Abstract
Here, I examine the epistemic relation between beliefs about the nature of sexual orientation (e.g. beliefs concerning whether orientation is dispositional) and beliefs about the taxonomy of orientation categories (e.g. beliefs concerning whether polyamorous is an orientation category). Current philosophical research gives epistemic priority to the former class of beliefs, such that beliefs about the taxonomy of orientation categories tend to be jettisoned or revised in cases of conflict with beliefs about the nature of sexual orientation. Yet, considering the influence of ideology on beliefs about socially significant phenomena, I argue for an epistemic reversal.
What is sexual orientation? This important question about the nature of sexual orientation has been of central interest to philosophers working on the social ontology of sexuality. We are also interested in the taxonomy of sexual orientation categories. What are the sexual orientation categories? Perhaps the socially dominant taxonomy is correct. Or, maybe we ought to endorse an alternative taxonomy.
In the social ontology of sexuality, beliefs about the nature of sexual orientation are generally given epistemic priority in relation to beliefs about the taxonomy of orientation categories. I refer to this treatment of the epistemic relation between beliefs about the nature of sexual orientation and beliefs about the taxonomy of orientation categories as the orientation-first view. In this paper, I argue that we ought to reject the orientation-first view in favor of the taxonomy-first view, which gives beliefs about the taxonomy of orientation categories epistemic priority in relation to beliefs about the nature of sexual orientation. In effect, I aim to reverse the epistemic order of a recently sprouted sub-field of philosophical inquiry.
My argument for the taxonomy-first view proceeds as follows. In the first section, I explicate the distinction between the orientation-first and taxonomy-first views. In the second section, I provide a dialectical reason to endorse the taxonomy-first view. In particular, I discuss a recent debate between Robin Dembroff and Esa Díaz-León about the nature of sexual orientation (Dembroff 2016; Díaz-León forthcoming). And I argue that while Dembroff and Díaz-León’s endorsement of the orientation-first view generates an impasse, the taxonomy-first view allows the dialectic to move forward. In the third section, I provide an independent argument for the taxonomy-first view by considering the influence of ideology on beliefs related to the nature and taxonomy of social properties.
1 The Orientation-First vs. Taxonomy-First Views of Sexual Orientation
In this section, I will explicate the orientation-first and taxonomy-first views of sexual orientation by distinguishing orientation facts from orientation taxonomy facts (Section 1.1), before describing how each view understands the epistemic relation between these two types of facts (Section 1.2). And I will note that the orientation-first view is endorsed in recent work on the metaphysics of sexual orientation (Section 1.3).
1.1 Orientation Facts and Orientation Taxonomy Facts
Orientation facts are facts about the nature of sexual orientation. With Díaz-León, let us suppose that sexual orientation is a property (Díaz-León forthcoming). The property of sexual orientation is instantiated in many persons, but it is not instantiated in sedimentary rocks, ferns, or prime numbers. Now, what is the nature of the property of sexual orientation? Answering this question will yield the orientation facts.
For example, if either Dembroff or Díaz-León’s theory of sexual orientation is correct, it is an orientation fact that orientation is a dispositional property. As I will discuss below, if Díaz-León’s analysis is correct, it is an orientation fact that individuals instantiate the property of sexual orientation in virtue of being disposed to have certain sexual desires (Díaz-León forthcoming). In contrast, if Dembroff’s analysis is correct, it is an orientation fact that individuals instantiate the property of sexual orientation in virtue of being disposed to engage in particular sexual behaviors (Dembroff 2016, p. 18).
In order to gain additional traction on the concept of orientation facts, let us consider facts about the nature of race.[1] Many persons have a race, unlike metamorphic rocks, seaweed, and composite numbers. Here, I will highlight two competing theories about the nature of race. On Quayshawn Spencer’s view, individuals have a race in virtue of being a member of a “human population partition,” i.e. a genetically significant division in the species homo sapiens (Spencer 2014, p. 1029–1032). In contrast, Sally Haslanger holds that individuals have a race in virtue of occupying a particular type of social position, more specifically: in virtue of being socially subordinated or privileged on the basis of perceived geographical ancestry.[2]
In contrast to orientation facts, orientation taxonomy facts include facts about what categories are orientation categories. For example, it is a fact about the taxonomy of orientation categories that woman is not an orientation category. My enduring admiration notwithstanding, I will even hazard to claim that there is no orientation category that corresponds to being exclusively attracted to Madonna. As will become important below, Díaz-León holds that homosexual, heterosexual, and bisexual are orientation categories, while Dembroff endorses a revisionary taxonomy that includes categories such as woman-oriented and female-oriented.
Here, it will be useful to continue the analogy with race. In addition to facts about the nature of race, there are facts about the taxonomy of race categories.[3] It is evident that lawyer and U.S. Citizen are not race categories. So, what are the race categories? Haslanger holds that taxonomies of race categories vary across social milieus (Haslanger 2012b, p. 308). For example, on Haslanger’s view, the taxonomy of race categories in early twentieth-century London is distinct from the mid-century taxonomy in Germany. Haslanger holds that, in the contemporary US, the taxonomy includes (at least) the categories of White, Black, Asian, and Latinx (Haslanger 2012b, p. 306). In contrast, Spencer holds that the taxonomy of race categories has remained constant across recent evolutionary history, and the taxonomy exhaustively includes the following categories: African, Caucasian, East Asian, American Indian, and Oceanian (Spencer 2014, p. 1030). Note that Spencer’s categories are (more than lexically) distinct from Haslanger’s categories. For example, Spencer categorizes Middle Eastern and South Asian individuals as Caucasian. But for Haslanger, the category White does not (at least straightforwardly) include Middle Eastern and South Asian individuals.
1.2 The Epistemic Relation Between Orientation Facts and Orientation Taxonomy Facts
The orientation-first and taxonomy-first views are distinguished by their answers to the following question: what is the epistemic relation between orientation facts and orientation taxonomy facts? The orientation-first view holds that beliefs about the nature of sexual orientation have epistemic priority relative to beliefs about the taxonomy of orientation categories. And the taxonomy-first view holds the reverse.
In order to cash out the notion of epistemic priority, suppose that, at T1, I believed that Doggo is a dog, and I also believed that it is metaphysically impossible to have a non-veridical perceptual experience. At T2, I had a perceptual experience in which Doggo’s ear was damaged, revealing part of a computer chip and some sleek mechanisms, and I formed the belief that I perceived (veridically or otherwise) that Doggo is composed of robot parts. For the purpose of example, let us hold fixed this belief about my perceiving Doggo to be composed of robot parts, and suppose that – given this belief – there is a conflict between the belief that Doggo is a dog and the belief that it is metaphysically impossible to have a non-veridical perceptual experience.
At T3, what ought I to believe? Perhaps I ought to retain the belief that Doggo is a dog and revise the belief that it is metaphysically impossible to have a non-veridical perceptual experience (such that, e.g. I come to believe that it is merely unlikely that any given perceptual experience is non-veridical). Alternatively, perhaps I ought to jettison or revise the belief that Doggo is a dog. I would not rule on what the norms of belief require in this case. If the belief that Doggo is a dog has epistemic priority in relation to the belief that it is impossible to have a non-veridical perceptual experience, then I ought to retain the former belief and jettison or revise the latter. And the reverse holds if the latter belief has epistemic priority in relation to the former. Or, perhaps both beliefs ought to be revised, in which case neither has epistemic priority in relation to the other. The point of this toy (poodle) example can be expressed as follows: belief B has epistemic priority in relation to belief B1 if and only if, in the case of conflict, belief B ought to be retained and belief B1 ought to be jettisoned or revised.
However, as will become evident below, the taxonomy-first view concerns the epistemic priority of beliefs of a certain class in relation to beliefs of another class. We might capture this with a modification to the above formulation: beliefs that are members of class C have epistemic priority in relation to beliefs that are members of class C1 if and only if, in the case of conflict, the class C belief ought to be retained and the class C1 belief ought to be jettisoned or revised. Yet, that formulation will be unduly strict for the purpose of explicating the taxonomy-first view of sexual orientation. Instead, what is important is a generic generalization: class C beliefs have epistemic priority in relation to class C1 beliefs just in case, as a matter of generic generalization, in the case of conflict, class C beliefs ought to be retained and class C1 beliefs ought to be jettisoned or revised. For example, it is plausible that an agent’s beliefs about their occurrent mental states have this sort of epistemic priority in relation to their beliefs about the external world. Here, as a matter of generic generalization, an agent’s beliefs about their occurrent mental states ought to be retained in the case of conflict with beliefs about the external world.
Notably, the above formulation of epistemic priority appeals quite broadly to norms of belief, that is, norms about what an agent ought to believe. This gloss is in the interest of neutrality. Here, I do not aim to assess whether the norms of belief are explained by a universally-demanding norm of rationality; there is a tradition in feminist epistemology critiquing the idea.[4] Neither do I aim to assess whether the norms of belief are purely epistemic; in addition to evidential reasons, there might be moral reasons to retain, revise, or jettison beliefs.[5] Instead of searching precisely for what the norms of belief involve, what matters in the context of this paper is finding what they demand. (In case that seems impracticable, note an analogy to the moral case. Even without complete knowledge of the correct moral theory, it is plausible that agents can know whether an action is right or wrong.)
To further clarify the distinction between the orientation-first and taxonomy-first views, it will be useful to continue the analogy with race. As noted above, Haslanger and Spencer endorse significantly different theories about the nature of race and the taxonomy of race categories. Importantly, Haslanger and Spencer also have significantly different views about the epistemic relation between beliefs about the nature of race and beliefs about the taxonomy of race categories.
Haslanger’s analysis of race relies on the belief that the taxonomy of race categories in the contemporary US includes the categories of Black, White, Asian, and Latinx. On this point, she claims:
We can all confidently identify members of different races. Martin Luther King, Nelson Mandela, Malcolm X, Toni Morrison, Oprah Winfrey, W. E. B. DuBois, Kofi Annan, Thabo Mbeki (insert here your choice of various friends and relatives) are Black. George Bush, Arnold Schwarzenegger, Margaret Thatcher, Golda Meir, Bertrand Russell, Vincent Van Gogh (insert here your choice of various friends and relatives) are White. Similar lists can be constructed for Asians, Latino/as, and other groups usually considered races. But if this is the case, then the terms “Black” and “White” pick out the best fitting and most unified objective type of which the members of the list are paradigms – even if I can not describe the type or my beliefs about what the paradigms have in common are false (Haslanger 2012b, p. 306).
The idea that Black, White, Asian, and Latinx are race categories is bedrock in Haslanger’s analysis. From here, Haslanger’s task is to analyze the categories. Haslanger argues that – notwithstanding any appearance to the contrary – the categories Black, White, Asian, and Latinx do not correspond to natural kinds (Haslanger 2012b, p. 306–307).[6] Instead, she argues that the race categories correspond to certain social positions.
In the context of this paper, it is not important to dive into the details of Haslanger’s social position analysis of race. Here, my point is that Haslanger gives beliefs about the taxonomy of race categories epistemic priority in relation to beliefs about the nature of race. For Haslanger, if an otherwise plausible theory of the nature of race has the result that Latinx is not a race category, that is reason to reject the theory of the nature of race – it is not reason to jettison the belief that Latinx is a race category.[7]
In contrast, Spencer gives beliefs about the nature of race epistemic priority in relation to beliefs about the taxonomy of race categories. As noted above, Spencer holds that race is a feature of human population genetics. More precisely, Spencer holds that “race” refers to a set containing the five most genetically significant human population partitions (Spencer 2014, p. 1026–1029). This theory of the nature of race is bedrock for Spencer.
With his genetic theory of race at hand, Spencer asks: what are the five most genetically significant human population partitions? Considering the empirical data, Spencer concludes that we ought to endorse “the Blumenbach partition,” which exhaustively includes the categories discussed in the previous sub-section: African, Caucasian, East Asian, American Indian, and Oceanian (Spencer 2014, p. 1030).
Although Spencer is interested in correlations between the Blumenbach categories and our ordinary categories, he allows that there are significant differences between the taxonomies. For example, Spencer notes that Latinx does not neatly correspond to any of the Blumenbach categories (Spencer 2014, p. 1033). Although Spencer does not say precisely how he handles this “mismatch” between the Blumenbach categories and our ordinary categories, he denies that Latinx is a race category.[8]
Here, my point is that Spencer gives beliefs about the nature of race epistemic priority in relation to beliefs about the taxonomy of race categories. For Spencer, if an otherwise plausible theory of the taxonomy of race categories leads to the result that race is not a feature of human population genetics, that is reason to reject the theory of the taxonomy of race categories – it is not reason to jettison the belief that race can be analyzed in terms of human population genetics.
By analogy with the epistemologies of Haslanger and Spencer’s theories of race, we can explicate the difference between the orientation-first and taxonomy-first views of sexual orientation. The orientation-first view holds that, as a matter of generic generalization, in the case of conflict, beliefs about the nature of sexual orientation ought to be retained and beliefs about the taxonomy of orientation categories ought to be jettisoned or revised. In contrast, on the taxonomy-first view, as a matter of generic generalization, in the case of conflict, beliefs about the taxonomy of orientation categories ought to be retained and beliefs about the nature of sexual orientation ought to be jettisoned or revised. Of course, that is all quite schematic.[9] In order to flesh out this outline, we will need an example of inconsistent beliefs about the nature of sexual orientation and the taxonomy of orientation categories. (That is, we will need an example analogous to the following inconsistent beliefs about the nature of race and the taxonomy of race categories: race is a feature of human population genetics, and Latinx is a race category.) In section two, I will provide further traction on the orientation-first and taxonomy-first views by explicating how Dembroff and Díaz-León’s theories of the nature of sexual orientation are inconsistent with (different) beliefs about the taxonomy of orientation categories, before considering how the orientation-first and taxonomy-first views recommend handling these inconsistencies.
With the distinction between the orientation-first and taxonomy-first views of sexual orientation at hand, I move to consider Dembroff and Díaz-León’s treatment of the epistemic relation between orientation facts and taxonomy facts.
1.3 Dembroff and Díaz-León’s Endorsement of the Orientation-First View
The work of Dembroff and Díaz-León is a touchstone in the social ontology of sexuality. Díaz-León seems to follow Dembroff’s stated methodology (Díaz-León forthcoming), which Dembroff describes as follows:
[This paper’s] target is twofold: (i) the everyday concept of sexual orientation, and (ii) the corresponding concepts associated with the taxonomy of sexual orientation (e.g. gay, straight). These concepts are highly interwoven, since the concept of sexual orientation constrains the taxonomy […] My project sets out to engineer a revised concept of sexual orientation that implies a new taxonomical schema of sexual orientation (Dembroff 2016, p. 2).
Here, Dembroff holds that the analysis of sexual orientation constrains the analysis of the taxonomy, but they do not mention the reverse. Additionally, Dembroff claims that the concept of sexual orientation “implies” a taxonomy of orientation categories (Dembroff 2016, p. 2, 4).
To clarify, the orientation-first view does not require remaining agnostic about the taxonomy of orientation categories until settling on a theory of the nature of sexual orientation. Indeed, Díaz-León and Dembroff consider the merits of various taxonomies in developing their analyses of sexual orientation. Instead, what is important to the orientation-first view is the claim that beliefs about the nature of sexual orientation generally ought to be preferred in the case of conflict with beliefs about the taxonomy of orientation categories.
While I aim to have shown that Díaz-León and Dembroff endorse the orientation-first view, my interpretation is open to the following complication. There is a difference between Díaz-León and Dembroff’s stated methodologies, on the one hand, and the methodological parameters under which their arguments proceed, on the other. And there is a worry that I have unduly highlighted the former. I deny that I have focused on Díaz-León and Dembroff’s stated methodologies at the expense of misinterpreting the methodologies that they actually employ. Yet, even if that were the case, it would still be an interesting result if Díaz-León and Dembroff had reason to endorse the taxonomy-first view, contra their stated methodologies. More importantly, as will become evident in the following sections, the question of whether we ought to endorse an orientation-first or taxonomy-first view of sexual orientation is of general significance to research on the social ontology of sexuality.
2 The Dialectical Consequences of the Orientation-First and Taxonomy-First Views
In this section, I outline a recent debate between Dembroff and Díaz-León about the nature of sexual orientation (Section 2.1), argue that Dembroff and Díaz-León’s endorsement of the orientation-first view generates an impasse (Section 2.2), and describe how the taxonomy-first view can resolve the impasse in the debate (Section 2.3). Additionally, I answer an objection to this argument for the taxonomy-first view (Section 2.4).
Here, note that the dialectical force of the taxonomy-first view provides reason to endorse it. This is the case for the following two reasons. First, although some dialectics close with an impasse, that would be a strange result at this point in the debate between Dembroff and Díaz-León. Social metaphysicians have just begun discussing the nature of sexual orientation. Moreover, Dembroff and Díaz-León share a queer perspective on the topic of orientation. Accordingly, it seems unlikely that the debate would have already reached an irresolvable impasse. Second, I join Dembroff and Díaz-León’s ameliorative project, which aims to produce sexual orientation concepts that – if deployed in our milieu – would have beneficial social and political effects.[10] And the impasse between Dembroff and Díaz-León is an obstacle to our shared ameliorative aims. For these reasons, we should search for ways to advance the dialectic.
2.1 The Debate between Dembroff and Díaz-León
In their groundbreaking work on sexual orientation, Dembroff argues for the following analysis of sexual orientation, which they refer to as bidimensional dispositionalism:
A person S’s sexual orientation is grounded in S’s dispositions to engage in sexual behaviors under the ordinary condition[s] for these dispositions, and which sexual orientation S has is grounded in what sex[es] and gender[s] of persons S is disposed to sexually engage under these conditions (Dembroff 2016, p. 18).
Dembroff’s analysis holds that the property of sexual orientation is dispositional, behavior-based, relative to both sex and gender, and thinly-relational. In turn, I will work through these aspects of bidimensional dispositionalism. (Also, I aim for the discussion to highlight and clarify some of the theoretical choice points in the sub-field.)
First, let us distinguish between behavior-based and desire-based analyses of orientation. Behavior-based views hold that individuals instantiate the property of sexual orientation in virtue of features of their sexual behaviors, not their desires. In contrast, desire-based views hold that individuals instantiate the property of sexual orientation in virtue of features of their sexual desires, not their behaviors. Dembroff endorses a behavior-based analysis of orientation.
Second, we can distinguish between dispositional and categorical analyses of orientation.[11] On categorical analyses, individuals instantiate the property of sexual orientation in virtue of their sexual desires or behaviors. In contrast, dispositional analyses hold that individuals instantiate the property of sexual orientation in virtue of their dispositions to sexual desires or behavior. We can combine desire-based and behavior-based views with either dispositional or categorical analyses.[12] For example, Dembroff endorses a behavior-based dispositional analysis, holding that individuals instantiate the property of sexual orientation in virtue of their dispositions to sexual behavior.
Third, analyses of sexual orientation must answer the following question: on the basis of attractions to which type(s) of features do individuals instantiate the property of sexual orientation? For example, suppose that an individual is exclusively attracted to short female women. Does the individual instantiate the property of sexual orientation in virtue of their attractions to sex-features, gender-features, and/or height-features? Dembroff argues that individuals instantiate the property of sexual orientation in virtue of their attractions to sex-features and/or gender-features, but not other types of features.[13]
Fourth, analyses of orientation are either (what I will call) thickly-relational or thinly-relational. On thickly-relational analyses of orientation, individuals instantiate the property of sexual orientation in virtue of being attracted to individuals with particular sex-features, gender-features, or other features, and themselves having particular sex-features, gender-features, or other features. In contrast, thinly-relational analyses of orientation hold that individuals instantiate the property of sexual orientation in virtue of being attracted to individuals with particular sex-features, gender-features, or other features, irrespective of their own sex, gender, etc. Dembroff endorses a thinly-relational analysis of orientation.
With Dembroff’s account at hand, I move to provide Díaz-León’s analysis of sexual orientation, as well as her argument against bidimensional dispositionalism. Motivated by the idea that “we could understand sexual orientations in terms of sexual preferences” (Díaz-León forthcoming). Díaz-León develops the following analysis of orientation, which she refers to as the desire view:
A person S’s sexual orientation is determined by the sex[es] and/or the gender[s] of persons for whom S is disposed to have sexual desires, under the relevant manifesting conditions (and S’s own sex and/or gender) (Díaz-León forthcoming).
And Díaz-León provides the following analysis of sexual desire:
A sexual desire (for men and/or women) involves the combination of a propositional attitude (of the form “S bears the relation of desiring towards proposition p”) plus a disposition to be sexually aroused by, or sexually attracted to, men and/or women (Díaz-León forthcoming).
With Dembroff, Díaz-León endorses a dispositional analysis of sexual orientation. Also with Dembroff, Díaz-León holds that individuals instantiate the property of sexual orientation in virtue of being attracted to individuals with particular sex-features and/or gender-features, but not other types of features. Contra Dembroff, Díaz-León’s analysis of orientation is thickly-relational and desire-based.
Note that on Díaz-León’s view, sexual desire is not (merely) a disposition to sexual behavior. This is the case because Díaz-León’s account of sexual desire includes a phenomenological element of arousal. For example, suppose that Simone desires to have sex with Dominique. For Díaz-León, this requires that Simone is disposed to experience sexual arousal related to Dominique. That is, Simone’s desire to have sex with Dominique includes a disposition to have a certain phenomenological experience. As phenomenological experiences such as arousal are not dispositions to behavior,[14] Díaz-León would deny that sexual desires are (mere) dispositions to behavior.
I highlight the phenomenological element of Díaz-León’s account of desire in order to make sense of the substantivity of the disagreement between Dembroff and Díaz-León. If Díaz-León were to analyze sexual desires as dispositions to sexual behavior, then the disagreement between Dembroff and Díaz-León could be glossed as follows: while Dembroff holds that orientation is a matter of first-order dispositions to sexual behavior, Díaz-León holds that orientation is a matter of second-order dispositions to sexual behavior (that is, dispositions to dispositions to sexual behavior). In that case, there would not be much of a disagreement. However, as Díaz-León holds that sexual desire includes a phenomenological element of sexual arousal, the deflationary gloss of the disagreement between Dembroff and Díaz-León is not available.[15]
For the purposes of this paper, I will focus on Díaz-León’s critique of Dembroff’s bidimensional dispositionalist analysis of orientation. In particular, Díaz-León argues that – unlike the desire view – bidimensional dispositionalism cannot accurately ascribe heterosexuality and bisexuality to individuals, i.e. bidimensional dispositionalism cannot capture the membership conditions of categories such as heterosexual and bisexual.
Díaz-León begins her critique by considering Alicia, a female woman, who is sexually aroused by both women and men. Given this feature of Alicia’s sexuality, Díaz-León holds that Alicia is bisexual. Díaz-León constructs the case of Alicia such that in the actual world, as well as in nearby possible worlds, Alicia is disposed to have sex exclusively with men. However, Alicia is disposed to have sex with both women and men in distant possible worlds. (In nearby possible worlds, Alicia is in a monogamous relationship with a particular man.) In order to capture the fact that Alicia is bisexual, Díaz-León claims that bidimensional dispositionalism must hold that individuals instantiate the property of sexual orientation in virtue of their dispositions to behavior in both nearby and distant possible worlds.
Next, Díaz-León considers Cary, a male man, who is predominately – indeed, almost exclusively – sexually aroused by women. Given this feature of Cary’s sexuality, Díaz-León holds that Cary is heterosexual.[16] Díaz-León imagines the case such that in the actual world, as well as in nearby possible worlds, Cary is disposed to engage in sexual activity exclusively with women. However, in some distant possible worlds, Cary is disposed to have sex with women and men. (In some distant possible worlds, Cary has a more experimental personality.) In order to capture the fact that Cary is heterosexual, Díaz-León claims that bidimensional dispositionalism must hold that individuals instantiate the property of sexual orientation in virtue of their dispositions to behavior in nearby possible worlds, but not in distant possible worlds.
With these cases at hand, Díaz-León presents the following critique of bidimensional dispositionalism:
The main worry for bidimensional dispositionalism can be put in the form of a dilemma: If we understand the account loosely enough, then we can count possible worlds where Alicia is not monogamous and has sex with women as being relevant for determining someone’s sexual orientation, and then the account would rightly capture the intuition that she is bisexual. But if we take this approach, then there seems to be no way of ruling out possible worlds where Cary feels like experimenting and has sex with some men, so the account could not capture the intuition that Cary is heterosexual. On the other hand, if we understand the relevant manifesting conditions more narrowly, and restrict the possible worlds to those where Cary does not feel like experimenting with men, then we should also restrict the possible worlds to those where Alicia is in a monogamous relationship with her male partner, but then Alicia would count as heterosexual, not bisexual. In conclusion, I do not see any way of modifying the account so that it can solve both counterexamples at the same time (Díaz-León forthcoming).
In short, there is no interpretation of bidimensional dispositionalism that ascribes bisexuality to Alicia and heterosexuality to Cary. For this reason, Díaz-León argues that we ought to reject bidimensional dispositionalism.
2.2 The Orientation-First View Generates an Impasse between Dembroff and Díaz-León
I think that Díaz-León has demonstrated that bidimensional dispositionalism cannot capture the membership conditions of categories such as heterosexual and bisexual. However, Díaz-León’s critique of bidimensional dispositionalism is only successful if categories such as heterosexual and bisexual are orientation categories. Dembroff can reply to Díaz-León by holding that bidimensional dispositionalism captures the membership conditions of real orientation categories such as female-oriented and woman-oriented, notwithstanding its treatment of categories such as bisexual.
So, in order to advance the dialectic, we need to know the taxonomy of orientation categories. However, as I demonstrate below, Dembroff and Díaz-León’s endorsement of the orientation-first view generates an impasse in their debate. By “impasse,” I refer to a dialectical situation in which (i) thinkers are rational to endorse their own arguments, (ii) which provide at least one thinker reason to deny premises in the argument of their interlocutor, and (iii) in which there is no mutually acceptable way to assess the truth or falsity of the disputed premise(s) in each argument.
As I hope is clear from the discussion in the previous sub-section, both Dembroff and Díaz-León are rational to endorse their own arguments. So, their dialectic has the first feature of an impasse. (Of course, lots of dialectics have this feature.)
Next, Dembroff’s endorsement of bidimensional dispositionalism provides them reason to deny the following premise in Díaz-León’s critique: heterosexual and bisexual are orientation categories. This is the case because bidimensional dispositionalism holds that sexual orientation is thinly-relational, such that individuals instantiate the property of sexual orientation in virtue of being attracted to individuals with certain features (irrespective of their own features). So, the dialectic has the second feature of an impasse.
Let us take stock. Díaz-León’s critique of bidimensional dispositionalism turns on the truth of the premise that heterosexual and bisexual are orientation categories, while Dembroff’s view requires its denial. (I will consider this point in more detail in Section 2.4.) In order to advance the dialectic, we need a mutually acceptable way to assess the premise.
Here is where Dembroff and Díaz-León’s endorsement of the orientation-first view generates an impasse. Suppose that Dembroff were provided with an argument for a taxonomy that includes the category of homosexuality. Because Dembroff gives beliefs about the nature of orientation epistemic priority in relation to beliefs about the taxonomy of orientation categories, if an argument for a taxonomy fact has the result that orientation is not thinly-relational, that is reason for Dembroff to reject the argument. It is not reason for Dembroff to jettison the belief that orientation is thinly-relational. Likewise, suppose that Díaz-León were provided with an argument for a taxonomy that includes categories such as woman-orientated and female-orientated. On the orientation-first view, Díaz-León would have reason to reject the argument, as thinly-relational categories are incompatible with Díaz-León’s version of the desire view. In short, on the orientation-first view, it is not the case that arguments for taxonomy facts supersede the implications of theories about the nature of orientation. For this reason, the dialectic between Dembroff and Díaz-León has the third and final feature of an impasse.
2.3 The Taxonomy-First View can Resolve the Impasse
Above, I argued that the orientation-first view generates an impasse in the debate between Dembroff and Díaz-León. Here, I will explain how the impasse can be resolved with the taxonomy-first view.
Recall that the taxonomy-first view holds that beliefs about the taxonomy of orientation categories have epistemic priority in relation to beliefs about the nature of sexual orientation. To clarify, the taxonomy-first view allows that an argument in favor of a particular taxonomy of orientation categories might rely on beliefs about the nature of sexual orientation. For example, in the present dialectical context, such an argument might rely on the belief – shared by Dembroff and Díaz-León – that orientation is dispositional. That is, the taxonomy-first view does not require that we suspend judgment about the nature of orientation until we have settled on a taxonomy of orientation categories. Indeed, without any knowledge of the nature of orientation, it might not be possible to distinguish between plausible candidates for orientation categories (such as homosexual and female-oriented) and implausible candidates (such as man and Black). Instead, the taxonomy-first view holds that beliefs about the taxonomy of orientation categories ought to be preferred in the case of conflict with beliefs about the nature of sexual orientation.
With that in mind, let us again suppose that Dembroff were provided with an argument for a taxonomy that includes the category homosexual. On the taxonomy-first view, such an argument could provide Dembroff with reason to revise their belief that sexual orientation is thinly-relational. Likewise, suppose that Díaz-León were provided with an argument for a taxonomy that includes the categories woman-orientated and female-orientated. On the taxonomy-first view, such an argument could provide Díaz-León with reason to revise her belief that sexual orientation is thickly-relational. Accordingly, the taxonomy-first view has the potential to advance the dialectic between Dembroff and Díaz-León.[17]
2.4 Objection and Reply
Above, I held that Díaz-León’s critique of Dembroff’s bidimensional dispositionalism turns on the truth of the claim that heterosexual and bisexual are orientation categories, while Dembroff’s account requires its denial. But here is an objection: Díaz-León’s critique of behavior-based accounts also applies to categories such as woman-oriented and female-oriented. That is – according to the objection – Díaz-León’s argument more generally demonstrates that sexual orientation categories can not be analyzed as dispositions to behavior.
As outlined above, Díaz-León argues that behavior-based views are unintuitive with respect to the membership conditions of categories such as heterosexual and bisexual. Without an explanation of these unintuitive results, it seems that Díaz-León has demonstrated that behavior-based accounts can not capture the membership conditions of categories such as heterosexual and bisexual.[18]
Yet, I deny that Díaz-León’s critique demonstrates that behavior-based views are unintuitive with respect to the membership conditions of categories such as woman-oriented and man-oriented. To be clear, I think that Díaz-León compellingly shows that Dembroff faces the following choice: either a wide range of possible worlds is relevant to orientation ascriptions, in which case both Alicia and Cary are woman-oriented and man-oriented, or a narrower range of possible worlds is relevant to orientation ascriptions, in which case Alicia is exclusively man-oriented and Cary is exclusively woman-oriented. Still, Díaz-León’s argument only poses a challenge to Dembroff’s view if the above options are unintuitive with respect to the membership conditions of categories such as woman-oriented and man-oriented.
However, I deny that ascriptions of thinly-relational orientations can be assessed for intuitiveness in the same way as ascriptions of homosexuality, heterosexuality, or bisexuality. On this point, note that Dembroff holds that categories such as man-oriented woman are not identical to categories such as heterosexual woman. In particular, Dembroff argues that the categories heterosexual, homosexual, and bisexual are cisnormative; that is, Dembroff argues that in order to be a member of the category heterosexual, an individual must be a cisgender woman exclusively attracted to cisgender men, or a cisgender man exclusively attracted to cisgender women. For Dembroff, this point applies to dominant as well as revisionary versions of the categories heterosexual, homosexual, and bisexual (Dembroff 2016, p. 5, 19, 24–25). Given that Dembroff’s thinly-relational categories are not (yet?) socially operative, I doubt that fine-grained judgments about the intuitiveness of their membership conditions are warranted. Indeed, if certain membership conditions seem unintuitive, this might be on account of problematically interpreting Dembroff’s categories through the lens of extant thickly-relational categories.
Here is another way to put the point. Because Díaz-León holds that we ought to revise extant categories, her account must – to some extent – answer to the current membership conditions of categories such as homosexual and bisexual. But Dembroff’s new categories do not generate an analogous constraint. Accordingly, I deny that Díaz-León’s critique of behavior-based theories of sexual orientation applies to accounts which endorse categories such as woman-oriented and man-oriented.
3 An Argument from the Influence of Ideology
Above, I argued that we ought to endorse the taxonomy-first view of sexual orientation on account of its dialectical import, that is, its potential to resolve the impasse between Dembroff and Díaz-León. While I think that the dialectical upshot of the taxonomy-first view is evidence in its favor, it will be useful to supplement the aforementioned dialectical argument. Considering the influence of ideology on beliefs about race, I will argue that ideology tends to distort facts about the nature of race, while highlighting facts related to the taxonomy of race categories. By analogy, I will argue that this aspect of ideology provides reason to endorse the taxonomy-first view of sexual orientation. Additionally, I will briefly consider another argument for the taxonomy-first view of social properties in general.
To begin, I will explicate the phenomenon of ideology, broadly appealing to the work of Tommie Shelby. On Shelby’s account:
A form of social consciousness is an ideology if and only if (i) its discursive content is epistemically defective, that is, distorted by illusions; (ii) through these illusions it functions to establish or reinforce social relations of oppression; and (iii) its wide acceptance can be (largely) explained by the class-structured false consciousness of most who embrace it (Shelby 2003, p. 183–184).
Here, Shelby distinguishes between ideological and non-ideological “forms of social consciousness.” The idea is that any given society will – for purposes such as coordination and stabilization – foster a particular form of social consciousness. Broadly, a form of social consciousness is a “coherent system of thought” of significance to social practices.[19] On Shelby’s view, an ideology is a normatively problematic form of social consciousness (Shelby 2003, p. 160).
For Shelby, (i) ideological forms of social consciousness are “epistemically defective.” On this point, Shelby considers beliefs involved in the racial profiling of Black individuals by police officers in the United States, claiming that “profiling, by tapping into longstanding stereotypes, revives and reinforces ideological beliefs about the inherent tendency of Blacks towards violence and sexual aggression” (Shelby 2003, p. 176). While the above beliefs involved in racial profiling are false, Shelby notes that the epistemic defects of ideologies are often more subtle, involving “distorted, biased, or misleading representations of reality” (Shelby 2003, p. 166). Shelby lists a variety of ways in which ideologies generate distortions: “inconsistency, oversimplification, exaggeration, half-truth, equivocation, circularity, neglect of pertinent facts, false dichotomy, obfuscation, misuse of ‘authoritative’ sources, hasty generalization, and so forth” (Shelby 2003, p. 166). Still, I think that this list is missing something. For example, as Shelby seems to note in a point about sensationalized reports of the racial distribution of crime rates, ideology can generate distortions by unduly highlighting certain facts. On this point, consider crime rates related to drug possession. Ideology generates a distortion by highlighting the fact that Black Americans are more often convicted of drug-related crimes than white Americans, while downplaying relevant contextualizing information such as the racial distribution of police-initiated searches.[20]
Next, (ii) ideology plays a role in creating and sustaining oppression. As Shelby notes, there is an intimate connection between (i) and (ii):
[I]deologies perform their social operations by way of illusion and misrepresentation. What this means practically is that were the cognitive failings of an ideology to become widely recognized and acknowledged, the relations of domination and exploitation that it serves to reinforce would, other things being equal, subsequently become less stable and perhaps even amenable to reform (Shelby 2003, p. 174).
The connection between ideological distortion and oppression is evident in the above example of racial profiling, as ideological representations of the character of Black individuals, in part, explain racial disparities in assessments of “suspiciousness.” To be clear, Shelby denies that eliminating racist ideologies would automatically abolish the practice of racial profiling; still, ideological representations are part of the problem (Shelby 2003, p. 187).
Next, Shelby holds (iii) that there is often an error-theoretic explanation for why individuals hold ideological beliefs. In particular, the following social structural explanation is often apt: ideological beliefs sustain unjust social practices (Shelby 2003, p. 170–172, 183, 188).[21] In case that sounds a bit conspiratorial, consider the following, pervasive case: a police officer unjustly attacks a Black individual, yet members of the public falsely believe – say, even with unambiguous video record of the disproportionate violence – that the officer was properly responding to an objective threat. Why would otherwise epistemically responsible individuals hold such unjustifiable beliefs? I think that there is a social structural explanation: the false belief sustains the unjust social practice of racial profiling, stabilizing and perpetuating oppressive, white supremacist social orders.
Here, ideology tends to frustrate the apprehension of facts about the nature of race, while highlighting facts related to the taxonomy of race categories. As noted above, ideology (i) falsely represents the character traits of Black individuals, which (ii) plays a role in sustaining the oppressive practice of racial profiling. In particular, ideology represents Black individuals as “parasitic, angry, ungrateful, and dangerous” (Shelby 2003, p. 161). This involves a false representation about the nature of race, viz., that race is predictive of character. Still, this is not quite enough for ideology to “work” in the case of racial profiling. Ideology must also highlight the fact that Black is a race category.[22] And while it is true that Black is a race category, highlighting this taxonomy fact in concert with the false representation that race is predictive of character results in a distorted representation that sustains the practice of racial profiling.
As noted above, (iii) there are often error-theoretic explanations for why individuals hold ideological beliefs. In the context of ideological representations that sustain the practice of racial profiling, this amounts to the following: on account of ideology, social agents are disposed to hold the false belief that race is predictive of character and the true belief that Black is a race category, which, in concert, produce a distorted representation that sustains the practice of racial profiling. In this way, social agents are disposed to hold a false belief about the nature of race in concert with a true belief about the taxonomy of race categories. Taking seriously the influence of ideology on the race-related beliefs of social agents, the above discussion provides reason to endorse the taxonomy-first view of race.
By analogy, I think that the influence of ideology also provides reason to endorse the taxonomy-first view of sexual orientation. On this point, perhaps it is enough simply to note that I do not see any relevant dissimilarities between ideological distortion in the case of racial oppression and ideological distortion in the case of sexuality-based oppression.
However, in case such an analogy is suspect, I will wrap-up this section by advancing an argument for the taxonomy-first view of social properties in general. The following argument is intended as a sketch. Here, I do not aim to provide a comprehensive defense of the taxonomy-first view of social properties in general. Still, the following supplementary remarks might shore up the analogy between the epistemology of race and sexual orientation properties. To begin, note that it is plausible that social properties are partly created by representations. For example, Ron Mallon holds that a social category exists just in case:
Representation: there is a term, label, or mental representation that picks out a category of persons C, and that representation is associated with – and figures in the expression of – a set of beliefs and evaluations – or a conception – of the persons so picked out.
Social conditions: many or all of the beliefs and evaluations in the conception of the role are common knowledge in the community (Mallon 2016, p. 58).
The former condition speaks to the contents of representations associated with a particular social category, while the latter explicates the conditions under which such representations manage to create a social category.
Now, in some cases, members of a social category come to have features that correspond to common knowledge representations, especially as related to representations involving category-typical features. This is the familiar causal “looping effect” involved in much social construction (Mallon 2016, p. 82). For example, members of the category queer are represented as typically gender deviant, which plausibly, in part explains why queer individuals break from mainstream gender norms more often than straight individuals. Yet, this sort of causal looping is not a necessary effect of representations involving category-typical features. This is obvious by considering the aforementioned ideological representations of members of the category Black. In short, the fact that there are common knowledge representations involving category-typical features might provide probabilistic evidence that members of the represented category in fact exhibit such features (Mallon 2016, p. 89–93). Yet, if there are common knowledge representations involving category-typical features, then the represented category exists. Or, at least, this is the case on Mallon’s above account.
Here, I would like to suggest that there is an intimate link between the epistemology of social properties and the representational aspect of their creation. If Mallon is correct that common knowledge representations create social categories, then it is unlikely that social agents will generally be mistaken about the existence of the categories that they represent. This is the case even if social agents are profoundly confused about the constructed status of represented categories, or the features in fact exhibited by category members. Here is reason in favor of the taxonomy-first view of social properties. As Dembroff and Díaz-León agree that sexual orientation is amenable to ameliorative revision, it is plausible that Dembroff and Díaz-León also agree that sexual orientation is a social property. In that case, the above argument provides reason to endorse the taxonomy-first view of sexual orientation.
4 Conclusion
Above, I provided an argument in favor of the taxonomy-first view of sexual orientation. That is, I argued that beliefs about the taxonomy of orientation categories ought to have epistemic priority in relation to beliefs about the nature of sexual orientation. Here is opportunity for future research. Without appeal to disputed facts about the nature of sexual orientation, we can work to explicate the taxonomy of sexual orientation categories.
Acknowledgements
In addition to the anonymous referees at the Journal of Social Ontology, special thanks to Elizabeth Barnes, Ross Cameron, Dominick Cooper, and Brie Gertler.
Bibliography
Basu, Rima (2019): “Radical Moral Encroachment: The Moral Stakes of Racist Beliefs”. In: Philosophical Issues 29, p. 9–23.10.1111/phis.12137Search in Google Scholar
Dembroff, Robin (2016): “What is Sexual Orientation?”. In: Philosophers’ Imprint 16, p. 1–27.Search in Google Scholar
Díaz-León, Esa (forthcoming): “Sexual Orientations: The Desire View”. In: Keya Maitra and Jennifer McWeeny (Eds.): Feminist Philosophy of Mind. Oxford: Oxford University Press.Search in Google Scholar
Haraway, Donna (1988): “Situated Knowledges: The Science Question in Feminism and the Privilege of Partial Perspective”. In: Feminist Studies 14, p. 575–599.10.2307/3178066Search in Google Scholar
Haslanger, Sally (2016): “What is a (Social) Structural Explanation?”. In: Philosophical Studies 173, p. 113–130.10.1007/s11098-014-0434-5Search in Google Scholar
Haslanger, Sally (2012a): “Gender and Race: (What) Are They? (What) Do We Want Them to Be?”. In: Sally Haslanger (Ed.): Resisting Reality: Social Construction and Social Critique. Oxford: Oxford University Press, p. 221–247.10.1093/acprof:oso/9780199892631.003.0007Search in Google Scholar
Haslanger, Sally (2012b): “A Social Constructionist Analysis of Race”. In: Sally Haslanger (Ed.): Resisting Reality: Social Construction and Social Critique. Oxford: Oxford University Press, p. 298–310.10.1093/acprof:oso/9780199892631.003.0010Search in Google Scholar
Haslanger, Sally (2012c): “Future Genders? Future Races?”. In: Sally Haslanger (Ed.): Resisting Reality: Social Construction and Social Critique. Oxford: Oxford University Press, p. 248–272.10.1093/acprof:oso/9780199892631.003.0008Search in Google Scholar
Haslanger, Sally (2012d): “What Are We Talking About: The Semantics and Politics of Social Kinds”. In: Sally Haslanger (Ed.): Resisting Reality: Social Construction and Social Critique. Oxford: Oxford University Press, p. 365–380.10.1093/acprof:oso/9780199892631.003.0013Search in Google Scholar
Jeffers, Chike (2013): “The Cultural Theory of Race: Yet Another Look at Du Bois’s ‘The Conservation of Races’”. In: Ethics 123, p. 403–426.10.1086/669566Search in Google Scholar
Mallon, Ron (2016): The Construction of Human Kinds. Oxford: Oxford University Press.10.1093/acprof:oso/9780198755678.001.0001Search in Google Scholar
Shelby, Tommie (2003): “Ideology, Racism, and Critical Social Theory.” In: The Philosophical Forum 34, p. 153–188.10.1111/1467-9191.00132Search in Google Scholar
Spencer, Quayshawn (2014): “A Radical Solution to the Race Problem”. In: Philosophy of Science 81, p. 1025–1038.10.1086/677694Search in Google Scholar
United States Department of Justice (2015): “Investigation of the Ferguson Police Department”. www.justice.gov, July 2019.Search in Google Scholar
© 2020 Matthew Andler, published by De Gruyter, Berlin/Boston
This work is licensed under the Creative Commons Attribution 4.0 Public License.
