Sexual Agency and Sexual Wrongs: A Dilemma for Consent Theory

Introduction

Consent looms large in discussions of sexual ethics. This is hardly surprising: nonconsensual sexual contact is a serious wrong, so when consent has not been given, one should refrain from sex. So much, so obvious. But one might think that there are important central ideas in sexual ethics that do not, and perhaps even should not, feature consent. If so, the near-exclusive focus on consent in discourse about sexual ethics is a mistake.

In this paper, we contribute to a growing body of literature that suggests that this is so.¹ We are focused on a particular kind of sexual violation, involving predatory sexual encounters to which victims freely consent. Such encounters, we think, are serious sexual violations—and explaining those wrongs falls under the project of sexual ethics—but they are not plausibly wrong because they violate consent-centric norms. Approaches to sexual ethics that exclusively emphasize consent, therefore, cannot accurately explain these violations.

The problem is not merely theoretical. It’s not just that we misunderstand what is going on in these cases. The misunderstanding can also produce a compounding harm: the consent framework itself can harm survivors of sexual predation, because it does not allow them to both recognize themselves as full agents, and, at the same time, acknowledge the moral violation they have suffered. They face an unjust dilemma: give up on the idea that they were wronged, or endorse a kind of agential demotion.

We’ll focus on two kinds of cases: exploitative sexual relationships between minors and adults, and adult sexual relationships involving significant power imbalances, such as relationships between athletes and coaches, students and teachers, or employees and employers.

The paper will proceed as follows. We begin in section 1 by motivating and characterizing the consent-emphasizing framework we shall resist. In section 2 we introduce our central cases—one involving sex between an adult and an underage teenager, and another involving sex between a professor and his student. We’ll argue that these cases involve sexual wrongs that are not well-explained by a lack of valid consent. The sexual wrongs we posit may not be recognized by their victims, so our diagnosis has implications on questions of first-person authority about sexual violation; we discuss them explicitly in section 3. Sections 4 and 5 introduce our key notion of agential demotion—an important and underappreciated cost of explaining sexual violations of the sort we are considering in terms of consent. The harms of agential demotion have a lot to do with the significance of ascription of competence to sexual decision-making, which we discuss in sections 6 and 7. In section 8, we relate our project to other recent discussions in sexual ethics.

1. The Consent Framework

We begin with some clarifications about how we are using the term “consent.” In the philosophical literature, consent is standardly thought of as a way of giving permission to some conduct or activity one would otherwise have a right against. In general, you have the right not to have your body sliced open with a scalpel, for instance. But if you consent to a medical procedure, the doctor may perform that action without violating your rights. Likewise, you have the right not to have your body touched in sexual ways, but you can permit that action by consenting to sexual contact. Following standard conventions, we specify that consent is “valid” when it gives a genuine moral permission. (If you agree to something under sufficient duress, or with insufficient understanding of it, you do not give valid consent, and thus retain your right against it happening.) For terminological specificity, we use “consent” such that genuine consent must be valid. (“Invalid consent” is not genuine consent; compare “counterfeit money.”)

One virtue of the consent framework is that it both provides a plausible diagnosis of the kind of sexual violation involved in paradigmatic rapes, and extends it to other similar cases in which consent is clearly violated.² Someone who sexually penetrates another person by force against their obvious will inflicts a grievous harm, subjecting the victim to a kind of contact against which they have a strong claim-right; the consent framework extends this story to cases where the victim does not actively resist, but also does not consent—cases, for instance, in which someone is incapacitated, and so does not attempt to fight off their assailant, or cases in which they comply out of fear for what will happen if they do not. Consent theory explains why victims in these cases also suffer a serious violation.

Criminal codes typically encode sensitivity to this particular kind of harm, in which lack of consent plays a central explanatory role. Although we are ambivalent about the way in which criminality is used to enforce sexual ethics,³ we have no quarrel with the idea that consent does play an important normative role that ought to be recognized in law, and it may well be that its legibility in the legal system is an important virtue of the consent framework.⁴

Emphasizing the similarity between forcible rape and other cases of sexual contact without consent has improved people’s understanding of sexual ethics. However, we are less impressed by an additional tendency we often see, extending the treatment to a further category of sexual violations that do involve at least apparent consent to sex. If the consent framework is the only way one knows how to talk about sexual wronging, then sound moral sensibilities generate pressure to describe some such cases—some cases involving sex between adults and minor teenagers, for instance—as nonconsensual. While we respect some of these motivations, we shall argue that denying consent in these cases is both a theoretical and a moral mistake.

2. Our Central Cases: Sexual Violations with Consent

We shall frame much of our discussion around two cases. Both involve sexual wrongs; the consent framework invites us to explain those wrongs via lack of consent. As we explain in the subsequent sections, we think that this is a mistake, both because it misunderstands what consent is, and, consequently, because it asks the victims in these cases to demote themselves, in their own self-conception, from full agency.

Our cases are fictional thought experiments, but we hope our readers will recognize them as realistic. Indeed, they are closely based on specific real-life examples that we know well.

Teenager Lily, a 15-year-old girl, goes to a show and meets Kyle, a 22-year-old man. They experience mutual attraction, and soon develop a sexual relationship. Lily is enthusiastic about their relationship, including the sexual activity, which she often actively seeks out. For example, she proactively engineers occasions where her parents will be away long enough for her to invite Kyle over to have sex with her. Lily’s attraction to Kyle is explained in part by their relative power imbalance: he has a car and a driver’s license, for instance, and she likes that. Kyle is also much more sexually experienced then Lily, and she likes that too. Lily is proud to have attracted an adult, whom she sees as more mature, worldly, and powerful than herself or her peers. And she derives a particular pleasure from behaving in ways that excite him and satisfy his desires.

Following pretty mainstream liberal intuitions (though not ones without their dissenters⁵), we condemn such sexual relationships. In particular, we think that Kyle seriously wrongs Lily in a way deeply connected to the sexual component of their relationship. But as we discuss below, we do not think, as many are tempted to, that this is because their sexual contact isn’t consensual.

Here is our second case.

Student Taylor, a 20-year-old university student, enrols in a course taught by Gregory, a professor in his mid-40s. Impressed by his knowledge and his charm, she gradually develops an infatuation with him, which he notices and cultivates. He starts mixing in personal messages along with his academic feedback in emails and office hour conversations; as their correspondence becomes more intimate, he also starts including sexual innuendo. All of this is flattering and exciting to Taylor, who never imagined that her sophisticated professor would have any interest in someone like her. One day, after Gregory expresses frustration with his marriage and “accidentally” admits overtly that he is attracted to Taylor, she kisses him. They go on to have a secret affair that lasts a couple months.

Many would condemn Gregory’s treatment of Taylor. We count ourselves among them: Gregory inflicts serious harm on Taylor, and universities have good reason to prohibit relationships like this one. Indeed, Gregory, as described, acts with more deliberate guile and manipulation than Kyle did. But we don’t think that we should justify such prohibitions on the grounds that relationships like this could never be consensual. We think that there often is consent in such cases, but that a significant sexual violation occurs nevertheless.⁶

A characteristic feature of both our cases is that we have victims of sexual misconduct who seem to be enthusiastic participants in the sexual relationship. This generates some anxiety about how to characterize them. Liberal principles may motivate one to defer to the individuals — who are we, one might ask, to tell people like Lily and Taylor what relationships they can and cannot freely choose to enter into?

While we disagree with such libertarian verdicts about which relationships are permissible, we recognize and respect a key motivation behind them: a concern about undue paternalism, and an insistence on recognizing people’s sexual agency, including that of teenagers and students. There is, we recognize, an important painful history of reactionary sexual moralism—and sexual minorities were its particular victims: mainstream intuitions about what kind of sexual relationships could really be valuable have played central roles in the maintenance of homophobic norms, for instance. Our central argument against the consent framework will turn out to be an expression of these kinds of agency-affirming values.

Our project is in no way a defense of relationships like those in our central cases. We have no quarrel with with the mainstream contemporary consensus that they are sexual wrongs. But we do dispute the mainstream explanation for these wrongs. We turn more specifically to this explanation now.

3. The Consent-Theoretic Explanation

The story we resist is a simple one: cases like Teenager and Student involve wrongs because they involve sex without consent. Lily, a legal child, cannot consent to sex with an adult like Kyle. And because Gregory has a substantial power imbalance over his student, she cannot consent to sex with him.

We think that this consent-theoretic explanation plays quite strong roles in the public imagination, and that it also sometimes tempts theorists and policymakers. We rarely see explicit commitments to it in the academic literature, but given its role in public consciousness, we consider it worthy of interrogation and discussion. We shall argue that the consent-theoretic story is both false and harmful.

The consent story is especially widespread in cases like Teenager. Indeed, according to a popular belief, so-called “age of consent” laws enshrine into law the idea that children and teenagers younger than a certain threshold age cannot consent to sex. This is a mistake. Criminal codes do not typically have much to say about whether young people can or do give consent; rather, so-called “age of consent” laws delineate the circumstances under which someone’s consent is relevant to whether one commits a criminal offense by having sex with them.⁷ Here, for instance, is the pertinent section of the Canadian Criminal Code:

when an accused is charged with [one of several sex crimes] … in respect of a complainant under the age of 16 years, it is not a defence that the complainant consented to the activity that forms the subject-matter of the charge.⁸

In other words, sexual contact with a child younger than the age of 16 years is a crime, whether or not they consent. This is not at all the same as saying that such children cannot consent. Even many official documents, however, do express this idea as concerning who can and cannot consent to sexual activity. Canada’s Department of Justice website’s page on “Age of Consent to Sexual Activity” does so, even though the actual code, quoted above, does not.⁹ We offer the Canadian Criminal Code as just one example, but it is quite typical in this respect.¹⁰

One sees similar invocations of consent in discussion of cases like Student. Those who advocate bans on student–professor sexual relationships, for example, sometimes express their rationale for them in terms of the limited possibilities for consent when there are significant power imbalances. Coleman (1988, 120–21), for instance, argues on psychoanalytic grounds that student attraction towards professors is a result of “transference,” concluding that there is no consent:

If the reason is transference, the dependent person is damaged per se because she never actually consented to sex. By definition she is unaware of unconscious motivations and thus is damaged by a sexual relationship she never actually agreed to—a sexual relationship with the powerful person rather than with the idealized person she created. (Coleman, 1988, 123)

Such ideas remain live. In a 2018 interview, Canadian novelist Heather O’Neill discussed the sexual harassment involved in relationships between university students and instructors, affirming that there “is no consent in a power relationship like that.”¹¹ Critics of prohibitions on student–teacher relationships also tend to focus on the consent question, supposing that such prohibitions presuppose that consent is impossible.¹² If you do a Google search around keywords like “consent sex power imbalance,” you’ll find many confident assertions along similar lines, e.g.:

• “Unequal power dynamics, such as engaging in sexual activity with an employee or student, also mean that consent cannot be freely given.”¹³

• “Power imbalances can make it difficult to recognize if consent is freely given, and can even make consent impossible.”¹⁴

• “Pennsylvania also recognizes that power imbalances in certain relationships make consent impossible, regardless of age.”¹⁵

• “Strictly from a legal context, a power imbalance negates consent. If there is the potential for any form of coercive influence, consent cannot be given.”¹⁶

We hope it is clear at this stage that the consent-theoretic explanation of the wrongs in our central cases is widely accepted; within mainstream discourse, it is the default view.

We reject the consent-theoretic explanation, for a variety of reasons. First, we’re impressed by extant arguments in the literature against assimilating the problems with power imbalances to a lack of consent,¹⁷ and the thought that too-inflationary notions of consent threaten to undermine consent’s potential moral explanatory power.¹⁸

Moreover, although the quotations given above attest to a widespread tendency to treat power imbalances like those at play between Taylor and Gregory as inconsistent with sexual consent, there are also reasons to interpret this thought as not quite serious. After all, it is very widely held, including by all parties to this debate, that sex without consent is rape, and that rape is a serious criminal matter. But even among those who support bans on student–teacher relationships, one does not typically see calls for criminal sexual assault prosecution in cases like Gregory’s.¹⁹ This, we think, reflects a tacit awareness that the consent framework is not quite the right one for these cases.

There is, to be sure, a terminological question at play here: should we use “consent” in a stronger way that rules out its application in our central cases, or in a weaker way that is easier to achieve, but less morally important? We want to suggest that the best way to progress on this question is to attend more explicitly to questions about what we want the idea of consent to do. As we understand it, consent is especially important for illuminating agency, autonomy, and competence. As we argue at length below, we think that the consent-theoretic explanation has implications in these domains that are both false and harmful.

The cases we focus on in this paper cast one tension into particularly stark relief. There are often divergences between first- and third-person assessments of consent, concerning whether an agreement is a genuine expression of someone’s autonomy, and whether it thereby renders an encounter morally permissible. In the following section, we consider the epistemic features of this tension by discussing an anxiety concerning self-knowledge. Shouldn’t people themselves be best-placed to know when they have consented, when they have refused, and when they are being mistreated? If so, aren’t Lily and Taylor’s participation in their relationships excellent evidence that they are permissible? In short, should we accept first-person authority about sexual mistreatment against oneself? We turn now to a discussion of several precisifications of that question.

4. First-Person Authority

There are many domains in which first-person authority (FPA) ought to be respected, for both epistemic and ethical reasons. We have both epistemic and ethical reasons to treat someone as the final and decisive authority on their gender identity, for instance.²⁰

For our purposes, FPA is understood to capture an epistemic difference between first-personal and third-personal perspectives in a given domain. When someone has FPA, they are in a better position to know things in that domain, compared to others.²¹ In this section we argue (i) that there is no FPA for judgments to the effect that one is not being sexually mistreated, (ii) that there is FPA for judgments to the effect that one is being sexually mistreated, and (iii) that there is also FPA for judgments to the effect that one is consenting or withholding consent.

Part of our skepticism about FPA for judging that one isn’t being sexually mistreated derives from the observation that, in some cases of sexual predation, the predatory harm itself is partly epistemic. Victims of predatory grooming, for example—as well as many others—are often strategically placed in bad epistemic positions by their abusers. Lauren Leydon-Hardy (2021, 140) calls this phenomenon “epistemic infringement.” Paradigm examples of epistemic infringement include predatory grooming, gaslighting, and propaganda. Such cases, by design, are ones where there is often no FPA about the abuse. These are cases where a perceptive third party might much more easily recognize the harmful dynamics of the relationship.²²

Failures of FPA in the relevant cases are not limited to epistemic infringement. In Teenager and Student, we don’t need to assume that Lily and Taylor are misled, or that their access to evidence about their situation is constrained, or that their self-trust is eroded.²³ Nevertheless, it may not occur to them until much later, if ever, that they have been wronged.²⁴

Strictly speaking, FPA is best thought of as applying to judgment types, rather than subject-matters. There is no incoherence in the idea, for instance, that one might have excellent epistemic grounds to know that one is in condition F, when one is, but be extremely poorly situated to know that one isn’t in F, when one isn’t.²⁵ To take one obvious example illustrating this asymmetry, let F be the condition that one is alive. If someone believes they themself are alive, that is extremely strong evidence that they are indeed alive; one should trust their testimony about such matters. There is FPA for beliefs that one is alive. But there is no FPA for the belief that one is not alive. Someone’s expressing such a belief is quite poor grounds for accepting it; third-party judgments are much more reliable.²⁶

The judgment that one has been sexually violated is one that enjoys pretty strong FPA. (It is not infallible, in the way the judgment that one is alive presumably is. But it is still quite reliable, and typically more so than third-personal judgments.) But the judgment that one has not been sexually violated may enjoy no such privileged status. In some cases, specific epistemic factors can make it particularly hard to recognize sexual violations from the first-personal perspective. It can be difficult, for instance, to acknowledge harm to oneself, especially when, like many teenagers, one has a strong aspiration to adult habits and responsibilities. In cases where victims of sexual violation minimize, downplay, or dismiss their experiences, it can be morally important not to take those declarations as the final authority on the matter.

We do, however, accept FPA for judgments about whether one consents to sexual activity. Subjects are in a better position to know whether they have consented (or withheld consent) than are third parties.

Why think that subjects are in a better position to know whether or not they are consenting as compared to third parties? The exact explanation will depend on one’s favored theory of consent, but we think that on quite a wide variety of approaches to consent, there is good reason to accept FPA. If one thinks that consent is attitudinal—e.g. constituted by (or analogous to) beliefs, intentions, or desires —then it is prima facie plausible that subjects enjoy privileged but fallible first-person access to their own consenting attitudes in the same way that they enjoy privileged but fallible first-person access to their own beliefs, intentions, and desires. Beliefs, intentions, and desires can be held privately, and in such cases third parties may have no access (let alone equal access) to those attitudes.

Furthermore, if consent is an attitude, it is not likely to be one that we simply undergo (e.g., coming to believe “it is cold” in response to walking outside during winter). It is closer to a mental act, more akin to Korsgaard’s (2009, 39) description of rational mental states as “active states of normative commitment.” Sustained normative commitments are not likely to be states that can be better (or equally) apprehended by third parties, given their close connection to one’s own reasons for acting. (Parrott (2015) provides an account where FPA about attitudes is justified by agential authority rather than epistemic access.)

Some theorists reject attitudinal approaches to consent, but not for reasons that undermine our case for FPA. Tom Dougherty, for example, has argued that consent requires communication, on the grounds that identifying consent with potentially private attitudes “[fails] to appreciate consent’s role in shaping how we are accountable to each other” (2015, 246). Thus, they argue that it is important to include communication in our definition of consent in order to explain how, why, and when permissions have been granted or revoked.

The alternative, then, is to adopt a performative account of consent, according to which consent is not a privately held state, but a communicative act. On one way of developing this view, it is central to consent that this communicative act expresses a particular kind of attitude—a desire or an intention or something similar. On this view, we have every reason to accept FPA, for the same reasons as on the attitudinal view: the consenter is best-placed to know what attitudes they have. People are also generally better-placed than others to know which attitudes their communicative acts express.

On another way of developing the performative view, certain communicative acts constitute consent themselves, without relying centrally on the private attitudes they express. One might suspect that this approach makes FPA about consent less plausible. If consent just is the uttered “yes,” then it might seem that the first-person does not enjoy privileged access relative to a third party. Given the publicity of the speech act, all parties are on equal footing to know when one has consented. But the prospect of “unintentional consent”—the exact cases that look like counterexamples to FPA—is dubious. Indeed, it is precisely these reasons that typically motivate stronger attitudinal constraints on performative theories.²⁷

Furthermore, the social emphasis on explicit, ongoing, and enthusiastic verbal or non-verbal communication presupposes FPA about consent: there is something deeply morally important that the consenter knows, something that their partner cannot know unless they are told. Namely, we need to be told whether our partner consents to begin with, whether that consent is sustained, and when (if) that consent is revoked.

Even if one were convinced by communicative approaches to consent that the consent-receiver has a special epistemic status that may rival that of the consent-giver, then, while this might undermine the letter of FPA, it would not do so in a way that challenges its broader role in our paper. It is one thing to say that consent-receivers may also have privileged access to facts about whether someone consented; this does not even remotely suggest that third-party judgments about consent, from people who did not participate in the conversation, or even witness it, could be more reliable than first-personal ones. The dialectical opponent we’re addressing here is someone who says that society broadly knows better than Lily and Taylor whether or not they consented to their sexual relationships. Arguing that their sex partners might also have important insights into whether they consented, perhaps comparable to their own, will not help challenge their professed consent.

Given FPA, Lily and Taylor’s judgment that they consented is quite a good indication that they did in fact consent. Of course, their judgment isn’t infallible. So it is still appropriate to consider and respond to reasons to doubt it in these cases; we shall consider several arguments along these lines throughout this paper.

Given our endorsement of FPA about consent and sexual wronging, and against FPA for judgments that one isn’t being sexually mistreated, our cases have the following features: the subject does not, at the time, know they are experiencing a form of sexual predation, but they do, at the time, know they have consented. We think that this is sadly quite common; for many survivors of sexual predation, understanding the predation won’t come until far later, if at all, especially in cases like ours, where consent is arguably present. Furthermore, the consent-based framework puts such survivors beginning to recognize their experiences in a dilemma: if the wrongfulness of the encounter implies that it was nonconsensual, then survivors cannot be right that their experience was one of being seriously sexually wronged and be right that they consented.

5. Agential Demotion

On the view we oppose, cases such as Teenager and Student are bad because they are not genuinely, fully, consensual. (So Lily and Taylor were self-deceived to think that they were.) Moreover, more generally, on this kind of view, when there is sex between adults and minor teenagers, or when there is morally impermissible sex between two parties with a significant power imbalance (as between a teacher and a student), there is some temptation to claim that the survivor was not even capable of consenting.

Thus, on a strong version of the consent framework, there is no room for survivors to both (a) articulate the moral impermissibility of the sexual activity they participated in, and (at the same time) (b) take themselves to have been full agents with the capacity to consent. The consent framework forces the unhappy choice over which to give up. The harm of giving up (a) is perhaps the more obvious; but we think that it is also a very serious loss to deny oneself the status of an agent fully capable of consenting.

We call this denial “agential demotion”:

A agentially demotes B when: (i) B is an agent capable of consenting, and (ii) A takes B’s agency to be compromised in a way that renders B incapable of consenting.

Agential demotion can also occur in domains outside of sexual ethics. An adult child might agentially demote their parent when that parent consents to a risky medical procedure. Someone might agentially demote their spouse when their spouse consents to an ill-advised financial contract. And of course, parents might agentially demote their teenage child when they discover that their child has been sexually active. In all cases, agential demotion is a failure to appreciate someone in their full capacity as an agent.

Agential demotion can occur either interpersonally or intrapersonally. When it occurs interpersonally, it can motivate A to act against B’s expressed interests, place restrictions on B’s sphere of influence, or undermine B’s standing as an agent in the eyes of others.

Something particularly interesting happens in intrapersonal cases, where A is B. Here, the consent-based framework requires survivors of sexual wrongs to agentially demote themselves in order to preserve their understanding that their experience was one of violation. This is an epistemic loss—a movement from knowledge to false belief—that can also involve psychological distress and loss of epistemic confidence. In some cases, agential demotion might, akin to a kind of “self-fulfilling prophecy,” lead to a true loss of agency. Given the role of social position in agency, third-personal agential demotion can be damaging to agency. But, in the context of the present discussion, we are particularly interested in the potential for compounding harms deriving from individuals regarding themselves as incapable of consent.

Survivors of predatory sexual relationships who agentially demote themselves are put in a very bad position. To be wrong about one’s own competence to consent is to be wrong about a deep and morally important feature of oneself. This (perceived) error might invite a skeptical line of reasoning in which one’s mistake was not one-off, but systematic: an indication of one’s limitations as an epistemic agent rather than the perverse nature of one’s circumstances. The epistemic failing might further interact with other negative self-regarding judgments common to many survivors of sexual predation: feelings that they are “bad,” “broken,” or “flawed.”

On some theories, positive self-regarding attitudes are necessary components of autonomy itself (see Govier (1993), McLeod (2002, 103–131)). On these views, a survivor who comes to believe that their judgments regarding their own competency are unreliable might plausibly lack the self-trust necessary for full autonomy. But, even if one does not think that self-trust is a necessary condition for autonomy, one might (and we think should) still be concerned about the undermining effects that such self-doubt has on one’s conviction to pursue a wide swathe of activities that one finds meaningful. If you doubt your own competence, you might come to distrust your desires more generally, yield decision-making to others, and forego opportunities for (non-predatory) sexual experiences and relationships.

In arguing for an ethical framework that leaves conceptual space for survivors to consent freely (indeed to consent freely, enthusiastically, and in an ongoing way) to morally impermissible sexual activity, we have two aims. The first is to resist a popular account of cases like Teenager and Student. The second is to circumvent the harms of self-directed agential demotion for survivors.

One of the central ideas in Linda Alcoff’s (2018) Rape and Resistance is the moral importance of allowing survivors to exercise agency in conceptualizing their experiences. Agential demotion curtails one’s ability to determine for oneself crucial aspects of sexual experience: whether one consented, whether one’s desires were genuine, the moral significance of the interaction, etc. Coming to believe that one’s agency was compromised at a time when one took oneself to be an agent fully capable of consenting can also interfere with epistemic confidence, which, in turn, can make it more difficult to recognize one’s desires, let alone pursue them.

A key harm of sexual violation, Alcoff argues, consists in

its effects on our capacity to become effective agents involved in the making of our sexual selves. Effective agency requires a sphere for exploration and experimentation and the hermeneutical space to generate one’s own interpretations of one’s experiences and desires. (Alcoff, 2018, 143)

Although Alcoff is more focally concerned with the impact that sexual violence has on our agency (whereas our cases are moral violations we would describe as nonviolent) the secondary effects of agential demotion are harmful along the same dimensions.

The relevant cases of agential demotion are ones in which the survivor consented (and so was capable of consenting), but was forced in their own self-conception to demote themselves from full agency in order to recognize their experience as one of sexual violation. Below, we shall offer arguments to support the idea that Lily and Taylor were fully competent to consent, and provide further explication of the compounding harms of agential demotion.

6. Competence and the Harms of Agential Demotion

One way to retain consent theory, in light of our cases, would be to deny that Lily and Taylor’s sexual relationships were consensual, holding that genuine consent is inconsistent with the predatory nature of those relationships. Maybe Lily and Taylor were not even capable of consenting, owing to the particular dynamics of their inappropriate relationships.

Someone might ask: Must the consent theorist really say that Lily and Taylor were incapable of consenting? Might they instead make room for the consent-based story they want to tell by saying that Lily and Taylor did not in fact consent? We do not find this strategy very convincing. If one thinks that Lily and Taylor could have consented but in fact did not, what more was it that one thinks that they would have needed to have done in order to consent? We can think of no plausible answer that vindicates the consent theorists’ judgments here, especially in light of the fact that Lily and Taylor take themselves to have consented. What other reason could there be to deny that valid consent is present in these cases, than the idea that the victims in question weren’t capable of it? (We note also that the modally strong language—“could not consent,” etc.—is ubiquitous in consent discourse, including many of the examples we quoted in section 3.)

A natural strategy to adopt in arguing that consent is impossible would be to focus on whether Lily and Taylor are competent to token or withhold consent. As discussed in section 3, “Age of Consent” laws codify the circumstances wherein sexual activity is a crime, rather than the age at which a person is competent to consent. However, public reception of these laws trend toward interpretations on which “Age of Consent” tracks competency. In a recent comparative study of consent legislation in Europe, Guangxing Zhu and Suzan van der Aa (2017) write that the age of consent “is an important legal mark that symbolizes when young people are considered capable to… give their ‘free and full’ consent to sex” and that “[f]rom the perspective of the children, as long as they are under the age of consent, they are considered incompetent to give valid consent to sexual activities” (Zhu and van der Aa, 2017, 15, our emphasis). Insofar as “age of consent” is received as a proxy for competence, it would be natural to think that the reason why Lily did not token consent is because she was not competent to do so. For Taylor, who is a legal adult, one might reason along lines similar to the popular literature we discussed in section 3: regardless of her age, power imbalance renders consent impossible.

Competence, it is widely recognized, is necessary for morally transformative valid consent.²⁸ Severe intoxication, for instance, is inconsistent with genuine consent, because it renders the intoxicated individual incompetent with respect to the decision to consent. Many have argued that the same is true of certain intellectual disabilities, and, one might think, perhaps the same goes for being underage, and in other circumstances wherein there is a large power imbalance.

People who are not competent to consent to sex might nevertheless still be regarded as agents broadly construed, insofar as they still have the capacity to act. Acts come in many forms: one may express agency by signing a contract or giving a speech, but also via much smaller and less conspicuous acts. Prolonged looking or furrowing a brow can be expressions of agency, as well as walking, talking, and raising one’s arm. (A notion of agency that does not build in ableist intuitions must not rely too heavily on either (i) assumed motor skills, or (ii) complex hierarchies of belief–desire pairs, extended diachronic planning, or intense self-scrutiny.)

Our notion of agential demotion is focally concerned with competence. When someone is deemed competent to perform an action (drive a car, token consent, teach) they are expected to pass a certain threshold of understanding and ability. This threshold is often thought to involve diachronic planning and an ability to predict the consequences of one’s actions (Freedman, 1975), as well as “the capacity for understanding and communication, and the capacity for reasoning and deliberation” (Buchanan and Brock, 1989, 23). Agential demotion occurs when someone is wrongly treated or thought of as lacking these agential competences.

If someone genuinely lacks capacities necessary for agency—if one is genuinely incapable of valid consent—then it is no error to treat them as if they are incapable of consent. But when someone is capable of diachronic planning, predicting consequences, understanding, and communication, and is nevertheless agentially demoted, that demotion mismatches reality. It is also often harmful. If a third party agentially demotes someone, they fail to grant them the respect and autonomy to which they are entitled. Or, if genuine agents are brought—by self-doubt, gaslighting, or other means—to agentially demote themselves, then they will take themselves to lack key agential capacities, which may even have a downstream effect in their actually losing agency.²⁹

Literature in biomedical ethics grapples with the difficulties of providing a satisfying theory of competence and articulating ways to measure it. Constructing protocols for testing competence is difficult but necessary, as consent (in particular “informed consent”) plays a central role in determining the ethical permissibility of medical interventions. Although analogies between sexual consent and medical consent are imperfect, we think that the discussion of competence in biomedical ethics can also be illuminating in relation to sexual ethics.

A central question in the bioethics literature on competence concerns whether we should think of it as a general capacity, or instead focus on more fine-grained, decision-relative capacities. Either view, we think, will imply that agential demotion is often a substantial harm.

Benjamin Freedman (1975) exemplifies the generalized competence view, holding that “responsibility” (his term for “competence”) should not be assessed relative to any particular decision:

[T]he “responsibility” which we require is to be predicated not on the nature of the particular choice, but on the nature of the patient/subject. What we need to know is whether he is a responsible man (“in general,” so to speak). (Freedman, 1975, 35)

On this view, denying that someone is competent to token consent implies that they lack the capacity for responsibility tout court. This is obviously an extreme verdict. By contrast, acknowledging that someone is generally competent to token consent prevents one from denying or disregarding someone’s true tokening of consent simply because one disagrees with the decision. If we deem someone “responsible” (i.e. competent), we are obligated to honor their tokening of consent. The only alternative is to deny them responsibility altogether. The idea that teenagers and college students are generally incompetent in this way is obviously a non-starter. But might there be a more circumscribed, domain-relative version of this strategy that rules out consent in our central cases?

7. Incompetence Within a Domain

Buchanan and Brock’s Deciding for Others gives an example of a domain-relative treatment of competence. “A competence determination,” they write, “is a determination of a particular person’s capacity to perform a particular decision-making task at a particular time and under specified conditions” (Buchanan and Brock, 1989, 18). Competence determinations vary relative to multiple axes: the domain of the decision, the complexity of information, changes in the subject or environment, and the stakes of the decision. According to the examples they provide, making medical decisions and making decisions about entering into contracts will belong to different domains, as will driving a car and solving differential equations. But Buchanan and Brock’s decision-relativity goes further than that—it’s not merely that different competences apply to different domains. Even within a single domain—the choice of whether to pursue a particular surgical procedure, for instance—whether one counts as having (enough) competence to decide depends on which decision is being contemplated. Higher degrees of competence may be necessary, according to Buchanan and Brock, to be competent to consent to a course of treatment judged harmful, than would be required for a course of treatment judged helpful. This is assessed by others—such as clinicians—rather than those making the decisions.³⁰

So Buchanan and Brock suggest that competence is both domain relative—one can be competent in some areas, but not others, and perceived-prudence relative—one can be competent to make a decision perceived as good, but not to make one perceived as bad, even within a single domain. Let’s consider each form of relativity in turn, when applied to cases like Lily’s and Taylor’s.

Start with domain-relativity. Perhaps Lily is competent generally, but incompetent to decide whether to have sex. This has the advantage of accommodating the intuitive fact that she manifestly is competent to make many other decisions. And Taylor, a legal adult, is plausibly competent to do even more things than Lily is—she can vote, and decide whether to get a tattoo—but incompetent to decide whether to have sex with her professor.

The most obvious way to do this would be to demarcate the realm of sexual decision-making as one to which their competence does not extend. But while this is slightly better than saying they’re incompetent in general, there are still, in our view, serious problems related to agential demotion with this strategy.

Sexual agency—and competence to make decisions in one’s sex life —is itself important. On the view we’re considering, although Lily is in a good position to know that she is a competent agent in general, she must doubt whether she is competent to make decisions in sexual contexts. This reduces the scope of Lily’s purported incompetence, compared to an account on which she is regarded as generally lacking competence, but it gives rise to versions of the same harms: Lily (1) loses knowledge, (2) is put in a position to doubt herself, and (3) becomes vulnerable to the psychological distress that can arise from (1) and (2).

Moreover, it is manifestly false that people like Lily have no competence to make sexual decisions. It is widely recognized, for example, that teenagers like Lily can be competent and responsible enough to decide whether to have sex with their teenage peers.³¹ And, just as we pointed out above that teenagers can be competent enough to be responsible for wrongdoing, so too can they be responsible for sexual wrongdoing; a teenager who rapes someone, for instance, is rightly held responsible for that action. So it is not true in general that teenagers are insufficiently competent to make sexual decisions.

And what goes for Lily, goes even more obviously for Taylor, who is a legal adult.

We suspect that positing a general incompetence in the sexual domain may share a common error with an influential (1996) paper by Heidi Hurd. While Hurd’s project is largely orthogonal to our main point, identifying the parallel and diagnosing the error may help illuminate our discussion of consent and autonomy.

In her “second identity thesis,” Hurd posited a close connection between the conditions necessary for culpability and the conditions necessary for valid consent. On Hurd’s view, it is precisely those circumstances which would have excused someone from moral blame for wrongful sexual contact, which are necessary for morally transformative consent to sex. (If one commits sexual violence on pain of the threat of death, one is, perhaps, not culpable—and if one “consents” to sex under a similar threat, that agreement is not morally transformative.) Hurd’s idea and the one contemplated here both suggest that certain factors—being a teenager, say—prevent one from responsible decision-making in a domain.

This idea has surprising consequences. We agree with contemporary liberal orthodoxy that there is no consent to sex in cases of severe intoxication. This thought, plus Hurd’s identity thesis, implies that severe intoxication is also inconsistent with culpable sexual assault—an obviously unacceptable conclusion. It is neither an adequate moral nor legal defense, against a rape accusation, that one was so drunk at the time that they have no memory of the interaction.

Hurd recognizes the revisionary implications of her view. She agrees that intoxication should be no excuse for sexual assault, but argues on this basis that it should also not invalidate consent. To do so, she suggests, might allow rape victims to unfairly “externalize responsibility for what they voluntarily do.” “On pain of condescension,” Hurd writes, “we should be loath to suggest that the conditions of responsibility vary among actors, so that the drunken man who has sex with a woman he knows is not consenting is responsible for rape, while the drunken woman who invites sex is not sufficiently responsible to make such sex consensual” (Hurd, 1996, 141).

To contemporary readers, Hurd’s stance can feel quite callous. We ourselves are not at all attracted to it. But we discuss it here for two reasons. First, we perceive a striking parallel between Hurd’s treatment and the idea that teenagers don’t have the relevant competence to make decisions about sex. And that line, we think, does represent a central thread in the contemporary imagination. The problems with Hurd’s view vividly illustrate where the thought goes, when taken to its logical conclusion. The implication that teenagers can’t be culpable for sexual wrongs shows that this cannot be correct. Hurd’s view and the one we are imagining share a common problem: an abstraction from the particular asymmetric interactions involved in sex and consent.

Second, although Hurd’s stark statement of her view does feel outdated, we do think that remnants of the thought persist, both in the public imagination and in the law. The website for Stop It Now, a non-profit organization dedicated to fighting child sex abuse, for example, carries the following text: “Even if a child or underage teen gives permission or acts willingly, this never implies consent. A child is never accountable.”³² “Accountable” is a notable word choice; it is a word associated primarily with blame. While those writing for Stop It Now were obviously doing so with the best of intentions, and we agree that adults shouldn’t have sex with children, we do not think that invoking questions about whether to blame the child for sex is the way to make the point (even if, as here, a negative answer is given).

There are material effects of this way of thinking. In 2017, a Minnesota woman was denied entry to a bar, because she was heavily intoxicated. She was lured under false pretenses to a private residence, whereupon she blacked out. When she woke up, the stranger she’d come home with was penetrating her vagina with his penis. She made a rape complaint, on the grounds—absolutely standard in recent decades—that incapacitation via intoxication is inconsistent with valid consent, and at first secured a conviction. The conviction was overturned in 2020, however, on the grounds that Minnesota’s criminal statutes involving incapacitation required that the intoxicating substances be “administered to that person without the person’s agreement.” Because she had voluntarily consumed the alcohol and narcotics in question, she did not count as “incapacitated,” under the restrictive statute in the Minnesota Criminal Code at the time.³³

We are not sure what motivated that restriction in the Minnesota law, but it’s not hard to imagine that the kind of argument Hurd offered, emphasizing “personal responsibility,” may have been a part of it. When the restriction was brought to salience in this case it was widely recognized to be intolerably victim-blaming, but the statute had been on the books for many decades at that time, and, as we have seen, it was used in 2020 to overturn a conviction for what we can easily recognize as rape.

What all of this suggests, in our view, is a separation of three kinds of questions: the conditions required for culpability for harm, the conditions required for validity of consent, and the conditions required for sex to be permissible rather than violative. Hurd collapses at least the first two, and the competence strategy we’ve been considering seems to imply this too. Consent theorists and much public discourse collapse the latter two. And there seems to be some temptation at least in the public imagination to collapse all three. But in order to maintain plausible verdicts about sexual harms and culpability for them, while avoiding the error and harm of agential demotion, we must distinguish all three.

8. Narrower Incompetence

Perhaps you are convinced by our case for the view that Lily and Taylor are not incompetent with respect to sexual decisions tout court, but still wish to pursue the domain-relative incompetence strategy by restricting the category to a particular kind of sexual decision-making.

For example, in recognition of the fact that teenagers can give valid consent to sex with their peers, one might restrict the domain to decisions concerning sexual activity with individuals who are older than then. Or perhaps the size of the age-gap is relevant: perhaps Lily is not competent to make decisions within the domain of “sexual activity with individuals four or more years older than her.” Even then, one might need to be more specific about the kind of sexual activity she is engaging in. Perhaps she is competent to consent to a kiss, but not to penetrative intercourse, so that the scope of the domain of her incompetence is narrowed further: Lily is held, perhaps, to be incompetent to make decisions in the domain “substantial sexual contact with individuals four or more years older than her.”

We do not find this strategy promising. The notion of a domain, we fear, is being gerrymandered according to one’s judgments of permissibility. If that is right, then judgments about Lily’s competence are being reduced to the claim that Lily is only competent to make sexual decisions in the domain: “sexual activity that is ethically permissible” or perhaps “sexual activity that does not put Lily in harm’s way.”

We have several concerns about the resulting picture.

First, it sacrifices the intended explanatory role that “competence” was supposed to play: Lily’s lack of competence was supposed to explain her inability to token consent. The domain-based strategy we’re contemplating says that she’s incompetent because the action contemplated—sex with Kyle—is a bad one. The badness of any such sex, then, can’t be explained by her incompetence. Which means in turn that it cannot be explained by her lack of genuine consent. In short, this strategy gives the order of explanation in the wrong direction.

There is a deeper problem. Attributing incompetence to decision-making in these cases via incompetence throughout a domain—no matter how finely specified, even assuming it could be delineated in a principled way—just gives the wrong result with respect to the moral agency of people like Lily. Although as we have made clear, we do agree with mainstream liberal intuitions condemning Kyle’s sexual relationship with Lily, we think that it is a serious misdescription of the way he wrongs her to assimilate it to nonconsensual sex. Lily’s agreement to have sex with Kyle—and her enthusiastic, affirmative actions that included seeking it out—are morally relevant to the situation. To say otherwise would be to say that Kyle’s action is morally equivalent to the possible case in which Lily refuses his sexual advances and he sexually penetrates her anyway. It would be to say, in other words, that that she has no say in the matter of whether he has her permission to touch her in sexual ways—clearly the wrong result.³⁴ The intuitive difference between the two cases cannot be captured in terms of domain-relative competence.³⁵

This brings us to perceived-prudence-relevance. The competence denier, we now see, faces pressure to say that in our cases, the victim lacks competence to consent, but retains competence to refuse, even with respect to the same question. So maybe the best strategy is to relativize competence, not (merely) to domains, but rather to the particular contents of decisions being made.

This is built into the account given by Buchanan and Brock: competence is relative to a benefit/risk assessment. For example, when a decision is made, if the “net balance [is] substantially better than [it is] for possible alternatives” then the level of decision-making competence required for consent to be valid is “low/minimal” (Buchanan and Brock 1989, 53). And conversely, if the “net balance [is] substantially worse than for another alternative or alternatives” then the level of decision making competence required for consent to be valid is “high/maximal.” In medical settings, this means that someone must be highly competent in order to consent to something that is very bad for them, but need only be minimally competent to consent to something that is substantially better (or best) for them.

But one would be prima facie suspicious if a doctor’s assessment of competence was perfectly aligned with their judgments about what is good for the patient. If a patient was deemed competent only to make decisions that their doctor judged would bring no harm to them—if the patient is free to choose anything they like, so long as it is the exact thing that the doctor thinks they should do—invoking the notion of “competence” begins to look like a cover for paternalistic attitudes. Competence assessments are important in part because they empower subjects to decide for themselves—this might include making decisions that deviate from third-party assessments about what is good for them.

What are the prospects for this strategy in denying consent in our cases? There may be sexual fates so heinous, or harmful, or debilitating, that no one is competent to consent to them.³⁶ The question is whether Lily’s and Taylor’s relationships are bad for them to such a degree that it would require a higher level of competence to consent to it than then in fact had. Intuitions may vary, but we certainly do not think so.

There is historical precedent for disagreeing with us in this matter, but it is now widely recognized as illiberal. In an American criminal proceeding in the 1960s, for instance, a court convicted a man of aggravated assault in what we would now recognize as a likely consensual BDSM³⁷ scene. (The “victim” did not complain; the case was brought to law enforcement by third parties.) The court held that “it is a matter of common knowledge that a normal person in full possession of his mental faculties does not freely consent to the use, upon himself, of force likely to produce great bodily injury.”³⁸ Few contemporary liberals would describe this case as just.

We think that the sexual relationships between Kyle and Lily, and Gregory and Taylor, were bad for Lily and Taylor, but they were not bad for them because they didn’t consent; nor were the encounters so heinous that Lily and Taylor were incapable of consenting. The consent framework does not allow room for analyzing these cases or similar cases in such a way. This is no marginal failure: cases like Lily and Taylor’s are common.

Furthermore, we think that the people who are wronged in cases like this are harmed in the first instance by the sexual violation, and then harmed again via agential demotion. This moral harm, we think, also leads to political errors. Some opponents of institutional prohibitions on student–professor relationships assume that they depend for their justification on the assumption that students are incompetent to consent to sex with professors, and draw out worries about agency, in the service of arguing against such bans.³⁹ One of the central implications of our paper is that it can be perfectly appropriate to condemn and forbid a class of relationships, without assuming that they would have to be nonconsensual.⁴⁰

9. Objections and Replies

10. Sexual Wrongs and Victim Preferences

By arguing for the limits of consent-based approaches to sexual ethics, we put ourselves in good company. Notably, Ann J. Cahill and Elise Woodard have introduced categories of consensual but nevertheless morally bad sex: unjust sex, and bad sex, respectively. Our project is different from theirs in important ways. Although we agree with Cahill (2016) and Woodard (2022) that the bad cases they consider are consensual and harmful, they explain the harms by invoking subversion of a person’s will or redirection of their desires to serve others’ ends. Our cases resist such a diagnosis; so there are instances of impermissible consensual sex wherein the moral badness is not explained in Cahill and Woodard’s terms.

Ann J. Cahill helpfully distinguishes between acts of sexual violence, and sexual acts that should be located in an “ethical gray area”: acts that are consensual, but unjust. Examples of unjust sex include ones in which “sex was the least bad option,” when there was “significant pressure to have sex with a partner,” and when “acquiescing to their partner’s sexual requests or demands was the easiest and/or quickest way to achieving one of their own needs or desires (sleep, for example)” (Cahill, 2016, 751). In these cases, a person’s will is recruited for the wronging party’s interests, pleasure, and ends.

Elise Woodard’s category of “bad sex” falls into at least three (related) types. The first includes sex that occurs as a result of psychological pressure and calculated trade-offs (Woodard, 2022, 309). The second kind captures encounters in which one consents to sex as a result of social coercion. If one is consenting as a result of the kind of pressure that arises from social expectations (e.g., that the person who pays for dinner is “owed” sex), then it may be a case of bad sex. Such expectations might be internalized or structural. Finally, Woodard describes cases of “epistemically unsafe sex,” wherein one suspects that one’s consent was modally irrelevant. One might token consent to avoid finding out that one’s consent is actually irrelevant—that the other party would have proceeded regardless.

Although they do not deny the presence of consent in these cases, Woodard and Cahill both explain the badness of bad sex in terms of trade-offs and coercion. The circumstances surrounding the sexual encounter are hostile to the individual’s ability to pursue their desires free from undue influence, and diminishes their sense of agency.

We also focus on an agency-related harm: agential demotion. But that is a secondary harm, related to how one is forced to conceptualize first-order sexual harms, given consent theory. The sexual violations themselves do not depend on coercion or trade-offs. Lily and Taylor conceived of their sexual activity as a true expression of their desires, and they were not self-deceived. They did not consent to sex to avoid exhaustion or emotional distress, nor did they suspect that their consent was modally irrelevant.

Nor did Lily or Taylor have a focal experience of being pressured. Of course, the first-personal experience of pressure needs to be distinguished from the tacit ways in which one conforms to social expectations or scripts, either by cultivating one’s desires in particular ways, or in how one acts. Lily and Taylor’s experience may be well-described in the latter terms (tacit conformity), but not the former (social coercion manifested as the experience of pressure). This, we think, distinguishes their experiences from Cahill’s rubric of unjust sex or Woodard’s rubric of bad sex: there are cases in which someone ends up badly harmed as a partial result of pursuing desires they genuinely regard as their own.⁴⁶

Bad sex and unjust sex involve circumstances in which sex was the least bad option. A subject must consent to sex, or else navigate the emotional fallout of their refusal, discover that their consent was modally irrelevant, or experience the discomfort of perceived norm violation, etc. In these cases, the wronged parties see themselves as being worse off relative to the comparative baseline wherein the wronging party never proposed or suggested sex.

By comparison, in our cases Lily and Taylor do not merely think that sex is their least bad available option. Rather, they are excited by the prospect; they see themselves as better off relative to the comparative baseline wherein sex was never even suggested, and better off not because the relationship is a means to a further good (a job, say, or promotion). Nevertheless, Lily and Taylor are wronged.

Diagnosing the kinds of harm that can arise in relationships with significant power imbalances is local, specific, and context sensitive. We think that it is morally relevant that Lily and Taylor desired their relationships with Kyle and Gregory, but we do not endorse the inference that the presence of desire—and the exercise of will in the absence of coercion—automatically puts a relationship in ethical good standing. As Alcoff writes, “Police, prosecutors, friends, assholes, ask: did you want it? Did you enjoy it? Or they may never voice these words but wonder all the same” (Alcoff, 2018, 131). Putting survivors in a position where they must answer those questions negatively, or else have their claims to harm dismissed, is a moral mistake.

So what does explain the wrongs that Kyle and Gregory commit? We’re more confident that there are wrongs here—and that the lack of consent isn’t the right explanation for them—than we are in any particular positive story about where the wrongs do come from. Indeed, we suspect that a pluralistic approach to diagnosing these wrongs may ultimately be correct; we see no reason to expect that any story as simple as the consent framework would be fully adequate to the complex moral landscape of these kinds of cases. Consequently, a complete story here would, of necessity, be outside the scope of this paper. But we can gesture towards some considerations that we find promising.

One idea we won’t pursue is that Kyle and Gregory commit non-directed wrongs. They plausibly do so—they each contribute to harmful and patriarchal social patterns, and make the world worse, irrespective of their effects on their sexual partners. But this is no explanation for the harms we have in mind. After all, we wish to explain why Lily and Taylor might eventually properly come to feel that they themselves have been wronged; non-directed wrongs clearly cannot explain that.

One can wrong a person without violating their agency. In general, one can harm someone by compromising their interests, and when one does so contrary to an obligation, one wrongs them. Not everything with interests is an agent. You can harm a plant by depriving it of sunlight. It is less obvious whether one can have obligations to plants, but if one can, one could wrong a plant by compromising its interests without violating its agency.

People, including teenagers and students, are agents. When one harms a person while violating an obligation one has to them, one wrongs them—whether or not one is acting against their preferences. Sexual relationships with adults are typically harmful to teenagers. We do not assume that there is a unified theoretical explanation for why this is so; we think that a more “particularist” treatment is quite plausible. Whatever their theoretical explanation, such harms are widely recognized, which is why our society includes a social order that obliges adults not to inflict such harms upon minors. And so likewise with relationships between students and professors. The obligation not to cause those harms is why these relationships constitute wrongs.

Here, we offer five additional tentative thoughts about the harms in our central cases; each would need to be worked out in much further detail to be fully convincing, but we include them as suggestive ways to extend our project in a more positive direction.

First, sexual relationships can involve, to greater or lesser extent, a pretense of egalitarianism. A student or teenager might feel empowered by their new standing as a sexual partner. This can make their relative social, educational, and economic disadvantages harder to recognize or acknowledge.

Second, sexual relationships are typically personally and emotionally intimate. There are exceptions, but our central cases certainly match this generalization. Sex involves deep vulnerability. Entering into a sexual relationship of this kind thereby places one in a mutually recognized position of trust. It is extremely plausible that this kind of relationship itself creates special duties of care that extend beyond the duties one holds to arbitrary members of one’s society, or students in one’s classroom. Gregory owes more consideration and care to the people he is intimate with than he owes to others. This may help to explain why, given their sexual relationship, Gregory owes it to Taylor—in a way he does not owe it to his other students—to conduct his personal life in a way that prioritizes her emotional well-being.⁴⁷

Third, it is plausible that there are special duties of care toward children in one’s society, or one’s students. Unlike duties of care that are generated by being in an intimate relationship (as described above) these are duties that adults and teachers owe to children and students because of their particular social role. This may help to distinguish the cases we focus on from other kinds of examples of sexually exploitative predation, such as that practiced by so-called “pick-up artists.” The “pick-up artist” who wrongs another patron at a bar by attempting to manipulate them does not plausibly do so because they are violating a role-specific duty of care.

Fourth, age-based criminal restrictions on sexual interactions and policies prohibiting student–professor relationships are relatively straightforward to implement, and carry less collateral damage, than would any policy we can think of that would ban things like pick-up artists. So there may be good practical reasons to focus on cases like ours, even if there is a broader pattern of similarly objectionable conduct.

Finally, while we are open to the idea that there are some examples of sexual relationships matching the patterns of our central cases where there is no serious harm done, we do think that there is good reason to think that harm is more the norm.⁴⁸ The testimony of many victims of such relationships is overwhelming evidence that at least often, they are harmful, even if they are consensual and freely chosen. In retrospect, for instance, it is obvious to many survivors of such experiences that they would have been better off if the relationship never been proposed. In other cases, survivors might justify their past decisions by continuing the pattern of relationship formation that they would otherwise discontinue. These outcomes are sufficiently common to render them foreseeable. One plausibly wrongs someone by engineering circumstances you have excellent reason to believe they will correctly later regret, whether or not the harm actually eventuates.

11. Conclusion

The consent framework for sexual ethics is helpful for moral understanding in many cases. It captures the continuity between forcible rape and other acts of sexual violence wherein an agent’s right against bodily contact is transgressed. It does not, however, capture many everyday cases in which an agent is harmed or wronged in the sexual dimension of their life despite their consent.

Our cases feature sexual agents who consented to sexual encounters that were nevertheless predatory and morally impermissible. We argue that the consent-based framework puts them in a dilemma: either give up on the idea that one’s experience was one in which one was seriously harmed or wronged, or else agentially demote themselves. The former distorts moral understanding and enables abuse, and the latter leads to the harms of agential demotion, including loss of epistemic confidence, psychological distress, and in some cases, the genuine loss of agency.

The way out of the dilemma is to weaken the connection between consent and sexual wronging. This is the way to understand the kinds of sexual wronging we focus on without agential demotion. It is both theoretically and morally important to explain these sexual violations, and to do so without asking their victims to deny their own agency.⁴⁹