Sex and Sexual Orientation, Gender and Sexual Preference

: On what we can call the “folk” conception of sexual orientation, sexual orientation is understood as sex­based attraction, that is, as (partly) attraction on the basis of the perceived sex of the person to whom one is attracted. However, in recent discussions, philosophers have either added gender to sex as the basis of sexual orientation, or have altogether replaced sex with gender. Moreover, this addition or replacement has gone – mostly – unargued for. This paper argues that a sex­based conception of sexual orientation remains plausible because (1) it is compatible with gender­based attraction, which I argue can be understood as a preference; (2) the reasons so far on offer for adding gender to sex (or for replacing sex with gender) are not convincing; (3) we have good evolutionary and non­evolutionary reasons for thinking that sex is the basis of sexual orientation; (4) we have good reasons to not add gender as a basis of sexual orientation; and (5) a sex­based conception of sexual orientation accommodates the various sexual orientations that have recently appeared, orientations in addition to the folk two (or three) of heterosexuality, homosexuality (and bisexuality), such as pansexuality, skoliosexuality, gynsexuality, and androsexuality. What emerges is a conception of sexual orientation based on the sex of the people to whom we are attracted, but that understands sexual­based attraction in broad enough terms to include surgically altered bodies.


I. Introduction
According to what we can call the "folk" conception of sexual orientation, sexual orientations are based on sex and are two or three in number (homosexuality, heterosexuality, and perhaps bisexuality).Moreover, which sexual orientation one has depends on the sex of the people to whom one is attracted (the orientees) and on the sex of the person with the orientation (the orienter).For example, and roughly put: if X is male and is attracted to other males, X is homosexual; if X is female and is attracted to males, X is heterosexual; if X is attracted to both, X is bisexual.This folk conception used to be commonly held, though its popularity has diminished, especially among younger generations and academics, who seem to have a more expansive understanding of sexual orientation both in terms of its numbers and its bases.In this regard, some philosophers have started either to add gender (which they don't take to be the same as sex) to sex as a basis of sexual orientation or to altogether replace sex with gender (evidence for this shortly).This paper argues against this change.It argues that the sexbased understanding of sexual orientation is compatible with genderbased attractions understood as sexual preferences that operate within a person's overall sexual orientation, much like other preferences, such as age, height, and skin color do.Hence there is no need to revise the folk conception as far as its basis in sex is concerned.
To elaborate, my aim is to defend one aspect of the folk conception, namely, that sexual orientation is based on the sex of the orientee.I do so by rejecting the suggestion that we should revise it to incorporate the orientee's gender either as an additional basis to sex or one that replaces sex.I reject this suggestion because, first, there are no good reasons for it, and second, because there are good reasons against it.Additionally, I defend the sexbased conception by arguing that there are good reasons for using sex as the basis of sexual orientation; that such a conception accommodates genderbased preferences and many of the various sexual orientations that have recently cropped up; and that it does not prevent us from engaging in moral, political, or legal change when it comes to sexual minorities.My account aims to preserve the sexbased aspect of the folk conception only with respect to the sex of the orientee and is silent on the sex of the orienter (though I allude to its importance in Section IV).It is also compatible with the existence of more sexual orientations than the folk conception countenances. 1n this section, I start with some clarifications about sexbased attraction, sex, the concept of sexual orientation and its purpose, and the folk conception of sexual orientation.Here, I rely on the distinction between the concept of sexual orientation and specific conceptions of it (e.g., the folk conception and ones developed by philosophers, like Robin Dembroff's bidimensional dispositionalism discussed in Section III 2 ).I then provide evidence for the shift to gender in philosophical understandings of sexual orientation.In Section II, I argue for the compatibility of the sexbased account with genderbased sexual attraction.In Section III, I reject the reasons on offer for why gender should be another basis of sexual orientation.In Section IV, I provide reasons in support of sex as a basis and, in Section V, I provide reasons against gender as a basis.

I.A. Sex and SexBased Attraction
Sexual attraction to others is prevalent among adult human beings.Through its lens, the world's objects are divided into roughly two kinds: those with which we would like to have sexual contact, and those with which we do not.Sexual attraction is the attraction that most adults feel upon happening upon an object with which they desire sexual contact.Moreover, sexual attraction is overwhelmingly sexbased: human beings are attracted to each other 1 On the importance of including the sex of the orienter, see Kathleen Stock, "Sexual Orientation: What Is It?" Proceedings of the Aristotelian Society 119 (3): 295-319, Section IV. on the basis of sex, whether the sex of the person to whom they are attracted is opposite to theirs, the same, or both.In this regard, we do not differ much from our nonhuman mammalian counterparts, whose sexual attractions are also overwhelmingly sexbased. 3ow do I understand "sex"?Biologically speaking, and in anisogamous species like human beings, one's sex is one's reproductive strategy (including one's chromosomes and hormones) to produce either large gametes (females, who produce eggs) or small gametes (males, who produce sperm).These strategies, with few exceptions, result in typically anatomical male and female bodies.I use "sex" in "sexbased attraction" to refer to attraction to these bodies, including their visible sex markers that are the result of these reproductive strategies, including, but not confined to, genitalia, breasts, and buttocks.So what is crucial to sexual attraction is not the orientee's ability to produce one or the other kind of gamete, but how this ability is, so to speak, reflected in the orientee's body -whether the body is male or female.This is because it is the sexed body, with its visible sex markers, on which sexual attraction is based and to which it is directed.This understanding of sexual attraction conforms to human beings' usual experience of sexual desire and attraction for each other.
Because it is the visible sexed bodies and their parts that are the bases and objects of sexual attraction, such attraction can be directed to bodies that have been surgically altered on the "outside," including trans people who have undergone sex transition surgery.For example, just as heterosexual men can be attracted to cis women who had breast augmentation plastic surgery, they can be attracted to trans women who have surgically externally altered their bodies from male to female (assuming everything else equal, such as not having beliefs that such bodies are unattractive).
Moreover, sexbased attraction does not rule out attraction on other bases: X's attraction to Y because Y is male only partially explains why X is attracted to Y. Y's other particular features (and perhaps also external factors such as where and when X and Y meet) are needed to fully explain X's attraction.Additionally, sexbased attraction does not make an anomaly of people who are attracted to, say, androgynous people, because their attraction is still sexbased.To understand this point, consider the difference between (a) and (b): (a) Sex is not a basis of attraction: some people's attraction to others is based on something other than sex.(b) Either (or, if there are more than two, any) sex is a basis of attraction: some people's attraction to others is sexbased but such that no sex is preferred over another.
(a) altogether disregards sex as a basis of attraction and allows for nonsex bases, such that some people are attracted to nonhuman beings, including plants and inanimate objects (e.g., objectophilia and dendrophilia -respectively, sexual attraction to inanimate objects and to trees -are apparently such orientations).(b) does not disregard sex, and it includes bisexual attraction and attraction to androgynous people (more on this in Section II).So to claim that attraction is sexbased is to claim the justdiscussed point (b), though it could allow for the possibility of nonsexbased attractions.

I.B. The concept of sexual orientation
I take sexual orientation to be that concept which sorts human sexual attractions into types based on the kind of object found to be sexually attractive -the kind to which one is sexually oriented: X has sexual orientation S1 because X is sexually attracted to objects of kind A, whereas Y has sexual orientation S2 because Y is sexually attracted to objects of kind B. Thus, sexual orientation's function is taxonomic, dividing people into types based on their sexual attractions. 4For this function to succeed, the number of orientations has to be limited at some point.It need not be two or three, per the folk conception, but it cannot be so many as to defeat the very point of categorization.Of course, the specific number of orientations cannot be decided beforehand, and it will emerge as the result of empirical and theoretical investigations.The trick is to land on criteria that make sense for the categorization of people's sexual attractions.One crucial claim of mine is that one essential criterion is the sex of the orientee and not their gender.5

I.C. The folk conception
I elaborate the folk conception using four additional claims that are compatible with its basic point and that reflect common experiences with sexual attraction.First, on the folk conception, the orientee's sex is one (albeit essential) reason why the orienter is attracted to the orientee.In individual cases of attraction, if, say, X is male, is homosexual, and is attracted to Y (also male), then one crucial reason for the attraction is that Y is male.Strictly speaking, however, and because sexual attraction is typically formed based on sensory cues (how Y looks, sounds, acts, etc.), X is attracted to Y because X perceives or believes that Y is male.That is, although sex is the basis of sexual orientation, for attraction to occur it is because the orienter believes or perceives that the object of their attraction is of the "right" sex, though X need not be, and typically is not, selfaware of this belief or perception (X does not usually think to X's self, "I believe that I am perceiving a male body").Of course, X might be mistaken in X's belief, in which case X's attraction might dissipate.For example, if X is attracted to Y thinking that Y is male but Y is actually female, then we would expect X's attraction to evaporate if X is not attracted to females or in the absence of special circumstances. 6Moreover, believing that Y is of the "right" sex is necessary, but not sufficient, for attraction because Y will need to have other properties on which basis X is attracted to Y. Sexual attraction does not differ in this respect from 4 This understanding of the concept is found in every philosophical discussion of sexual orientation.For example, John Corvino writes, "The attempt to categorize people into their sexual interests appears to have ancient roots" ("Orientation, Sexual," in Sex from Plato to Paglia: A Philosophical Encyclopedia, edited by Alan Soble [Westport, CT: Greenview Press, 2006], 728-734, at 730; my emphasis).Edward Stein also agrees that sexual orientation divides people into kinds, and focuses on whether they are natural or social (The Mismeasure of Desire: The Science, Theory, and Ethics of Sexual Orientation [Oxford and New York: Oxford University Press, 1999], ch.3).And Dembroff's view, discussed in Section III, considers taxonomy one crucial goal of the concept.other desires and (even) emotions, such as X's desire to eat a dish because X believes it to be tasty.
Second, sexbased attraction need not be just to genitalia (or, in the cases of male heterosexual desire and female homosexual desire, just to women's genitalia and breasts), but also to other body parts and to the whole body, just as long as the body is perceived to be of the desired sex (indeed, some people might not be attracted to genitalia at all, some even finding them offputting).We can say that the orientee's sexed body circumscribes the orienter's attractions, excluding some and including others.So if John is gay with a foot or armpit preference, he would be attracted to men's feet or armpits, not to women's.John's attraction to males is because they are males: it is their maleness that is the basis of John's attraction to others, and from which John's other, specific sexual preferences emanate.
Third, it is compatible with the folk conception to not claim universality on its part.This is so in two ways.First, as mentioned, it is possible for some sexual orientations to not be based on sex at all.Second, sexual orientations need not be always absolutely discrete.Someone can be heterosexual yet have an occasional attraction to members of the same sex.For instance, in prisons heterosexuals are sometimes willing to have sex with members of the same sex due to various factors, including sexual deprivation and assertion of domination.College students are also famous for sexually experimenting.And sometimes people are just sexually curious.In short, the folk conception is compatible with orientations admitting of exceptions. 7ourth, sexual orientations are different from sexual preferences, most of which are also sex based but operate within a person's sexual orientation, as the above examples of foot and armpit preferences indicate.The distinction between orientations and preferences is fairly common, both among lay people and academics: it is commonly accepted that people have "types" and that these types "operate" within people's sexual orientations, such as a heterosexual woman preferring the "biker" or "surfer" type, and a homosexual woman preferring slender women. 8This distinction is crucial for this paper's argument, as becomes clear below.

I.D. Philosophers on sexual orientation
Philosophers have not discussed sexual orientation much, but when they have, they have tended to understand it as attraction based on sex.One definition from 2006 is that a "person has a sexual orientation by virtue of the sex of his or her actual or preferred sexual partners."Another, from 2013, is that sexual orientation "is usually understood to refer 7 This is not an ad hoc move on my part.It reflects the fact that many people have stable orientations yet "dabble" with the other sex.Lisa Diamond has argued that women's sexuality is somewhat fluid, which she defines as "situationdependent flexibility in women's sexual responsiveness."But she also sees fluidity as "a component of female sexuality that operates in concert with sexual orientation to shape women's sexual desires" ( to one's persistently recurring sexual desires for members of the sex that attracts one."A third, from 1988, is that a homosexual "is someone, male or female, who is erotically attracted to members of his/her own sex."And "sex" is in the title of one excellent anthology on homosexuality from 1997: Same Sex: Debating the Ethics, Science, and Culture of Homosexuality. 9 Although these philosophers, as they should, raise questions (left mostly unanswered) about the adequacy of these definitions, the definitions reflected the folk understanding of it. 10here is now a noticeable shift in the new philosophical literature on understanding sexual orientation, which is that the word "gender" is either added to "sex" or replaces "sex."Sexual orientation is now understood to be sexual attraction to others on the basis of their sex and gender or on the basis of only their gender, despite the fact that there is no agreement on what "gender" means.Recent examples are some of the essays in The Routledge Handbook of Philosophy of Sex and Sexuality (2022), 11 in which almost all of those that deal with sexual orientation take it as a given that sexual orientation is based partly or exclusively on gender."Take it as a given" is deliberate wording on my part, because the authors do not question this new understanding of it (as philosophers should).So whereas the handful of past philosophers who started with sexbased understanding of sexual orientation discussed whether it is an adequate understanding, these new authors do not address the philosophical ins and outs of understanding sexual orientation's basis on gender.One crucial exception is Robin Dembroff's 2016 essay, which argues for and develops an account of sexual orientation based on both the sex and gender of the orientee (and omits reference to the sex or gender of the orienter).I discuss it below. 12wo possible explanations for this shift in this recent philosophical trend are, first, that the definitions seem to reflect the new cultural tendencies that "queer" organizations are adopting, including using "gender" in place of, or in addition to, "sex."Governmental institutions are also adopting these definitions.For instance, in my home country, the United States, the Department of Labor defines "sexual orientation" as referring to "an individual's physical, romantic, and/or emotional attraction to people of the same and/or different gender." 13The second explanation is Dembroff's own essay, which has been highly influential in this small subfield of philosophy, evidenced by the fact that almost every work on sexual orientation published after 2016 has cited the essay (and rightly so, given its claims that gender is a basis of sex and that we should dispense with the orienter's own sexgender).
Consider now some examples of this shift among philosophers.Joy Cameron understands heterosexuality "as a robust, primary desire for romantic and/or sexual 9 The first is John Corvino's ("Orientation, Sexual," 728).The second is Wilkerson's ("What Is 'Sexual Orientation'?"195).The third is Michael Ruse's (Homosexuality, 1 [Oxford: Basil Blackwell, 1988).The fourth is the anthology edited by John Corvino (Lanham, MD.: Rowman & Littlefield, 1997). 10The main exception to this past trend is Stein, who "provisionally" defines sexual orientation as "indexed to the sexgender of the people a person is attracted to" (The Mismeasure of Desire, 34).But the book's arguments and conclusions seem to ignore the gender part and discuss sexual orientation in terms of sex. 11The Routledge Handbook of Philosophy of Sex and Sexuality, edited by Brian Earp, Clare Chambers, and Lori Watson (New York and London: Routledge, 2022).An even more recent example is Arina Pismenny, "Pansexuality: A Closer Look at Sexual Orientation," Philosophies 8 (4): Article 60, 2023 (https://doi.org/10.3390/philosophies8040060). 12Dembroff, "What Is Sexual Orientation?" 13 https://www.dol.gov/agencies/ofccp/faqs/lgbt#Q19.
intimacy with persons who are of a different sex and/or gender."14Christie Hartley understands sexual orientation under "current conditions" to involve "genderbased attraction" but believes that in a world without gender sexual orientation might be sexbased, which, according to her, would be okay from a feminist perspective. 15Esa DíazLeón understands sexual orientation as "determined by … the sex(es) and gender(s) of the people for whom the person is disposed to experience sexual desires."16As a final example, Lisa Diamond characterizes it as based on only gender, not sex, because it is unclear that sex is what drives the kind of attraction found in talk of sexual orientation since it is often at "least in part socially coded gender expressions that may or may not 'line up' with sex in a straightforward way that draws us to a person."17Note that Diamond's reason cuts both ways, for we might as well claim that it is at least in part the orientee's sex that draws us to them, so it is unclear why only gender is the basis of sexual orientation in Diamond's account.
It is these kinds of claim that this essay challenges by defending a sexbased account of sexual orientation.My motivation for challenging these claims is that this shift to "gender" talk is puzzling for at least two reasons.First, it is unclear what "gender" means, an unclarity that should not be news to philosophers as they are aware of it given their debates about gender and the various accounts of it that they have offered.These accounts include gender as identity (which gender one identifies with in light of existing social gender categories), gender as socialization (being socialized into gender norms for men and women), gender as a social position (that one's gender depends on which social position one occupies at any given time given the gender categories of one's society), and gender expression (one's gender presentation given the social conventions surrounding gender in one's society). 18oreover, all these conceptions of gender face difficulties and objections. 19So when we claim that sexual orientation is genderbased attraction, either we are unclear on what we are saying or we are saying that sexual orientation is based on any or all these things that we now think gender is, despite their attendant difficulties.Though the latter option is conceptually possible, it can also be seen as avoiding the difficult issue of pinning down gender and deciding whether it is indeed a basis of sexual orientation.
The second reason why this shift is puzzling is that, with the exception of Dembroff's account, we have not been given reasons for the addition of gender to sex, let alone for gender replacing sex, especially since the sexbased understanding of sexual orientation does not, and never has, crowded out the role that some forms of gender can play in sexual attraction -as I argue below, the sexbased understanding of sexual orientation is compatible with attraction to others on the basis of their gender, especially gender as gender presentation.
So I will argue that the sexbased conception of sexual orientation remains plausible.This is so for four reasons: (1) it is compatible with attraction to others on the basis of gender, a compatibility that helps us see how it handles some alleged sexual orientations that have recently appeared on the cultural scene; (2) the reasons so far given for adding gender are not convincing; (3) we have good reasons for thinking that sex is the basis of sexual orientation; and (4) we have good reasons to not add gender to sexual orientation.I elaborate these reasons in turn.

II. SexBased Sexual Orientation Is Compatible with GenderBased Attractions
To understand why a sexbased understanding of sexual orientation is compatible with attraction on the basis of gender, we need to rely on the abovementioned crucial distinction between sexual orientations and sexual preferences.Sexual preferences are preferences that someone typically has within the parameters of their sexual orientation.For instance, a gay man is likely to have sexual preferences that reflect the kind of men to whom he is attracted, such as having a certain height, body weight, body shape, skin tone, and hair color.He can, more relevant to my point, be attracted to men that appear or act in certain ways as far as their gender presentation is concerned: he might be attracted to, say, hypermasculine men (think, e.g., bald, bearded, hairychested, tattooed, muscular); or to feminine men (think, e.g., slender, young, hairless, effeminate).Gay men can then have these different preferences based on gender while sharing the same sexual orientation.If gender presentation is one way of understanding gender, and if gender presentation is characterized by properties such as those in the justlisted examples, the sexbased understanding of sexual orientation does not preclude having genderbased attractions, understood as preferences.Moreover, gender as gender presentation is surely a crucial form of gender since most people's sexual attraction to others partly depends on how the latter gender present.Thus the sexbased understanding of sexual orientation can construe genderbased attraction as a preference.
Additional support for this point comes from the historical fact that gender presentation has varied, and continues to vary, across cultures and times, as sexbased attraction has remained constant.Straight women were attracted to wigwearing men at the end of the 17th century in France, just as they were to mascarawearing men in ancient Persia.And straight men were attracted to headtotoe covered women in Puritan New England, just as they are to thongclad ones on the beaches of contemporary Greece.Sexbased attraction remains constant across the variations in gender presentation while accommodating these variations.Such variations in genderbased attractions, then, can be plausibly understood as preferences that vary depending on culture and time period.Thus, genderbased attractions can be plausibly construed as preferences.
Understanding "gender" as referring to forms of gender other than gender presentation is also compatible with the sexbased understanding of sexual orientation, even if they sound farfetched.Consider two examples.Someone can have a preference for males who, without any surgical alteration or even other forms of gender presentation (such as clothing and hair style), selfidentify as women.This would be a preference based on gender as gender identity, and their orientation would be either straight or gay (depending on their own sex), given that, by stipulation, the orientee has not undertaken any visible changes to their sex or general appearance. 20And someone can have a preference for transwomen because their gender presentation and prevalent social norms put them at one of the lower rungs of the social hierarchy, as is often claimed about the social position of trans people (this would be preference based on gender as social position). 21How common such preferences are is an empirical issue, but -rare or common -the sexbased understanding of sexual orientation need not exclude them.
Note that the argument about the compatibility of genderbased attraction with the sexbased conception of sexual orientation does not depend on having a true definition of gender (I did, after all, claim that it is unclear what gender is).It can be understood instead as a conditional claim: whether by gender we mean this or that, it is compatible with the sexbased conception.
Can the sexbased conception of sexual orientation accommodate some recently identified sexual attractions?Its ability to do this supports its plausibility by showing that it is in step with the times, so to speak.
Consider first gynandromorphophilia, which is the attraction, almost always among cismen ("gynandromorphophiles"), to transwomen who gender present as feminine (gynandromorphs): males (usually slender) who have long hair, breasts, and generally effeminate features but who retain their penises.It might seem that gynandromorphophilia is a genderbased attraction because the men in question are attracted to males who gender present as women.If so, the sexbased conception might not accommodate it.However, this is too quick because although gynandromorphs are biologically male, their visible sex markers are mixed (they have breasts and penises).Since the men attracted to them are attracted to their bodies, it is not obvious that the attraction is solely genderbased.Sex underlies it.
The empirical evidence indicates that most gynandromorphophiles are straightidentifying men attracted to cisgender women and show little to no interest in cisgender men, though they are sexually aroused by their sexual partners having a penis. 22This indicates that although the folk conception of sexual orientation cannot account for this attraction, a more expansive sexbased account -one that does not insist that, say, breasts always have to come with vaginas -can.We can then think of gynandromorphophiles as either having a sexual orientation all on their own or, if they are also attracted to cisgender women (as many of them are), as having two sexual orientations: they are both heterosexual and gynandromorphophiliac, much like we can think of bisexuals as having two sexual orientations: they are both heterosexual and homosexual. 23onsider next pansexuality, which can be characterized as "romantic and sexual attraction to all genders, or regardless of gender, on the basis of 'hearts not parts.'" 24his characterization understands pansexuality as attraction to all genders, to all sexes, or to both.If it is gender attraction, then the sexbased conception of sexual orientation accommodates pansexuality's attraction by claiming that pansexuals are attracted to all genders, much like bisexuals are attracted to both men and women. 25Pansexuals would differ from bisexuals in being attracted to more than the two main genders.If pansexuality is also attraction to all sexes, then pansexuality would also be attraction to people whose phenotypical appearances are the result of intersex conditions.This would still be sexbased attraction, except that the attraction is not confined to the typical sexed bodies of nonintersexed males and females.It is unclear whether pansexuality is about gender or sex or both.But the sexbased conception is able to account whichever form it takes.Thus, as far as sex is concerned, if pansexuality is attraction to all sexes, it would still be sex based.
Pansexuality, then, does not require a rethinking of sexual orientation as sexbased.However, both pansexuality and gynandromorphophilia revise the folk conception of sexual orientation by changing the number of sexual orientations: sexual orientations are more than two (or more than three, depending, again, on how we understand bisexuality).In this regard, my argument does not preserve that part of the folk conception of sexual orientation according to which there are two or at most three orientations.
Consider next androsexuality (attraction to masculinity), gynosexuality (attraction to femininity), and skoliosexuality (attraction to gender queer, transgender, or nonbinary people). 26Androsexuality and gynosexuality, insofar as they do not reject any particular sex as attractive, can be considered forms of bisexuality or pansexuality but with genderbased preferences, namely, for people whose gender is masculine (androsexuality) or feminine (gynosexuality).Skoliosexuality, also insofar as it does not reject any particular sex as attractive, can be considered a form of bisexuality or pansexuality, but with a preference for people who don't have a clear gender presentation or for those whose gender presentation does not match their sex or societal expectations of gender presentation.
Consider next asexuality (the lack of sexual attraction to others), 27 polyamory (being in, or the desire to be in, simultaneous multiple relationships with other people, some or all of whom can be in relationships with each other), 28 and demisexuality (attraction to others after developing romantic feelings for them). 29Are they sexual orientations on the sexbased conception?
Whether asexuality is an orientation depends on how broadly or narrowly we understand sexual attraction.If sexual orientation is attraction or lack of attraction on the basis of sex, then asexuals would have a sexual orientation, but understood, so to speak, negatively, as not being attracted to others on the basis of their sex (or any other basis).If sexual orientation is (positive) attraction to others on the basis of their sex, asexuals would not have an orientation because they do not experience this attraction.I leave this issue open but note that even if asexuals do not have a sexual orientation, they can still have a sexual identity: they are individuals who do not feel sexual attraction to others.Moreover, they can be considered a group on the basis of this identity, which might be needed for legal or social recognition.
Polyamory is either the de facto state of being in multiple simultaneous relationships or the belief or desire that one should be in such relationships because, say, monogamy is unethical or less than ideal.Polyamory, then, is not about sexual attraction but about the number of simultaneous relationships that one is or desires to be in.Thus, it does not seem to be a sexual orientation on the sexbased account (or on any other basis of sexual attraction), and a person of any sexual orientation can be polyamorous.Although ameliorative arguments can be given for polyamory's being an orientation, they are not convincing, as I show in Section III.A regarding gender as a basis of sexual orientation. 30emisexuality is sexual attraction to others after developing romantic feelings for them.Because demisexuality is only a temporal variation of any one of the existing orientations, demisexuals can be heterosexuals, homosexuals, bisexuals, pansexuals, and gynandromorphs, but such that their ability to experience sexual attraction is contingent on their first experiencing romantic feelings (this is the temporal part) for those to whom they will eventually be sexually attracted.
Before concluding this section, let's consider how the sexbased conception handles anomalous cases.Suppose that Marge, a heterosexual female, is attracted to someone -Pat -whom she believes is a woman but comes to find out that he is male.Marge's attraction to Pat is based on her belief that Pat is a woman, given Pat's gender presentation.What is Marge's orientation? 31The case is underdescribed as presented, because we need to know whether Marge's attraction is oneoff and what happens to her attraction when she knows that Pat is male.(a) If Marge experiences attractions to women occasionally or frequently, then we can say that she is bisexual.Her attraction to Pat would then be like her other womenattractions, except that in this case her belief that Pat is a woman is mistaken.Now, if her attraction to Pat dissipates after the discovery, Marge's gender preferences cleave to the gender binary: she is bisexual, but she likes her men looking like men and her women looking like women.If her attraction persists, then Marge would be a bisexual with no gender preferences (she is skoliosexual).Finally, (b) if her attraction to Pat is one off, Marge is heterosexual and her attraction to Pat is an exception (see the third point in Section I.C.).
The sexbased account, then, can handle the above sexual attractions: some can be understood in ways that do not necessitate the revision of the sexbased understanding of sexual orientation; some can be understood as sexual orientations all on their own; others as gender preferences that exist within the sexbased attraction framework; while others seem to not be sexual orientations at all.

III. There Are No Good Reasons for Adding Gender as a Basis of Sexual Orientation
The recent literature is unclear as to why gender should be a basis of sexual orientation.The most important account to date of why sexual orientation should be based on both sex and gender is the aforementioned one by Dembroff, who calls the account "bidimensional dispositionalism" (BD).According to BD, gender -specifically, gender expression 32 -is an additional basis to sex.Indeed, "bidimensional" is the name of this account precisely because "sexual orientation tracks both gender and sexattraction"; "sexual orientation concerns both gender and sexattraction"; and, "sexual orientation pertains to both gender and sexattraction" (10, 31, and passim).Because Dembroff's account has been influential, and because it provides explicit reasons for why gender should be the basis of sexual orientation, I devote some space to it.BD states that "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[ses] and gender[s] of persons S is disposed to sexually engage under these conditions" (31).BD, then, has two parts, the first concerning dispositions (which I do not discuss), and the second concerning the bases of sexual orientation, which are the sex and gender of the orientees.BD does not require that every person must have a sexual orientation based on both sex and gender; instead, it leaves room for someone to have an orientation based on only one or neither.What is crucial is that sexual orientation is based on sex and gender, not other features of a person (31-32).
Dembroff offers two reasons for why sex and gender are such bases.The first is general and has to do with why BD should be accepted as an account of sexual orientation.The second is more specific and has to do with why we should not regard either sex or gender as a mere sexual preference.

III.A. Dembroff's reasons for BD
Dembroff''s argument for BD seems to be as follows: (1) A plausible (true, acceptable, good?) account of sexual orientation should serve a number of crucial purposes.(2) BD serves these crucial purposes.(3) Therefore, BD is a plausible (true, acceptable, good?) account of sexual orientation.Dembroff identifies four purposes that an account of sexual orientation should "ideally" serve.First, it should clarify "the criteria for ascribing sexual orientation, as well as how these criteria translate into a taxonomic schema of sexual orientation."Second, it should be "consistent with relevant social scientific research, in particular, research concerning sex and gender."Third, it should reduce or eliminate "the presumption that cisheterosexuality is the normatively standard sexual orientation and all queer sexual orientations are normatively deviant."Fourth, it should be "conducive for establishing legal and social protections for persons who have queer sexual orientations" (7-8).Unlike the first and second purposes, the third and fourth clearly embody political, social, moral, and legal goals, hence reflect Dembroff's methodology of conceptual engineering, according to which the development of a concept depends on which purposes we would like the concept to serve. 33Moreover, Dembroff is explicit that the first two purposes are important because of the last two: the first two are crucial "for developing a concept that serves the social and political purposes stated in [the second and third purposes]" (8).
Dembroff lists the third and fourth points as purposes that a plausible conception of sexual orientation should meet because the current conception of sexual orientation excludes queer sexual orientations and because it has a difficult time classifying them -"How should a gender nonconforming, transgender, or intersex person (or their partners) describe their sexual orientations?"(8) -a reason that Dembroff repeats in addressing why gender or sex are not sexual druthers (which I discuss in Section III.B).Dembroff thus emphasizes "the need to recognize the community of persons who are exclusively (or strongly) attracted to transgender individuals, or who are themselves transgender and seeking persons with these attractions."This indicates to Dembroff that "the current taxonomy of sexual orientation simply fails to capture their sexual orientation, since they experience strong or exclusive attraction specifically to persons who are not cisgender" (18).
The idea, then, is that if we conceive of sexual orientation as referring to the orienter's sex (in addition to the orientee's), if we conceive of it as only sexbased, and if we conceive of sex as binary, then neither an intersex nor a gender nonbinary person would have a category under which to place themselves in this taxonomy.Moreover, if one is cisgender but is attracted to intersex or gender atypical people, one would not be able to locate one's sexual orientation under the current scheme simply because it is not locatable.Finally, it is also unclear, according to Dembroff, how laws against discrimination would protect such individuals given the current conception of sexual orientation.Thus, a revised conception would help meet these social, political, and legal purposes by reducing, or even altogether getting rid of, what Dembroff calls "cisheterosexuality" -the socialcummoral fact that heterosexuality based on sex (not gender) is the norm and other sexual orientations are not.
There are, however, two main difficulties with Dembroff's approach.The first is that the third and fourth purposes, without proper argumentation, beg the question against those who do not consider gender (as opposed to sex) the basis of sexual orientation.That is, granting that we want our conception of sexual orientation to do political and moral work, it is unclear why this work needs to specifically dislodge cisheterosexuality as the norm, as opposed to instead dislodging only heterosexuality.To claim that we need to downgrade the former, as opposed to the latter, assumes that sexual orientation should track both sex and gender.But this is what needs to be argued for.The point is not that there is no such thing as cisheteronormativity or that we should not morally address it.The point is that we need a separate argument for the idea that both sex and gender are bases of sexual orientation.Dembroff does provide such arguments, though only regarding why sex and gender are socially salient categories and not mere preferences (see Section III.B.).
The second main problem with the third and fourth purposes is the very assumption that we want our conception of sexual orientation to do political and moral work.This assumption is fraught not because homophobia and transphobia are off the moral hook, but because it is unclear why the burden of (morally, politically, etc.) addressing them should be on the shoulders of a conception of sexual orientation.After all, a plausible suggestion is to provide a conception of sexual orientation that is apolitical or amoral, and on its basis build political and moral arguments.This is, moreover, what philosophers have done in the past.They have assumed that a proper conception of sexual orientation classifies people into attraction based on sex, and then argued that nothing, as such, about nonheterosexual people or their nonheterosexual sexual actions merits moral censure or political and social discrimination. 34Moreover, this approach has not impeded political gains for gay people: samesex marriage, for example, is the law of the land and employment discrimination on the basis of sexual orientation is universally legally prohibited in the United States (under Title VII, especially).And, crucially, all this was achieved without having to change the conception of sexual orientation so that it is inherently political.If there are continued social and other forms of discrimination against the populations that Dembroff is concerned about, it might be simply a matter of time for us to overcome them, just as it was a matter of time before political and moral gains were attained for gays and lesbians, and might have nothing to do with our conception of sexual orientation.
There is another reason why the assumption that our conception of sexual orientation should do moral and political work is fraught, namely, that it is unclear whether it can succeed.Indeed, it faces problems both as a conceptualcummoral claim and as an empirical one.First, conceptually and morally speaking, sexual orientation is not sufficient for such protections: zoophiles, necrophiles, and pedophiles should not be socially accepted or legally protected just because their sexual attractions are considered sexual orientations. 35Empirically speaking, it does not seem likely that the general population would come to accept such sexualities if they were to be declared sexual orientations (notwithstanding attempts to the contrary by the new movement of "minor attracted individuals" whose name seems designed to elicit social acceptance 36 ).Of course, there is no consensus that such sexual attractions are orientations.But my claim is conditional: if they were to be considered sexual orientations, that alone would not be sufficient for social protections.
Second, sexual orientation is not necessary for social protections, because the answer to a proper recognition of people with sexual desires for, say, gender nonbinary people, need not be the expansion of the taxonomy of sexual orientation.It could instead be the expansion of the taxonomy of people's sexual preferences or identities.Such an expansion would move away from both cisheterosexuality and cishomosexuality as genderbased preferences to include other genderbased attractions, thereby leaving intact in the process the taxonomy of sexbased sexual orientations.In this way, people's genderbased preferences would not be conceived of as exhausted by the gender expressions of (and for) cis people, and would include the gender expressions of (and for) trans and gender nonconforming people.Moreover, this would not be a radical shift given that we have always known that there are various preferences within each sexual orientation.
Empirically speaking, if people who are attracted to, say, gender nonbinary people are socially unaccepted or might be legally discriminated against, the cause need not be a particular conception of sexual orientation (as based only on sex or on both sex and gender), but society's slowness in understanding and accepting new preferences.That is, social acceptance of various genderbased sexual preferences might have less to do with our conception of sexual orientation and more to do with various social factors.Indeed, I suspect that social acceptance does not depend on whether a type of sexual attraction is a preference or an orientation, but on its moral legitimacy.To return to the age example, suppose that we claim that age, and not only sex, is the basis of sexual orientation, so that there are sexual orientations like pedophilia and gerontophilia (sexual attraction to old people).This is unlikely going to change the fact that pedophilia will continue to be viewed as morally abhorrent, because what society cares about is which actions or desires are harmful to others, regardless of whether the actions or desires stem from an orientation and regardless of whether the orienters were "born this way" -after all, being born a certain way does not as such provide moral justification for immoral actions and desires.So it is unclear, to say the least, whether claiming that gender expression is a basis of sexual orientation would afford social protections to those with the alleged sexual orientation.

III.B. Dembroff's reasons for why gender is a basis of sexual orientation
I mentioned that the political purposes that Dembroff lists require an argument to the effect that sexual orientation should be based on sex and gender.Dembroff provides two such arguments.The first is that sex and gender are apt bases for sexual orientation because they are socially significant categories when it comes to sexual orientation (12).Dembroff believes that, ultimately, this social significance is arbitrary, stating that "[i]t is not clear why attraction to certain sexes or genders is considered relevant to one's sexual orientation, but not attraction to a certain hair color, race, or economic status."However, Dembroff goes on to accept that "sex and gender are, for better or worse, particularly salient social categories with respect to sexual orientation," so "we find ourselves in the position of classifying persons' sexual orientations on the basis of their sex and genderattractions and not on the basis of their other attractions" (12).So despite the fact that the causes of or reasons for their salience are unclear, sex and gender are socially salient categories when it comes to our beliefs about sexual orientation, and this is what we have to work with.
The difficulty with this reasoning, however, is that if indeed sex and gender are arbitrary bases of sexual orientation, Dembroff opens up their account to the charge of incoherently continuing this arbitrariness at the expense of those people whose sexual attractions have additional bases to sex and gender.Why not race, age, economic class, and even species for that matter?Dembroff seems to think that because "persons with particular sex and genderattractions [are] more vulnerable to discrimination than persons with attractions to persons with a certain hair color or economic status" (12), this justifies focusing our conception of sexual orientation on sex and gender.But the examples of eye color and economic status make the case easy for Dembroff.Examples of preferences on the basis of age, race, and species indicate that they, too, face social, moral, legal, or other forms of discrimination (some justified and some not).Indeed, it is ironic for Dembroff to reject such bases: because Dembroff wants to engineer the conception of sexual orientation, why not make it more inclusive?By opening the door to gender as an additional basis to sex, and by admitting that sex and gender are ultimately arbitrary bases, Dembroff's account would not be able to keep sexual orientation's bases nonarbitrarily confined to sex and gender.
Dembroff's second argument for why gender and not only sex is a basis of sexual orientation is that because genderbased attraction is not a mere preference, we should consider it part of sexual orientation.Dembroff gives two reasons for this claim.The first is the "frequency with which people experience sexual attraction to not only individuals with particular gendered features, but also to individuals with particular primary and secondary sex characteristics."Dembroff adds that this is "simply the observation that, in addition to attraction to particular genders, persons can also be exclusively interested in partners with particular sex characteristics" (18, my emphases).
This reason, however, is not convincing.Given that sexual attraction has always been partly based on gendered features, why decide to make gender a basis of sexual orientation as opposed to a preference?That is, Dembroff does not argue why genderbased sexual attractions are not preferences.After all, we have always known that people usually come genderpackaged, so to speak, a fact considered perfectly compatible with a sexbased understanding of sexual orientation -for instance, understanding homosexuality as sexbased includes genderbased attractions, such as butch lesbians' attraction to feminine women and Ancient Greek men's attraction to barely pubescent boys.So to claim that genderbased attraction exists says nothing new.This reasoning can be further supported by substituting another basis of attraction for either sex or gender in Dembroff's quotation, such as racial belonging or age, as follows: the "frequency with which people experience sexual attraction to not only individuals with particular gendered features, but also to individuals with particular racial (or age) characteristics."By Dembroff's reasoning, these should also be part of our conception of sexual orientation, especially since Dembroff believes that sex and gender are arbitrary bases.
Thus, that there are individuals with attractions to transwomen or ciswomen does not suffice to show that gender is a basis of sexual orientation, only that it is a basis of their preference.This remains true even if the preference, as Dembroff emphasizes, is exclusive, because exclusivity does not as such add bases to one's sexual orientation.A gay man who has an exclusive preference for tall men does not render his sexual orientation "gaytall"; he's still gay, albeit with an exclusive height preference.Moreover, there are far more individuals whose preferences for gender expressions are not exclusive: many straight women, for example, are attracted to men with long hair (think rock stars, hippies, surfers) even if they prefer men with short hair.So Dembroff's first reason does not convince that sexual orientation should be based on gender in addition to sex.
Dembroff's second reason for why genderbased attraction is not a mere preference is "the need to recognize the community of persons who are exclusively (or strongly) attracted to transgender individuals, or who are themselves transgender and seeking persons with these attractions" (18).However, granted that such individuals require recognition, it is unclear why the recognition should come through a revised conception of sexual orientation, and not, say, from a recognition of their exclusive preferences, as I have been arguing.Moreover, such strong or exclusive preferences are often part of someone's sexual identity -how one sexually identifies or is identified given one's sexual orientation and preferences 37 -as opposed to being part of their sexual orientation (note that Dembroff accepts the distinction between sexual identity and orientation [10]).For example, many gay men claim that they are "tops" or "bottoms."Many lesbians identify as butch or femme.Many straight and gay people identify as BDSMers.Such preferences, if strong enough, need not count as part of sexual orientation, but as part of sexual identity (which preferences are part of identity and which are not is an issue worth pursuing), and it is unclear why gender -especially gender expression -is special in this regard.
Before concluding this section, it is worthwhile to consider one additional argument for why gender might be a basis of sexual orientation, one not explored by Dembroff but related to their second purpose that a plausible conception of sexual orientation should be "consistent with relevant social scientific research, in particular, research concerning sex and gender."Here, one can argue that if some people claim that they have a particular sexual orientation, we must take their word for it.So if some people claim that they have a sexual orientation based on the orientees' gender, then we must accept it as a sexual orientation.Such an argument would be consistent with the relevant research about sex and gender in that it takes into account what people claim their orientations to be (though people's testimonies would need to be filtered through properly designed studies).The argument would also be a version of epistemic firstperson authority, but applied to sexual orientation instead of gender. 38According to this view, a person's sincere selfidentification about their gender takes epistemic precedence over what others claim that person's gender to be.Similarly, one's sincere selfidentification about their sexual orientation takes epistemic precedence over what others claim that person's sexual orientation to be.
This view, however, faces the obvious objection that people might be mistaken about their sexual orientations.In the past, and owing to the power of the closet, many gay people thought that they were straight (and many acted on their beliefs by having sex with the opposite sex, and marrying and raising families with them).But we would readily agree that they were mistaken, and that, in fact, they were gay.Moreover, there is much (understandable) confusion in the nonacademic (and even academic!)world about differences among sexual orientation, sexual preferences, and sexual identities, which makes it easy for people to claim various sexual orientations when their claims could be construed as referring to sexual preferences or identities. 39Thus, we cannot just take people's word for what their sexual orientations are.
I have argued in this section that we have no good reasons to revise our conception of sexual orientation to add gender expression as a basis.

IV. Why Sex Is the Basis of Sexual Orientation
The arguments in the previous section reject an ameliorative account of sexual orientation.But then what is the point of the concept of sexual orientation?What work does it do?Per the remarks in Section I, I take its primary function to be descriptively classificatory: to classify into types human beings' sexual attractions.It is a morally and politically neutral concept that sorts our sexual attractions into types, and on whose basis further moral and political work can be done.That is, once we have the descriptive concept in place, further moral and political work can be done on its basis.
My claim is that sex is a good criterion for classifying sexual attractions for various reasons.I will start with and focus on one evolutionary, and thus historical, reason.I then briefly offer a second, nonhistorical reason, which is that sexbased attraction explains people's attractions given three crucial human values: procreation, pleasure, and love.
As with all life on Earth, evolution by natural selection is the process by which the human species has adapted and persisted over time and by which the continuation of the human species has been made possible. 40Given this, sexual desire has evolved such that we find each other sexually attractive on the basis of sex.This has been, and continues to be, mostly true of the opposite sexes: male and female human beings -respectively, the spermproducing sex and the eggproducing sex -have evolved to find each other sexually desirable given the role such desire has played in enabling procreation.There are, then, obvious explanations for why heterosexual sexual desire has been selected for and is adaptive.This aspect of our evolved, biological nature aligns with that of other mammals.Without such an account, we are hardpressed to explain why the overwhelming majority of people's sexual desires are for the opposite sex.
Moreover, human beings did not (and do not) procreate as, say, fish do, with females releasing eggs fertilized by the sperm released by males.We procreate through physical contact, through copulation.This explains why the genitalia are especially sensitive and pleasurable compared to other body parts: individuals with more sensitive genitalia had more sex, so with time more sensitive genitalia were selected for.Procreation also explains the desire to copulate, as opposed to just masturbate (even when the orgasms of masturbation can be more intense than those of partnered sex), and for why the genitalia (and female breasts) themselves are found sexually attractive.Because we need to copulate to reproduce, the desire for physical contact with another has been selected for, and if the genitalia were, somehow, undesirable, such physical contact would have been harder to attain.
Before continuing, note three things.First, this account does not imply that people desire to have sex only when they desire to procreate; people desire to have sex for a variety of reasons. 41The account merely explains why sexbased sexual desires are prevalent.Second, it does not imply that exclusive heterosexuality is the only or best strategy as far as advantageous selection is concerned.Even among nonhuman species, there is samesex sexual activity, which has its own explanations for why it is advantageously selective.Third, the account is not normative or prescriptive -it does not dictate how things should be.It only provides a historical narrative of how sexual desire came to be mostly heterosexual (and why, incidentally, it makes sense for an account of sexual orientation to include the sex of both the person with the orientation and the sex of the people to whom one is oriented).
The (uncontroversial) point so far is that sexual desire for and attraction to the opposite sex, including the genitalia, has been selected for or is adaptive because of its role in continuing the species.The more controversial point is that homosexual sexual desire and any other nonheterosexual desires (I use homosexual desire as my main example) are derivative of the heterosexual one -again, with no moral implications.As derived from or an offshoot of heterosexuality, homosexuality can be either adaptive or adaptively neutral.There are some interesting evolutionary explanations for why homosexuality might be adaptive. 42One is the famous kinselection hypothesis, that homosexuals continue to be found among populations because they help raise their kin's offspring.Another explanation is that samesex desire has strengthened relationships among samesex individuals, such that "the social assistance of peers and higher status companions may have increased the likelihood of access to resources and may have provided allies to help ward off attacks from other conspecifics." 43uch theories are controversial among evolutionary biologists and others, and homosexuality might be, for all we know, adaptively neutral.If it is adaptively neutral, homosexual desire is likely an offshoot or derivative of heterosexual desire.This is because, as in heterosexual attraction, in homosexual attraction the sex of the people to whom homosexual people are oriented remains the basis and object of their sexual attraction -homosexual men are attracted to human males because they are males, and homosexual women are attracted to human females because they are females.Moreover, as in heterosexual desire, the sexual parts and those near them, especially the buttocks, are crucial objects of attraction: 44 gay men sexually enjoy interacting with the sexual parts of men, and lesbians sexually enjoy interacting with the sexual parts of women, just as heterosexuals sexually enjoy interacting with the sexual parts of the opposite sex.If homosexual desire were unrelated to heterosexual desire, its similarity to patterns of heterosexual attraction would be a striking fact about homosexual attraction.Why wouldn't the focal point of attraction for homosexual men be, say, men's forearms instead of their genitalia and buttocks?That homosexual people exhibit the same pattern of attraction as heterosexual people do is best explained by the hypothesis that homosexual desire is, somehow, an offshoot of heterosexual desire, even if it is adaptively neutral. 45It bears repeating that the adaptive neutrality of homosexuality is a claim with no moral implications, so that discrimination against homosexuals would not be justified on the basis that homosexuality is not adaptively selected for or that it is an offshoot of heterosexuality.Sex is a taxonomical, not a moral, norm.
If this section's reasoning is plausible, understanding sexual orientation as based on sex is supported by sex's role in our evolutionary history.It is not an accident that in heterosexual and homosexual desire the genitalia and other body parts are found attractive because of the orientee's sex.Sex, then, delineates the kind of object one is sexually oriented to, thus providing a nonarbitrary criterion for classifying our sexual attractions.This is true especially for the descriptively neutral concept of sexual orientation (as opposed to an ameliorative one) that I have been defending.
Sex, however, is a good reason for sorting sexual attraction not only because it is nonarbitrary but also because it essentially (albeit partially) explains why some people choose their partners to pursue three crucial values that humanity holds dear, values that seem to be not contingent on specific social configurations: procreation, sexual pleasure, and love. 46First, classifying sexual attraction by sex tells us whose attractions are of the kind that leads to procreation, and whose don't: heterosexuals pursue those sexual partners with whom they can procreate (when not prevented by, e.g., age, contraception, biological malfunctions).Of course, procreation can occur with partners whose is of the kind to which one is not sexually attracted, but it usually occurs with members of the kind of sex to which one is attracted.In this regard, the sexbased concept of sexual orientation explains that some people procreate with others because the latter are of a particular sex.
Second, classifying sexual attraction by sex tells us who is likely to seek sexual pleasure with whom: heterosexuals seek sexual pleasure with the opposite sex, homosexuals with the same sex, bisexuals with both, while others with intersex people.Of course, one can experience sexual pleasure with members of the kind of sex to which one is not attracted, but it usually occurs with members of the kind of sex to which one is attracted.In this regard, the sexbased concept of sexual orientation explains that some people experience sexual pleasure with others because the latter are of a particular sex.
Third, classifying sexual attraction by sex tells us who is likely to love whom given that sexual attraction is often (though not always) the precursor of love:47 heterosexuals, because they are attracted to the opposite sex, are likely to love them; homosexuals, because they are attracted to the same sex, are likely to love them; and so on. 48Of course, one can fall in love with members of the kind of sex to which one is not attracted (and one can be sexually attracted to people one does not love, even hates), but it usually occurs with members of the kind of sex to which one is attracted.In this regard, the sexbased concept of sexual orientation explains that some people fall in love with some others because the latter are of a particular sex.
Sex is not, of course, a sufficient criterion by which people choose their partners, but it is essential and seems to underlie and limit all other criteria, such as gender presentation and age. 49For example, a heterosexual man's attraction to older women is attraction to older women, not also to older men.That is, even if someone's sexual preference is for old people, that preference is circumscribed by the orientees' sex: people are usually attracted to old men or to old women, not to old people, period.The same is true for body size (and weight, etc.).Age, body size, and body weight are not bases of sexual attractions independent of sex and, like gender, seem to be preferences.
Thus, a sexbased conception of sexual orientation is supported by two crucial reasons: an evolutionary reason, and one that shows the importance of sexbased attraction to three crucial human values.
To conclude this section, recall Dembroff's first purposes that an account of sexual orientation should "ideally" serve: it should clarify "the criteria for ascribing sexual orientation, as well as how these criteria translate into a taxonomic schema of sexual orientation," and it should be "consistent with relevant social scientific research, in particular, research concerning sex and gender."I have so far argued that one criterion for ascribing sexual orientation to someone is the orientee's sex, and that this criterion classifies people as having different sexual orientations depending on the orientees' sex.Moreover, and given that any conception of sexual orientation will have to allow for sexual preferences, the sexbased conception of sexual orientation is "consistent with relevant social scientific research ... concerning sex and gender" because it takes sexbased attractions to be primary while classifying genderbased attractions as sexual preferences.

V. Reasons Against Gender as a Basis for Sexual Orientation
The arguments in Section II, Section III, and Section IV would not show much if there were also not good reasons against gender as a basis of sexual orientation.For otherwise one can plausibly claim that even though there are good reasons for sex as the basis of sexual orientation and that there are no good reasons for gender as a basis, we could be conceptually generous and admit gender as an additional basis.
There are, however, at least two reasons to be conceptually parsimonious.First, I have argued that there are good reasons for why sex is a basis of sexual orientation.But without a reason for adding gender and not other properties we face a slippery slope: Why stop at gender?Why not add more bases to sexual orientation, such as age, racial and ethnic belonging, skin color, facial features, body shape and size, and other properties that can also be bases, given that people definitely exhibit preferences along their lines?(Perhaps this is why Dembroff believes that sex and gender are ultimately arbitrary.)There needs to be a way to stop this slide because we could reach a point where every person has an almost unique sexual orientation consisting of the sum of their sexual preferences, which would defeat the taxonomic purpose of sexual orientation as a concept. 50By contrast, and as I have argued in the previous section, sex is not arbitrary.
The second reason, which I offer tentatively, is a response to the following claim: if gender is different from age, skin color, height, and so on, it would make sense to consider it, but not the others, a basis of sexual orientation.However, gender seems to also "behave" in the same way as the others do, in that it is circumscribed by one's sexual orientation.Just as one is not usually attracted to youth, period, but to young men or young women (or both if one is bisexual), one's genderbased attraction is also circumscribed by sex.For instance, if gender means gender presentation, people's genderbased attractions seem to lie within their sexual orientations: heterosexual men are not usually attracted to just anyone who gender presents as a woman, but to females who so gender present.Homosexual men are not usually attracted to just anyone who gender presents as a man, but to males.Many bisexuals also have their gender preferences limited by their sex preferences: whether they are attracted to a gender presenting man depends on whether that person is male. 51f we assume that gender means masculinity and femininity, the same seems also generally true.For example, heterosexual women's attraction to masculinity is as it is instantiated in males, not females; some gay men have types that are genderinflected, such as the attraction to "bears" (hypermasculine men) or to twinks (slim, youthful men, with little to no body hair); and some lesbians prefer "butch" women.If gynandromorphophilia is a sexual orientation, the gynandromorphs' femininity is attractive because of a penis along with femalelooking breasts and other physical traits. 52 offer this reason tentatively because it needs empirical research for full support.For instance, if gender refers to masculinity and femininity, there is no research that I know of that shows, for example, that heterosexual men are more attracted to masculine women than they are to feminine men.Or, if gender means gender identity, I know of no research that shows that some people are attracted to others simply on the basis of the latter's gender identity, regardless of their sex.The jury is still out.Nonetheless, from what we do know about such attractions (per the previous two paragraphs), sex seems to play a limiting role.
Thus, the need to avoid a slippery slope and the strong possibility that sex limits genderbased attractions are two reasons against adding gender as a basis of sexual orientation.

VI. Concluding Remarks
I have argued that attraction on the basis of sex makes sense from an evolutionary perspective; that it helps explain the basic human pursuits for procreation, sexual pleasure, and love; that it accommodates attractions on other bases, such as gender presentation and age; that it handles other purported sexual orientations; and that it does not prohibit our ability to politically and morally advocate for the recognition of various sexual preferences and identities.I have also argued that the reasons on offer for adding gender are not convincing unless we are willing to also admit other bases, such as age, race, ethnic belonging, and body size.
While I hope that my argument is convincing, I also hope that it generates additional debate about these issues.Future fruitful engagement would have to address whether there is a plausible distinction among orientations, preferences, and identities, and, if there is, which features of people go into which category; whether there are reasons other than those I rejected in this paper for why gender (and which referent of "gender") should be a basis of sexual orientation; whether these reasons admit other features (e.g., age) of people as additional bases of sexual orientation; and whether evolutionary accounts provide good, or even conclusive, reasons for sexual orientation classifications.Moreover, future engagement would also have to address why we need to classify people by sexual orientation (or even, more generally, by their sexual interests), and how many sexual orientations we would be willing to accept without undermining the taxonomic function of the concept. 53t the end of the day, there is no God or universegiven reason as to why we have to classify people by sexual orientation, or as to why these, but not those, features are the bases of sexual orientation.Classifying people by sexual orientation and agreeing on its bases are, ultimately, human decisions.But these decisions need to be backed by good arguments, not merely asserted, taken for granted, or blindly following cultural trends.This is, after all, what it means to do philosophy.