consider white privilege. Although experiences of whiteness are heterogeneous, we find it plausible that at least some white people who deny that they benefit from white privilege are mistaken about their own identity. Thanks to Gabbrielle Johnson and an anonymous referee for incisive discussion here.15 See also Warnke’s (Citation2008) ‘identity minimalism.’16 For more on the well-established dynamics of navigating cognitive dissonance, see Gawronski and Strack (Citation2012).17 So far as we can tell, Sen’s discussions of particular cases are without exception critical of efforts to prioritize thick, intersectional, or combinatory identities. He is especially critical of combinations involving race, ethnicity, or religion, which he consistently portrays as ‘sectarian’ relative to various broader, superordinate identities, such as citizenship, class, or a ‘global sense of belonging’ (Sen Citation2007, 182; see also Citation2007, 14, 167; Citation2008, 11; Citation2009a, 142).18 Consider also the typically pernicious effects of construing LGBTQAI+ identities, and sexual identities in general, primarily as a matter of choice with little room for experimentation and discovery. For a nuanced analysis of the role of choice, interpretation, and discovery in sexual identity, see, e.g., Wilkerson (Citation2009).19 Even with respect to the partition of Indian and Pakistan, for example, Parul Sehgal (Citation2022) makes a compelling case that the eruption of violence is not best understood in terms of a single-factor opposition between Hindus and Muslims. He suggests it must also be understood as intersecting with gender and sexuality, for it involved pervasive inter-religious and intra-religious ‘sexual terrorism.’Although explicit endorsements of the singular model are hard to find, one can, by contrast, find many cases that plausibly involve the explicit endorsement of rigid prioritization, but these examples tend to strike us as the result of reasoned choices or efforts at rational persuasion, as in Jaimes and Means’ call to prioritize their American Indian identity over their gender, or calls to prioritize class over race (§1), and the cases we discuss in what follows.20 We do not claim that all violent political acts constitute ‘extremism.’ Some are justified acts of resistance to oppression. Our point is only that individuals are often recruited to groups espousing rigid prioritization through reasoned, narrative processes (rather than being swept up in sudden, mindless carnage).21 See, e.g., John S. Mosby’s (Citation1907) letter to Samuel Chapman. Alternatively, perhaps Southerners like Mosby (assuming we can take their subsequent disavowals of slavery at face value) should have bracketed their group-affiliative concerns altogether and simply done what was required by justice. See note 8.22 See also the comments and replies published alongside Gries and colleagues’ target article.23 A referee for Inquiry suggested that such remarks about Sen are ad hominem. They are not. They are reports of our failure to find any bibliographic evidence that Sen cites or otherwise engages with intersectional scholarship.