I. CONSENT, AUTONOMY, AND THE GATE OPENER MODEL

The idea that the power of consent promotes and protects personal autonomy is widely endorsed and generally taken as a starting point for further theorizing rather than as a claim in need of any detailed defense.Footnote ⁸ Those who seek to explain the connection between autonomy and consent in more detail generally emphasize the important positive and facilitative role that consent plays, in contrast with the negative role that is played by autonomy rights.Footnote ⁹ They start from the plausible assumption that individuals should, for example, have control over their own bodies, such that they can exclude others from touching or interfering with their bodies against their will. Given the importance of being able to exclude others from these spheres, they are protected by autonomy rights, which place others under duties not to interfere with or use their bodies. However, by excluding others, these rights also prevent us from engaging in any of the wide array of important and valuable forms of interaction that involve bodily contact: a checkup at the dentist, a hug, a game of football, and so on. This is where the power of consent appears to play a crucial role. Seana Shiffrin puts the idea in the following way:

On this view, the importance of personal autonomy justifies rights that serve as barriers to interaction. Yet given the importance of all the forms of interaction that are ruled out by these rights, the power of consent must also be regarded as “part and parcel [of] any plausible conception of an autonomous agent.”Footnote ¹¹ Thus, the power of consent provides a form of positive control over our interactions with others that might be viewed as the necessary flip side to the negative control established by our autonomy rights.

Call those areas over which an individual should have the ability to exclude and include other people at will their legitimate domain. According to what I will call the Gate Opener Model of consent, the power of consent's central function is to serve the power-holder's interests in having control over whether other people can enter their legitimate domain.Footnote ¹²

This line of thought is intuitive. So far, however, the Gate Opener Model is ambiguous between two ways in which the combination of our negative rights and power of consent might foster personal autonomy. According to the first, our power of consent protects autonomy by giving individuals non-normative control over what happens to them. On this view, my negative rights and my power of consent give me control over whether you can act in a certain way. More precisely, they make it true for some action ϕ that you will not ϕ unless I give you consent.Footnote ¹³ For instance, they might make it the case that you will not use my car or will not give me an injection unless I give you consent. Here, your recognition of and guidance by my rights and powers is crucial because the normative entities that are my rights and powers are causally inert, and so cannot prevent or allow you to do anything. However, the value of this recognition is essentially instrumental. It is because of this recognition that I have control over how you interact with me. Call this view of autonomy's relation to consent the causal account.

A second way in which our power of consent may be thought to contribute to personal autonomy is by equipping individuals with normative control.Footnote ¹⁴ Here, what matters is not that my power of consent gives me control over what you will do, but rather, that it gives me control over whether you will violate a directed duty that you owe to me by acting. We need, as Shiffrin suggests, the ability to make various forms of interaction morally possible.Footnote ¹⁵ On this account, the underlying value of autonomy is characterized in such a way as to include the value of having control over the normative shape of my life. Call this account of the relationship between autonomy and consent the normative account.Footnote ¹⁶

It is not always clear whether those who appeal to the importance of autonomy to underwrite the power of consent are appealing to the causal or the normative account, or to some combination of the two. In any case, I will argue that even a combination of these two considerations does not exhaust the valuable functions of consent.

Begin by considering the limitations of the causal account. On this view, the fact that other people recognize my power of consent has significant instrumental value, insofar as it ensures that others do not enter the spheres protected by my rights against my will. To be sure, this will often be of great value to me. If other people will only touch me or use my property if I have given them my consent, then I have an important form of control over what happens to me. Yet having this kind of non-normative control is not the only thing at stake when others recognize and regulate their behavior in accordance with my consent. To see this, consider the situation of married women in the (not so distant) past when it was widely believed that the permissibility of a husband having sex with his wife did not depend on his acquiring her consent. According to the causal account, the reason this mattered was that, to the extent that women were not regarded as having normative power over moral claim-rights to sexual autonomy, women lacked actual control over whether they had sex or not.

Now this lack of non-normative control was clearly of great importance. However, appealing to this lack of control alone does not provide a fully illuminating description of the problem. To see why imagine a world in which a woman (call her Mary) acquires a practical power. Unlike a moral normative power, which gives an agent direct control over some part of the domain of moral reasons and requirements, a practical power gives an agent the ability to control or limit the actions that others can perform. In this case, imagine that Mary's practical power perfectly aligns with her justified rights and power of consent: she can, through an act of will, prevent others from acting in ways that violate her autonomy rights. For instance, when Mary's husband (Ted) attempts to have sex with her, she can prevent him by establishing a protective barrier between them.

When Mary has this practical power, she has both negative and positive control over her sexual interactions with others. Not only can she prevent others from making unwanted contact with her, but she can also willingly engage in sexual relations if she wishes to do so. Nevertheless, we intuitively think there is a serious problem in Mary and Ted's relationship if Mary must rely on this practical power. Surely Mary should not need to make use of her practical power in managing her sexual interactions with Ted. Yet on the Gate Opener Model, Ted's failure to recognize Mary's power of consent does not register as a matter of concern. Insofar as she has a practical power, she has control over whether Ted can make sexual contact, and this perfectly replicates the function of the moral power of consent.

This point generalizes. If the normative significance of consent is reduced solely to the fact that recognition of this normative power gives us non-normative control over how others act in the spheres protected by our autonomy rights, then nothing would be lost if they did not recognize our possession of this normative power so long as we had alternative mechanisms through which to ensure their compliance with our will. Moreover, none of this necessarily depends on fanciful philosophical examples involving practical powers and the like. While inconvenient, we could establish the relevant kind of non-normative power through elaborate personal security systems, weapons, and biochemical enhancements. These might similarly enable individuals to prevent others from interfering with their person or property against their will. Yet if we needed to rely on such methods when interacting with others, it would tell us that our relations with those others are deficient in important respects.

In response, someone might suggest that the example of Mary is problematic from the point of view of autonomy because, in addition to the value of having actual control over what happens to us, we also value having normative control. That is, we value the ability to control whether it is permissible for others to, for example, make sexual contact with us. This is at least in part because we value the ability to shape our relationships with others, and this includes having control over whether our interactions with them will wrong us.

Yet we can incorporate this concern without addressing the underlying problem. We can assume that Mary has both a practical power and a normative power. At the same time as controlling whether Ted is physically able to make sexual contact with her, we can assume that Mary also intentionally controls whether he will wrong her by making such contact. It does not follow that Ted must recognize Mary's normative power of consent. In such a case, Mary can control whether Ted can make sexual contact with her and can also control whether he will wrong her by doing so while Ted fails to recognize and be practically guided by Mary's moral power of consent. Insofar as this is the case, Mary's relationship with Ted suffers from the same intuitive defect highlighted above.Footnote ¹⁷

The upshot of the foregoing is that the valuable functions of A's power of consent cannot be reduced to (i) serving A's interests in having non-normative control over whether B can enter A's legitimate domain, (ii) serving A's interests in having normative control over whether it is permissible for B to enter A's legitimate domain, or (iii) a combination of (i) and (ii). It also matters to us that those with whom we interact recognize and are practically guided by our normative power of consent. Call this the interpersonal recognition of our power of consent. A full account of the valuable functions of consent should explain the importance of this interpersonal recognition. Appealing solely to our autonomy interests fails to provide such an explanation. Thus, we should consider what further role the power of consent may play within our interactions and relationships.

II. VALUABLE RELATIONSHIPS AND SHARED UNDERSTANDING

We have seen that appealing solely to our autonomy interests fails to explain the importance of the interpersonal recognition of our power of consent. To make progress, I think that we need to move away from a focus on the ways in which the power of consent serves the interests of the power-holder alone, and consider how consent affects the nature and value of interactions between individuals. To foreshadow the central idea, I will argue in Section III that a shared understanding between A and B that B treats A's power of consent as a deliberative constraint on acting within A's legitimate domain is constitutive of a non-instrumentally valuable form of relationship between them. To situate and support this claim, the present section argues for the more general claim that, in many contexts, a shared understanding between individuals that they are appropriately responsive to the reasons and requirements that are operative in that context is partly constitutive of a valuable form of relationship.

My account begins with the observation that the nature and value of many of our interactions and relationships does not only depend on the way in which individuals’ actions serve or set back other's interests. It also depends on how agents deliberate about how to act, that is, on the way in which individuals take one another's interests, needs, and choices into account when engaged in practical reasoning. One context in which this is particularly clear is within special relationships. For example, we believe that a friendship between Abbey and Ben is only a true or valuable instance of friendship if Abbey and Ben are disposed to deliberate about matters that fall within the parameters of their relationship in a certain fashion. Abbey and Ben ought, for example, to recognize and be guided by relationship-based reasons to provide emotional support, to keep one another's confidences, to help in the pursuit of one another's projects, and so on. If Abbey fails to recognize and respond appropriately to these reasons then she fails to relate to Ben as a friend, and this fact undermines the value of her friendship with Ben.

Call this the deliberative constraint on the value of relationships.Footnote ¹⁸ In contexts where deliberative constraints obtain, the value of an interaction or relationship is partly determined by whether the individuals involved recognize and are practically guided by the reasons that apply to them. While perhaps more familiar in close relationships such as friendship, deliberative constraints of a similar kind plausibly hold in a wide range of contexts. As P. F. Strawson famously argued, we generally attach a great deal of importance to other people's intentions and attitudes toward us. We are, as Strawson puts it, deeply concerned with whether other people's actions “reflect attitudes towards us of goodwill, affection, or esteem on the one hand or contempt, indifference, or malevolence on the other.”Footnote ¹⁹ For instance, I will feel a different kind of resentment toward someone who deliberately treads on my hand to injure me, compared with someone who treads on my hand accidentally. Similarly, we will seriously resent someone who drives down a busy street while drunk even if they do not cause anyone any harm because this action reflects a failure to give sufficient weight to other people's significant interests.Footnote ²⁰

The evaluative significance of our practical deliberations plausibly flows from our capacity to recognize and respond to reasons. The fact that we can recognize and respond to the reasons given to us by one another's interests means that the way we deliberate about those reasons affects the meaning and value of our interactions and relationships.Footnote ²¹ Of course, what constitutes giving one another's interests an appropriate role within our deliberations will vary significantly, depending on both the nature of the relationship in which we stand and the specific context in which we are acting. But the general idea is that, insofar as it sensibly matters to me whether other people relate to me appropriately, the value of my interaction or relationship with them partly depends on their abiding by certain deliberative constraints.

As Strawson further observed, our concern for how other people relate to us is not merely an abstract concern but plays an important role within our concrete interactions and relationships with others. For Strawson, our perceptions and beliefs about the attitudes and intentions of other people underwrite the reactive attitudes. The reactive attitudes are emotional responses such as resentment, betrayal, and gratitude, which we feel when we judge that other people's attitudes and intentions either do or do not reflect an adequate degree of “good will” toward us. To experience such responses is, Strawson wants to remind us, “what it is actually like to be involved in ordinary inter-personal relationships, ranging from the most intimate to the most casual.”Footnote ²²

I now want to argue that there is a further way in which our perceptions and beliefs about how we relate to one another affect our interpersonal relations. The basic idea is this. The value of many forms of interpersonal relationship does not only depend on whether the participants in fact deliberate about one another's interests appropriately. It also depends on whether the participants have a shared understanding that they each meet these deliberative constraints.

Here again, the example of friendship is illustrative. As I have said, the value of a friendship between Abbey and Ben will be partly determined by whether they deliberate in accordance with the normative requirements of friendship. I am now further claiming that the value of Abbey and Ben's friendship is partly determined by whether Abbey and Ben have a shared understanding that they each meet the deliberative constraints of friendship. The idea is that part of what it is to have a good friendship with someone is to have a shared understanding that both parties are (at least in general) appropriately responsive to the relationship-based reasons that friendship provides. To help see this, compare a “friendship” between Abbey and Ben in which, while they both in fact consistently act in accordance with the requirements of friendship, they are also regularly suspicious of the other's motives in acting. Such a relationship does not look like a flourishing example of friendship. By contrast when Abbey and Ben have a shared understanding that they both deliberate and act as friends should, then they achieve an especially valuable form of relationship.

What exactly is it for two agents to have a shared understanding of the relevant sort? At its heart, shared understanding is a type of common belief or common knowledge. In the central case, the value of shared understanding is realized between two agents, A and B, when (i) A and B both deliberate in accordance with the normative requirements that apply to them (given the specific context and relationship), (ii) A and B both believe that they each deliberate in accordance with the normative requirements that apply to them, and (iii) A and B both believe that the other person also has the belief specified in (ii). To be sure, there is significant space for misunderstandings or failures of different kinds, and this yields complexities in the precise contours of the values that are realized in any given relationship. Nevertheless, this general account is sufficient for my present purpose, which is to explain the sense in which shared understanding is sometimes partly constitutive of valuable interactions and relationships.

While it is intuitively important that friends and spouses have a shared understanding that they are each appropriately responsive to the normative standards that govern their relationship, some may doubt whether shared understanding is important outside of such contexts. However, while the value of shared understanding is plausibly of greater significance in special relationships, I think that the underlying source of the value has less to do with the importance or intimacy of specific kinds of relationships, and more to do with the fact that two individuals can recognize one another as reason-responsive agents.Footnote ²³ Indeed, a general concern for achieving a shared understanding is, I think, manifest in a wide range of the norms that govern social interaction. Consider, for example, the role of good manners and politeness in expressing respect and gratitude,Footnote ²⁴ the importance of practices of explanation and justification,Footnote ²⁵ and the role of blame, apology, and forgiveness in managing interactions and relationships when things have gone awry.Footnote ²⁶ While no doubt answering to several different values, I think the importance of such norms reflects, at least in part, the value we attribute to achieving a shared understanding in our interactions. In any case, what matters here is the extent to which the value of shared understanding is at issue in the context of consensual interactions, and it is to this issue that we now turn.

III. THE RELATIONAL MODEL OF CONSENT

According to the Gate Opener Model, the power of consent's central function is to equip A with control over whether B can enter her legitimate domain. Thus, according to the Gate Opener Model, my power of consent fulfills its main function if, for example, it gives me control over whether a lover can make sexual contact with me, or whether a friend can borrow a book. Yet as we saw in Section I, individuals could in principle have both non-normative and normative control over such interactions without other people recognizing or being practically guided by their moral power of consent. And as the example of Mary and Ted served to highlight, this failure to be practically guided by another's power of consent can seriously undermine the value of a relationship, even if the consent-giver retains complete control over whether others can enter her legitimate domain. So far, I have left the nature of this impairment unanalyzed, relying on the intuitive force of the example to motivate the need to consider alternatives. What the discussion in Section II brings into view is the idea that the means through which A achieves control over her legitimate domain in relation to B will make an important difference to the nature and value of A and B's interactions and relationship.

This brings us to the Relational Model of consent. At the heart of the Relational Model is the idea that A's power of consent does not merely give A control over how B can act, but rather enables a valuable mode of interaction between A and B over which A has control. Specifically, the Relational Model claims that, at least in some central cases, A and B realize a valuable form of relationship if they have a shared understanding that B treats A's consent as a deliberative constraint on entering A's legitimate domain. In outlining the Relational Model, I will begin by setting out the central features of the view before considering some examples and then addressing some possible objections.

To understand the Relational Model, begin with the observation that it is by treating A's consent as a deliberative constraint on entering A's legitimate domain that B accords the appropriate weight to A's control interests. A's significant control interests ensure that she should have control over whether B can make sexual contact with her, or borrow one of A's books, and this control should not depend on morally arbitrary features of A or her environment, such as whether she possesses practical powers or other means of non-normative control. Rather, A's significant control interests ensure that B has sufficient reasons to recognize A as having normative authority over her legitimate domain.

Where B does treat A's moral power of consent as a deliberative constraint, he meets one important condition of a valuable interaction with A. However, as I argued in the previous section, the value of many of our interactions and relationships does not only depend on whether individuals in fact deliberate about one another's interests appropriately. It also depends on whether they have a shared understanding that they meet these deliberative constraints. For instance, the value of friendship is partly constituted by A and B's shared understanding that they deliberate in accordance with the normative requirements of friendship. In the case of consent, this requires not only that B is practically guided by A's consent, but further that A and B maintain a shared understanding that B recognizes and is practically guided by A's consent. Thus, according to the Relational Model, the nature and value of A and B's interaction depends on both (i) whether B treats A's consent as a deliberative constraint on entering A's legitimate domain, and (ii) whether A and B have a shared understanding that B treats A's consent as a deliberative constraint on entering A's legitimate domain.

Insofar as this kind of shared understanding is partly constitutive of valuable interactions and relationships, we have interests in being able to realize this kind of shared understanding. We might call these further interests “relational interests.” Relational interests can be defined as interests that concern how we relate to one another. Specifically, relational interests are second-order interests that concern how we take one another's first-order interests into account when reasoning about what to do. In the present context, the Relational Model of consent claims that A and B have relational interests in being able to interact within A's legitimate domain, in what would normally be a rights-violating way, while (i) B is appropriately responsive to A's significant control interests, and (ii) maintaining a shared understanding that B is appropriately responsive to A's significant control interests. For instance, A and B have interests in having sex, or in B's borrowing a book from A, while maintaining a shared understanding that B gives the appropriate deliberative role to A's control interests. As I noted just above, B is appropriately responsive to A's control interests if he treats A's consent as a deliberative constraint. Thus, A and B have relational interests in being able to interact while maintaining a shared understanding that B treats A's consent as a deliberative constraint on action. So, while the Gate Opener Model focuses solely on A's interests in having control over whether B can act in certain ways, the Relational Model further appeals to A and B's mutual interests in being able to engage in a valuable form of interaction. All being well, B's recognition and responsiveness to A's consent does not only ensure that A has an adequate measure of control over her legitimate domain, but further enables A and B to interact in an especially valuable way.

To illustrate the central idea, consider the example of sex in more detail. There is no doubt that it would always be extremely valuable for individuals to have control over whether others can make sexual contact with them, however this control is achieved. Yet the positive value of our sexual relations does not only depend on our acquiring sexual gratification while maintaining de facto control over whether our sexual partners can make physical contact with us. Neither is it sufficient that we merely have control over the normative fact of whether it is permissible for someone to make sexual contact. Beyond this, it is partly constitutive of valuable sexual relations that sexual partners are appropriately sensitive to one another's significant interests in having control over whether they have sex. Thus, B's recognition of and guidance by A's power of sexual consent is a deliberative constraint on valuable sex between A and B, and vice versa. Furthermore, this deliberative constraint is something that sexual partners are aware of and care about. We do not merely regard it is an interesting fact about a sexual partner's psychology that they recognize and are practically guided by our consent. Moreover, we know that our partner, like us, will care about whether we are responsive to their consent. As such, we will not only want to know that our partner is guided by our consent, but also want our partner to know that we are guided by their consent. For these reasons, when A and B have a shared understanding that they are guided by one another's sexual consent they realize an especially valuable kind of interaction.

Alternatively, imagine that Carla wants to borrow a book from her friend Dan, a treasured gift from Dan's mother. If Dan were only concerned to have physical control over whether Carla could borrow the book, then these interests could be satisfied through various non-normative mechanisms (e.g., door locks, alarms, etc.). If Dan has a further interest in controlling the normative status of Carla's action, then this normative interest can be satisfied so long as Dan can waive (by an act of will, say) Carla's duty not to borrow the book. Notably, the satisfaction of neither of these control interests requires that Carla and Dan recognize interpersonal norms of consent that govern whether and when Carla can borrow Dan's book. However, even if Dan has complete control over whether Carla can borrow the book, it is still intuitively important that Carla seek Dan's consent. The Relational Model explains this by appealing to the value of shared understanding. Specifically, managing their interaction through interpersonally recognized norms of consent enables Carla to borrow the book while Carla and Dan maintain a shared understanding that Carla recognizes and that is practically guided by Dan's consent. And this kind of shared understanding is partly constitutive of a valuable relationship between them.

According to the Relational Model of consent, then, a central function of consent is to enable a non-instrumentally valuable form of interaction and relationship between individuals when they interact in close quarters. As we saw in Section II, the reasons and normative requirements that we recognize, and understand one another as recognizing, partly determine the nature and the value of the relationships that we stand in. For instance, I simply cannot be someone's friend unless I am generally disposed to respond to the relationship-based reasons that my friendship gives me. In the same way, the Relational Model of consent claims that we could not realize a valuable kind of relationship if we did not recognize one another as possessing a moral power of consent. That is because treating one another's consent as a deliberative constraint on action and having a shared understanding that we treat one another's consent as a deliberative constraint on action is often partly constitutive of the value of an interaction or relationship. This shows that, while it is in principle possible that both our non-normative and normative control interests could be protected without interpersonal recognition of the power of consent, we would be losing something of great significance by abandoning our practices of giving and receiving consent.

Note that the values of autonomy and control remain central to the Relational Model. After all, the object of A and B's shared understanding makes essential and direct reference to the power-holder's control interests. According to the Relational Model, the power of consent enables interactions between A and B in which they have a shared understanding that B is appropriately responsive to A's control interests. It is precisely because it is good for A to have control over her legitimate domain that it matters to A and B's relationship whether B recognizes and is practically guided by A's power of consent, and whether A and B have a shared understanding of this.

At the same time, the Relational Model offers a clear explanation of why the interpersonal recognition of consent matters in cases such as that of Mary and Ted introduced in Section I. Recall that Ted does not recognize that Mary has an authoritative power of consent over the permissibility of his having sex with her. However, Mary does possess a practical power, which enables her to control whether Ted can access her legitimate domain. Why, then, does it matter that Ted fails to recognize Mary's normative power of consent? According to the Relational Model, by failing to recognize and be practically guided by Mary's power of consent, Ted makes it impossible for them to realize a valuable kind of interaction and relationship. That is because it is partly constitutive of valuable sex and ongoing sexual relationships that sexual partners have a shared understanding that consent functions as a deliberative constraint on their making sexual contact.

Importantly, the scope of my claims about the Relational Model is limited. My central examples are of consensual interactions within what we might call “close personal relationships,” such as sexual relationships and friendships, and it is within these contexts that it is most obvious that the power of consent is valuable in part because it serves our relational interests. Indeed, it seems clear that the nature and value of a friendship or sexual relationship would be quite different if individuals did not treat one another's consent as a deliberative constraint or have a shared understanding of this fact. However, given the wide range of settings in which it might be valuable for an agent to possess a power of consent—for example, in arm's-length economic relations, parental consent and substitute decision-making, and consent given by collective agents and corporations—it is very plausible that the constellation of interests, values, and functions that underpin norms of consent will vary depending on the context. As such, while the Relational Model may be well suited to some contexts (e.g., friendships and sexual relationships) it may be less well suited to others (e.g., arm's-length economic relations).

Having said this, I do think that achieving a shared understanding is of value in a wide range of cases, at least within interactions between natural persons. As I suggested in Section II, while shared understanding is plausibly of special importance within ongoing and intimate personal relationships, the possibility of this value derives from the fact that two individuals are reason-responsive agents. And as Strawson and others have convincingly argued, we generally attach great significance to the attitudes and intentions of others in their interactions with us, quite independent of whether we already stand in a personal relationship with them. For instance, it will not only matter to me that I have control over whether the surgeon I have just met can operate on me, but also whether the surgeon recognizes and is practically guided by the fact that I should have control over whether they operate. For much the same reasons, the surgeon is likely to want to acknowledge that they are indeed guided by my legitimate power of consent. Thus, it does not seem implausible to think that even in less personal contexts of interaction our relational interests may contribute to the explanation and justification of appropriate norms of consent. However, fully defending this claim, and identifying the contexts in which our relational interests are significant, would take me beyond what I can hope to achieve in the current article.

To bring out the distinctive features of the Relational Model, it may be helpful to contrast the view with another proposal that emphasizes the importance of the consent-receiver in a consensual interaction. In a recent article, Renée Bolinger argues that one of consent's important functions is to manage and mitigate moral risks. To demonstrate the significance of this function, Bolinger imagines a “world without consent.”Footnote ²⁷ In this world, Bolinger stipulates, B's ϕ-ing will not violate A's rights only if A stably and autonomously desires that B ϕ. Now imagine that Arnold requests that Beth perform a surgery that, while not strictly necessary, would be likely to improve Arnold's life. Insofar as Beth is a conscientious agent who prioritizes refraining from violating others’ rights, she may be reluctant to go ahead with the surgery. That is because, whatever Arnold says, Beth lacks direct access to the internal states of Arnold (viz, Arnold's stable desires) that determine whether Beth will be violating Arnold's rights by performing the surgery. It might be that, despite Arnold's outward assurances to Beth, Arnold is mistaken about his all-things-considered desires (perhaps changing his mind about the surgery as soon it begins). Alternatively, Arnold could be deceiving Beth about his desires to benefit from her wrongdoing. For instance, if Beth does violate Arnold's rights she will owe him compensation, compensation that other members of the moral community can be called upon to enforce on Arnold's behalf. Given this situation and given that Beth does nothing wrong in refusing to assist Arnold, she may well choose to pass on such a high-stakes “moral gamble” and thus refuse Arnold's request. Yet of course, this leaves Arnold unable to benefit from the surgery.

As Bolinger points out, consent can positively alter this situation by allowing Arnold to “own” the risks that Beth faces. If certain forms of communication are publicly recognized as acts of consent, which grant the relevant permissions and thus prevent Beth from violating Arnold's rights, then Arnold can give consent to Beth in a way that undercuts the gamble she previously faced. However, Bolinger points out that to play this role in facilitating interaction, consent must have certain features. In particular, Beth must be able to track the facts that determine whether she has received a consent-based permission from Arnold. For this reason, Bolinger argues that conventionally recognized acts of consent will generally be sufficient to provide a consent-receiver with a consent-based permission, independent of the consent-giver's intentions or other attitudes.Footnote ²⁸

Bolinger's arguments seem to me to be broadly convincing. She persuasively argues that, without the power of consent to mitigate the moral risks involved in social interactions, we would be much less willing to interact in all sorts of valuable and beneficial ways. Nevertheless, Bolinger's account does not explain why we regard the interpersonal recognition of consent as significant. So long as Beth can be reasonably assured that performing the surgery will not violate Arnold's rights, the risks of wronging Arnold or owing him compensation for violating his rights are mitigated. Yet Beth could have this kind of assurance without treating Arnold's consent as a deliberative constraint or achieving a shared understanding with Arnold that this is the case. For instance, this kind of assurance could be derived from a permission granted by a third party. If a legitimate legal system recognized Cathy as having the normative power to permit medical interventions to Arnold, as it sometimes will for good reason if Arnold is not competent to make medical decisions, then Beth could be reasonably assured that she will not violate Arnold's rights and Arnold can benefit from Beth's surgical intervention.

To be sure, on the assumption that Arnold is a competent adult, Beth's reasonable assurance that she will not violate Arnold's rights can generally only be secured through her recognition of and guidance by Arnold's consent. Nevertheless, the example highlights that the considerations that motivate Bolinger's position are quite different from those at the heart of the Relational Model. The benefits central to Bolinger's account concern what we might call the outcomes of interaction, such as the benefits of surgery to Arnold's quality of life. For instance, Bolinger plausibly argues that if Beth faces a high degree of moral risk, it is much less likely that Beth will be willing to operate and thus less likely that Arnold will benefit from the surgery. By contrast, the Relational Model is motivated by the value of a certain kind of interaction. Indeed, I have been arguing that it does not only matter to us that we can enjoy the outcomes of interaction—such as sexual pleasure, or access to good reading material—while maintaining control over those interactions. We can add that it does not only matter to us that we can receive these benefits without, as individuals, taking moral risks. Beyond this, the value of some such interactions is partly constituted by our having a shared understanding that we give sufficient weight to one another's legitimate control interests, and thus, that we are practically guided by one another's power of consent.

None of this is supposed to represent an argument against Bolinger's central claim. Rather, my aim is to highlight the way in which Bolinger's account and the Relational Model diverge, and to bring out the distinctive features of the Relational Model. The aim of both accounts, I assume, is to draw attention to the power of consent's role in realizing values and protecting interests that are routinely neglected within the existing literature on consent, and to highlight the implications of these considerations. Neither I nor Bolinger argue or assume that the power of consent serves only one function, value, or set of interests.Footnote ²⁹ And there is no obvious reason not to be pluralists about the considerations that underpin the power of consent.Footnote ³⁰ Indeed, as I suggested earlier, given the wide range of contexts in which a normative power of consent is likely to be of value, it seems plausible that the power will be underwritten by different considerations in different contexts.

Before turning to consider some implications of the Relational Model (in Section IV) let me address some possible objections. To begin, it might be objected that the Relational Model misconstrues the value of consent, by endowing it with a much more significant role than it in fact plays. One way of stating this objection is as follows: A's consent is valuable because it can make B's ϕ-ing permissible. Insofar as we want to interact without wronging one another this is important. However, this is the full extent of the role that consent plays. While a variety of further considerations will be relevant to the value of our interactions and relationships, these considerations are not relevant to the value of consent.Footnote ³¹

For example, it might be objected that is implausible to think that the value of any sexual interaction is fully determined by whether sex is consensual. B's giving sexual consent to A ensures that B does not commit the wrong of rape. This is obviously important. But many further factors will be relevant to the value of a sexual encounter, such as whether there is an emotional connection between A and B, or whether A and B are adept at satisfying one another's sexual desires.

However, this way of stating the objection mischaracterizes the Relational Model. I am not claiming that a shared understanding between A and B that B is practically guided by A's consent will ensure that their interaction is valuable along all dimensions. For instance, I am not claiming that A and B's having a shared understanding that B is practically guided by A's sexual consent (and vice versa) ensures that A and B's sexual interaction is good, valuable, or meaningful in all possible respects. Rather, I am claiming that A and B's having this kind of shared understanding is, at least usually, one valuable feature of sex. The presence or absence of other factors, such as an emotional connection between A and B, will also make a difference to the nature and value of a particular sexual interaction.

Nevertheless, the objector may deny that A's consent plays any valuable role beyond ensuring that B does not wrong A. In assessing this suggestion, we need to tread carefully. Claiming that A's power of consent is only important because it can prevent B from wronging A severely underdescribes the situation. On the assumption that there are normative considerations that explain and justify why B will normally wrong A by ϕ-ing, and why A can give B a consent-based permission to ϕ, we want to know what these considerations are. Indeed, a central aim of this article is to try and get clear about what these considerations are. A proponent of the Gate Opener Model claims that the relevant considerations are A's (non-normative or normative) interests in having control over whether B can ϕ. I have argued that, in addition to A's control interests, A and B's relational interests are also relevant.

Thus, it is true that according to the Relational Model, consent plays a more significant role in our interactions than is often assumed. It not only ensures that A has control over how B can act within A's legitimate domain, but further enables A and B to realize a valuable mode of interaction. My argument for the latter claim was motivated by the observation that there is something problematic about cases in which B fails to recognize or be practically guided by A's power of consent, such as in the case of Mary and Ted. However, it might be objected that one can accept that there is something problematic about the relationship between Mary and Ted but deny that this problem is directly related to Mary's consent. For instance, it might be claimed that, while Mary can object to Ted's behavior insofar as he is thoughtless, inconsiderate, or disrespectful, these are distinct objections to that of their sex being non-consensual.Footnote ³² Thus, if Mary intends to allow Ted to make sexual contact with her (as she might in the example without Ted's being guided by her consent)Footnote ³³ then she cannot complain that sex with him was non-consensual, even if she can still lodge complaints about Ted's thoughtlessness, inconsiderateness, or lack of respect. As such, it might be claimed, there is no need to appeal to a thicker account of the value of consent to explain what is problematic about this relationship.Footnote ³⁴

At this stage, we have two competing explanations as to why the interpersonal recognition of consent is significant. How should we decide between these competing views? Indeed, at this point, it may be unclear what is at stake in the dispute. I think the most plausible way of understanding the substantive disagreement is in terms of whether relational interests are at least sometimes relevant to the explanation and justification of the norms that govern the power of consent and consensual interactions. A proponent of the Gate Opener Model will insist they are not relevant. They will say that while shared understanding may be a valuable feature of some interactions and relationships, only our control interests are relevant to the determination of the moral, social, or legal norms that govern consensual interactions.

Against this, let me highlight three reasons to think that relational interests are relevant to the explanation and justification of norms of consensual interaction. First, note that we are envisaging that the defender of the Gate Opener Model accepts that there is something problematic about the relationship between Mary and Ted in virtue of Ted's disregard for Mary's moral power of consent. They simply claim that whatever exactly the nature of that problem is—whether Ted is thoughtless or inconsiderate or disrespectful—it has nothing to do with the explanation or justification of the norms that govern consensual interaction. However, I assume that a proponent of the Gate Opener also accepts that whatever the problem is, it would be resolved by Ted's recognition of and guidance by Mary's moral power of consent. Furthermore, I assume that at least in the central cases, a defender of the Gate Opener Model will agree that the only way to resolve the problem is through Ted's guidance by Mary's moral power of consent. That is, Ted will avoid being thoughtless or inconsiderate or disrespectful only if he recognizes and is guided by Mary's power of consent.

Yet if it is true that Ted's recognition of and guidance by Mary's power of consent is necessary to avoid the problematic mode of interaction at issue, it seems difficult to claim that our interests (say, in avoiding disrespectful interactions) are not relevant to the determination of norms of consent. On the assumption that Mary and Ted have an interest in avoiding the problematic mode of interaction, and that this is achieved (at least in part) through Ted's responsiveness to Mary's consent, surely these interests will inform the norms that govern consensual interactions. But if that is the case, a defender of the Gate Opener Model will have ceded the crucial point, namely, that interests other than Mary's interests in having control over whether Ted can enter her legitimate domain are relevant to the explanation and justification of the norms that should govern consensual interactions between them.

Second, the claim that our relational interests are relevant to the explanation and justification of norms of consent aligns with a familiar thought about rights and wrongs. In particular, it aligns with the thought that an adequate account of rights and wrongs will reflect a fundamental concern for how we relate to one another.Footnote ³⁵ While this thought has been given a variety of articulations by authors who emphasize different aspects of our interpersonal interactions and relationships, the general idea that we are concerned with how we relate to one another is a familiar starting point in our moral and political theorizing. Thus, the idea that A and B's relational interests should factor into the explanation and justification of the norms that govern A's power to determine whether B will wrong her by ϕ-ing is supported by this familiar thought about rights and wrongs.

Finally, it is important to reiterate that the central claim I have sought to defend in this article is that our relational interests are relevant to the explanation and justification of norms of consent in some contexts. For instance, insofar as friends and sexual partners have relational interests in being able to interact in close quarters while maintaining a shared understanding between them that they are appropriately sensitive to one another's significant control interests, these relational interests help to explain and justify the norms of consensual interactions that obtain between friends and sexual partners. For this reason, it is not an objection to the Relational Model that there are some contexts, such as arm's-length economic relations, in which relational interests are not intuitively relevant to the explanation and justification of norms of consent. After all, as I suggested above, our norms of consent are plausibly context sensitive.

IV. THE RELATIONAL MODEL AND THE ONTOLOGY OF CONSENT

I have argued that a complete account of the values, interests, and functions that underpin the power of consent must incorporate our relational interests and the value of shared understanding. At least within close personal interactions and relationships, a shared understanding that we treat one another's consent as a deliberative constraint is partly constitutive of the value of those interactions and relationships. If this is correct, our relational interests should inform ongoing discussions about the ethics of consent. To bring this out, this final section briefly highlights some implications of the Relational Model for one debate, namely, the debate about the ontology of consent.

Broadly speaking, there are two camps in this debate: those who believe that acts of consent are mental states or mental acts,Footnote ³⁶ and those who believe that acts of consent require some form of communication.Footnote ³⁷ In arguing for the former view, several authors have explicitly appealed to the idea that, because the power of consent is grounded in considerations of personal autonomy, we must be capable of giving consent without engaging in any form of communication. For example, Kimberly Ferzan argues as follows:

Similarly, Heidi Hurd claims that:

Bolinger's recent account offers a forceful challenge to this line of argument.Footnote ⁴⁰ As discussed above, she argues that for A's consent to play its risk-managing function between A and B, B must be able to track the facts that determine whether A has given him a consent-based permission. Because A's mental states are not observable by B, they fail to pass this test. However, Bolinger proposes that an easy way to solve this problem of “normative opacity” is “to hold agents accountable to a conventional mapping of facts they can track (communicative behaviours) onto the facts of primary interest (moral permissions).”Footnote ⁴¹ As such, Bolinger argues that it must be sufficient for A to provide B with a consent-based permission that A engages in a conventionally recognized communicative act of consent.Footnote ⁴²

If I am correct that consent also serves the consent-giver and consent-receiver's relational interests, this opens another line of response that, like Bolinger, meets proponents of the mental state view on their own territory, by arguing from the interests and values that underpin consent's normative significance to the norms that govern the power of consent and consensual interactions. Specifically, if I am right that a shared understanding between A and B that B treats A's consent as a deliberative constraint is partly constitutive of the value of, for example, a sexual interaction, or a friendship, then the norms governing what constitutes an act of consent in these contexts will be informed by this value.

What does this imply? For one, it provides additional support for the claim that A's acts of consent must be observable by B. If B is to be practically guided by A's consent and treat A's consent as a deliberative constraint on entering A's legitimate domain, then B will need to be able to track the facts that determine whether A has given B consent. Thus, the problem of “normative opacity” that is a feature of mental state views does not only leave B facing an unreasonable degree of moral risk, as Bolinger points out, but also disrupts B's ability to meet a condition on a valuable form of interaction with A.Footnote ⁴³ Insofar as A and B both have an interest in B's being able to meet this condition, this supports the claim that the facts that determine whether A has given consent to B must be observable by B.

What is more, the Relational Model's concern for shared understanding between A and B provides additional support for an observability condition. Indeed, I suggest that the value of shared understanding motivates a publicity condition on acts of consent.Footnote ⁴⁴

The idea is that if A and B are to have a shared understanding of whether B treats A's consent as a deliberative constraint, it needs to be as transparent as possible between A and B whether A has given B consent. A full articulation and defense of this condition would require another paper. But to see the intuitive idea, consider the example of traffic lights. If Abbey and Ben can both clearly observe a set of traffic lights, and both know that they each have a clear view of the traffic lights, then it is public between them whether they face a red or a green light. This not only enables Abbey and Ben to be practically guided by the lights when deciding whether to stop or go, but also to interpret one another's actions in light of this public information. To map this onto the case of consent, we can imagine that Abbey has control over the traffic lights. If Abbey wants to allow Ben to enter her legitimate domain Abbey can turn the light green. Otherwise, she can leave it red. These publicly observable signals not only enable Ben to be guided by Abbey's consent, but also make it as transparent as possible between Abbey and Ben whether Ben treats Abbey's consent as a deliberative constraint.

Importantly, the publicity condition does not claim that β constitutes an act of consent only if A and B have a shared understanding that A β's, or that B receives a consent-based permission from A only if A and B have a shared understanding that A β's. Because shared understanding involves beliefs about one another's mental states this would reintroduce the problem of normative opacity.Footnote ⁴⁸ Rather, the publicity condition is motivated by the idea that A and B are more likely to be able to achieve a valuable form of shared understanding if consent is public between them. As with the goal of managing moral risk, the easiest way to meet the publicity condition will be through publicly recognized conventions of consent. However, a fuller discussion of such conventions will have to await another occasion.