1. The Puzzle of Third-Party Coercion1
Here is a paradigmatic case of consent-undermining coercion:
Coercion: Badguy reliably threatens Goodguy that unless Goodguy pays Badguy a sum of money, Badguy will beat up Goodguy. Goodguy agrees to pay, and pays.
Assuming that no other non-standard complications are in place (such as Badguy for some reason being entitled to the money, and so on), Goodguy’s consent is not valid. If Goodguy transfers the money to Badguy, Badguy does not acquire the relevant property rights in the money (as he would have had the consent been unproblematically valid). If Goodguy signed a document seemingly undertaking a commitment to pay, no such duty has been created (as it would have been had Goodguy’s consent been unproblematically valid).
Compare Coercion to the following two cases:
Bystander: Badguy reliably threatens Goodguy that unless Goodguy pays Bystander a sum of money, Badguy will beat up Goodguy. Goodguy agrees to pay Bystander, and pays.
Bodyguard:2 Badguy reliably threatens to beat up Goodguy. In order to defend himself, Goodguy hires Bodyguard, agreeing to pay him a sum of money for his services.
In both these cases, Goodguy’s consent to transfer the money is caused by Badguy’s wrongful threat, and at least in that sense, coerced. Still, I’m going to take it for granted that in Bodyguard — at least if details are filled in in a natural way — the consent is valid. Goodguy owes Bodyguard the money, and if he already paid, the money now fully and unproblematically belongs to Bodyguard. Bystander is perhaps not quite as clear, but it seems to me closer — assuming Bystander doesn’t know about the threat — to Bodyguard than to Coercion. The puzzle of third-party coercion is to explain how it is that Bodyguard and maybe also Bystander so starkly differ from Coercion.
Let me spend a couple of paragraphs saying more, first, about why this puzzle is puzzling, and second, about why it is important.
The puzzle of third-party coercion is puzzling because, to put it crudely, either consent does its thing, or it doesn’t. Coercion, we seem to think, undermines consent. It severs the tie between, say, someone uttering the words “I consent” (or more realistically, something like “Sure, you can have it” or “Here you go” or “Take it”) and them really consenting. Think of competence conditions for consent. If someone is incompetent to consent (to the relevant thing in the relevant situation), then, so the natural thought goes, they can say, “Sure, you can have it”, but this will just not amount to consent. If it doesn’t amount to consent, it just doesn’t; it’s null and void, it’s as if nothing consentish has happened. This fact — that no consent has been given — seems indifferent to the identity of the purported consent-receiver. And it is very tempting to think of coercion in a similar way: Coercion undermines consent. Under Badguy’s threat, Goodguy’s utterance “Take the money” does not amount to consent. Nothing consentish has happened (later on, we’ll have to qualify this statement). And that fact, too, it seems, should be indifferent to the identity of the money-receiver. If Badguy’s coercion renders Goodguy’s “consent” null and void, then how can Bystander or even Bodyguard be entitled to Goodguy’s money? It’s not as if Goodguy has consented — really consented — to this transfer. But Bodyguard is entitled to the money. Hence the puzzle.
This puzzle, it seems to me, is sufficiently interesting in its own right to merit philosophical exploration. But it is important also for two other reasons. The first one is that consent under morally problematic background conditions is ubiquitous. If we endorse a picture of coercion that makes use of a moral baseline — namely, if we think, as we have reason to, that whether a conditional offer amounts to coercion depends on what the party on the receiving end is already entitled to3 — then any case in which someone’s consent is partly caused by conditions that fall short of perfect justice is in danger of being proclaimed a case of third-party coercion. This is also why we probably can’t settle for classifying Bystander together with Coercion: Too many of us stand with regard to too many real-world transactions in something like Bystander’s position. Merely insisting on Bystander’s blamelessness — while surely important — will not suffice to save the moral permissibility and validity of all those transactions. Without a fuller understanding of third-party coercion cases, then, there is a danger that almost all cases of consent in the real world are invalid.4 We must avoid such a conclusion if at all we can.
The second reason why the third-party coercion puzzle is important is that making progress on it will likely amount to making progress on understanding consent in general and other cases of flawed consent in particular. We already have one such example — for coming up with a plausible thing to say about third-party coercion cases will, it will turn out, require revisiting the clean intuition emphasized above: that either consent does its thing or it doesn’t.
In the next section, I discuss a natural suggestion: that consent is relational in a way that helps with the puzzle. I distinguish between different ways in which it may be thought that consent is relational, arguing that the one needed for this thought to solve the puzzle amounts less to a solution to the puzzle than to its renaming. In the following (and central) section, I provide my suggested solution: I give reasons to understand consent (here and elsewhere) contrastively, and I show how doing so solves the puzzle of third-party coercion. Towards the end of that section, I also revisit the relational view and comment on its relation to mine.
Throughout, I use the examples I started with, where consent plays a role in the transfer of property or some such. In the relevant recent literature, though, such examples are less central than examples having to do with consent to sex.5 But I think there are weighty methodological reasons to start with other examples. I dedicate section 4, in which I also outline those methodological reasons, to a preliminary discussion of how what I say elsewhere in the paper applies also to consent to sex.
Before proceeding, though, several quick preliminaries about consent, and a methodological remark.
First, consent’s paradigmatic normative upshot is turning an otherwise impermissible action into a permissible one. But consent may have other normative implications as well, beyond this paradigmatic one.6 It may make some actions — interventions by third parties, for instance — more morally desirable, and some less. It may affect the weight of reasons already in place. It may affect questions about whether or not compensation for an action is owed, and if so, how much. And so on. Still, the puzzle of third-party coercion is — at least in the first instance — about the paradigmatic role consent has of turning otherwise impermissible action permissible and making certain transactions valid, so this is what I focus on below. When other normative upshots are relevant, I say so explicitly.
Second — and to an extent, a particular instance of the previous point — consent may not just make an otherwise impermissible action permissible. It may, even if it fails to make the action permissible, make it less seriously impermissible. You may think, for instance, that boxing is so harmful to boxers that it remains impermissible even given the boxers’ consent. Still, even if this is so, it will be very hard to deny that the boxers’ consent makes a moral difference: Even if one boxer punching the other is still morally wrong, surely it is (significantly) less seriously wrong than just one person sucker-punching another, precisely in virtue of the boxer’s consent.7 The sometimes neglected scalarity of consent8 can be expected to be relevant in our context as well, of course — you may think, for instance, that the consent in Bystander is, as it were, less valid than in Bodyguard, but more valid than in Coercion (so that in Bystander there are some, but not all, of the implications of fully valid consent). And I agree that the scalarity of consent is relevant here as elsewhere (and I briefly return to it below). Still, I think it’s clear that merely noting the scalarity of consent will not solve the puzzle of third-party coercion, as what we want to say about the consent to pay Bodyguard in Bodyguard is that it is entirely valid.
Third, it is important not to overdo the force of coercion. Coercion does, it is true, often undermine consent, render it invalid, and so on. But at least as often, it does not do so by annihilating the consent entirely as if (to put things in the words I use above) nothing at all consentish has taken place. Consider Goodguy’s after-the-fact thoughts in Coercion. He may — in suitable circumstances — be proud of himself, for instance, for responding in the way that was safest for himself and for others. But presumably, such pride only makes sense with regard to one’s actions — actions that one, in a sense, self-authors, actions that are an expression of one’s autonomous agency.9 This fact — that even in paradigmatic cases of coercion it’s not as if the consent is entirely non-present — will be important below.
Lastly, let me make my methodology explicit: I start, as can be seen above, with intuitions about rather clear cases (Coercion and Bodyguard). I take it to be an adequacy constraint on an account of coercion that it respect them — and if not, that it offer a very plausible debunking explanation thereof.10 The same will be true — though to a lesser extent — about other cases that will come up from time to time, cases about which my initial intuitions are not quite as robust. In all of these cases, it’s also going to be important that the account delivers not only plausible verdicts on the relevant cases, but does so for what upon reflection seem good reasons. And of course, other considerations, too, will be relevant — for instance, how well different accounts cohere with plausible general things to say about the nature of normative powers, how explanatorily deep and wide they are, and so on. Philosophical accounts — here and elsewhere — are to be evaluated both comparatively and holistically: Theories earn and lose, as it were, plausibility points on account of all these considerations (and others, too), and we should go for the theory that after all is said and done has the most plausibility points. In what follows, I will try to convince you that the theory I put forward as a solution to the puzzle of third-party coercion is that theory.
2. Going Relational
Perhaps we should think of consent as essentially relational. It’s not as if consent is valid or invalid. Rather, it’s valid vis-à-vis someone, or invalid vis-à-vis someone else, or more valid vis-à-vis one person than vis-à-vis another. (More simple cases, where we are tempted to say that consent is either valid or invalid full-stop, are seen as cases where consent is valid vis-à-vis everyone, or vis-à-vis no one, respectively.) And you see how this can help with the puzzle of third-party coercion: Perhaps Badguy’s threat renders Goodguy’s consent invalid vis-à-vis Badguy, but the consent remains valid vis-à-vis Bystander and vis-à-vis Bodyguard.11 I’m not going to reject this suggestion — in one way of understanding it, the story I’m going to end up telling can be seen as a version of the relational view. Still, it’s important to spend some time on this suggestion now, mostly in order to distinguish between different ways in which consent may be said to be relational (the distinction I will introduce between the relationality of content and of force seems to me of independent value) and in order to clearly see the crucial details that are missing from current versions in the literature. After presenting my view in section 3, I return to the relational view and the relations between it and my solution to the puzzle of third-party coercion.
Now, consent is clearly relational in its content. I can consent that my friend Gooddriver use my car but not that my other friend Poordriver use it. If so, it is now permissible for Gooddriver but not for Poordriver to drive it. It is natural to think of this point in terms of rights.12 My property right in the car includes a right against Gooddriver that she not drive the car, and a right against Poordriver that he not drive the car. By consenting that Gooddriver drive the car, I have waived my right against Gooddriver that she not drive the car, and so now, absent some other complications, it is morally permissible for her to drive the car. I have not, however, waived my right against Poordriver that he not drive my car, so that right is still in place and Poordriver ought not violate it.
Perhaps this, then, is the thing to say about third-party coercion cases? In Coercion, Goodguy’s consent that Badguy take his money is coerced and so invalid. Goodguy has not, in this case, waived his right (against Badguy) that Badguy not take his money. In Bodyguard, though, the situation is more complicated: Goodguy’s consent that Bodyguard take his money is valid. Vis-à-vis Bodyguard, Goodguy has successfully waived his right (that Bodyguard not take his money), though the consent remains invalid vis-à-vis Badguy. And perhaps this is the situation regarding Bystander in Bystander as well.
I don’t think, though, that such a story suffices as a solution to the puzzle of third-party coercion. In order to show this, I need to distinguish between two ways in which consent may be relational. I am not aware of anyone in the consent literature making the distinction we need here, so I’m going to introduce it in the context of duties, where it is more standard, and then apply it to consent. Duties, then, may be relational in their content, or they may be relational in whom they are owed to. If I (unproblematically) promise to pay you a sum of money, then my obligation is now regarding you (to pay you the money), and furthermore this obligation is owed to you. In this simple case, then, the two ways of being relational coincide. But they needn’t: If I promise you to look after your children, say, then the content of the duty has to do with your children, not with you; still, I owe this duty to you, not to them (it is you, not they, who possess the correlative right against me that I take care of them; it is you, not them, whom I wrong if I break my promise).
We can draw a similar distinction when it comes to permissive consent. Here, too, in many cases the two kinds of relationality coincide. If I give you my consent that you use my property, then you play a double role here — both in what I consent to (namely, that you use my property) and in that it is you to whom I give my consent — in rights-waiving terms: I waive my right against you that my property be used (by you). We can call these your role in the content of the consent and in its force, respectively. And in the consent case, too, we can imagine cases where the two roles come apart, though here more imaginative creativeness may be needed. Perhaps, for instance, I don’t have a right against your very young daughter that she not step on my lawn — perhaps I only have a right against you that your daughter not step on my lawn. If so, I may consent to you that she step on my lawn (this need not entail also my consent to you that you step on my lawn). Here, you don’t play a role in the content of the consent, but you do play a role in its force. For your daughter, the situation is reversed.
Armed with this distinction, then, what kind of relationality is invoked by the relationality-of-consent solution to the puzzle of third-party coercion? Invoking the relationality of the content of consent is plausible, but, I will now argue, cannot succeed in solving the puzzle of third-party coercion. And invoking the relationality in consent’s force — while a step in the right direction — does not go far enough unless it can rely on a story of the kind I supply in section 3.
That the content of consent is often relational cannot be doubted, as we’ve already seen. I may consent that Gooddriver use my car but not that Poordriver use it. And sexual consent is, of course, paradigmatically relational in this way. So there is no problem in invoking such relationality in our context as well. But consider:
Anonymous: Badguy sends Goodguy a text message reliably threatening that unless Goodguy pay that anonymous guy standing over there a sum of money, Badguy will beat up Goodguy. Goodguy agrees to pay that guy, and pays.
What Goodguy consents to is to pay that guy over there. The consent is thus relational in its content: to pay that guy, rather than someone else, say. But the person Goodguy consents to pay is picked out ostensibly (by pointing) or by a description (the person at the corner) rather than by name (Bystander, or Badguy). And surely, that in itself is no flaw — we can and often do consent in this way (say, to that woman swimming next to me borrowing my fins, or to this guy selling me coffee taking my money). And this means that the content of the consent in Anonymous remains the same whomever that guy turns out to be, and in particular whether he turns out to be Badguy or Bystander. But the status of the instance of consent is different in these two cases. So the relationality of the content of consent cannot explain the moral difference between Coercion and Bystander.13
If something about the relationality of consent does this job, then, this should be the relationality of its force. The thought seems to be that in Anonymous, Goodguy’s consent (with the fixed content that that guy take the money) is invalid vis-à-vis Badguy, but could be valid vis-à-vis others like Bystander. If so, then in the version of Anonymous where that guy turns out to be Badguy, no valid consent vis-à-vis him has been given and the situation is as it is in Coercion. But in the version of Anonymous where that guy turns out to be Bystander, valid consent vis-à-vis him has been given, and the situation is as it is in Bystander.
This story is on the right track, but as things stand, it is explanatorily unsatisfying. What explains why it is that Goodguy’s consent is valid vis-à-vis Bystander but not vis-à-vis Badguy? The coercive element in Badguy’s behavior is present in both. How can it succeed in rendering the consent valid vis-à-vis one but not the other? The presence of consent or something resembling it, with it being valid vis-à-vis Bystander but not vis-à-vis Badguy — while not impossible — does call for explanation. And notice again that the explanation cannot be in terms of the content of the consent, as the discussion of Anonymous shows. Similarly for Bodyguard: If Goodguy gives Badguy his consent to pay Bodyguard, that consent is invalid. But if Goodguy gives such consent to Bodyguard, it is valid. How does the suggested story explain this difference? Without such a story,14 the relational attempt at a solution to the puzzle of third-party coercion seems less like a solution and more like another name for the problem.15
I return to the relational view — and to how my theory naturally completes it — in section 3.4 below.
3. Going Contrastive16
The key to solving the puzzle of third-party coercion, I now want to suggest, is to appreciate that consent is contrastive. It’s not that an agent consents to something. Rather, they consent to one-thing-rather-than-another. Often, the contrast remains implicit and is clear in context: When I give my consent to Gooddriver driving my car, I consent to her driving it rather than her not driving it (with pretty much all else remaining constant). When I consent to the person making my coffee taking my money, I consent to him giving me coffee and taking my money rather than me keeping my money but also remaining thirsty. At other times, filling in the contrast is less obvious, and can help clear away confusions. This is the case, I will argue, with third-party coercion cases. I develop this idea here as follows: In section 3.1, I elaborate on why we should understand consent contrastively. In section 3.2, I fill in the details of the contrastive understanding of consent and give the official statement of my account. In section 3.3, I revisit the different cases already discussed, showing how thinking of consent contrastively sheds light on them, and so solves the puzzle of third-party coercion. In section 3.4, I return to the relational view, and I show how the contrastive account fills in the details of (the most plausible version of) that account. And in section 3.5, I hint at other ways — not directly related to third-party coercion — in which understanding consent contrastively is helpful.
3.1 Why Go Contrastive?
First, then: Why think that consent should be understood contrastively?
One reason has to do with how theoretically productive the (suggested) contrastive nature of consent is. Here, then, the proof of the pudding will be in the eating: After I use the idea of contrastive consent in an analysis of third-party coercion cases, it will be clear, I hope, that going contrastive is indeed theoretically productive. And the point of section 3.5 is to give a feel of how theoretically productive thinking of consent contrastively is in other contexts as well, so that invoking consent’s contrastive nature in a discussion of third-party coercion is in no way ad hoc. Indeed, I hope that such a discussion will also make it clear how natural it is to think of consent as contrastive in this way.
But even before eating the pudding, I believe we can say the following: Consent is very closely tied to the will, to desires, indeed to preferences. The nature of the connection is perhaps not entirely clear, but that there is such a connection cannot be plausibly denied. And preferences (to pick perhaps the clearest case) are essentially contrastive: You always prefer one thing over another. At times, the contrast remains implicit and can be completed from context: If, at a restaurant, you note the desert you prefer, it’s clear that you prefer it over the other options on the menu. Similarly, if asked whether you want to go to the dentist, it seems natural to say that you want to go to the dentist rather than to continue suffering that toothache (after all, without such a desire, how do we explain the fact that you’re calling to make an appointment?), but of course if the alternative had been not to have the toothache to begin with, you don’t desire to go to the dentist rather than that. And if these other phenomena in the vicinity of consent are contrastive, it is a plausible hypothesis that so is consent. At least, it is plausible enough to have a taste of the pudding.17
And here is another major advantage of understanding consent contrastively (to an extent, it is a particular instance of the first): It allows greater leeway and subtlety in understanding complex situations. In Coercion, for instance, suppose someone asks you, “Well, which is it? Has Goodguy consented to pay Badguy, or hasn’t he?” With a simple, non-contrastive understanding of consent, you should be torn: On the one hand, Badguy clearly doesn’t get to keep the money, so the consent must not be valid. On the other hand, Goodguy may rightly take pride in consenting (because this was the only way to avoid, say, danger to himself and others). So something consentish must have taken place. Going contrastive allows you to say the following: Goodguy consented to paying-rather-than-being-subject-to-Badguy’s-violence. And this consent was the right call, which explains why he can take pride in it. On the other hand, Goodguy never consented to paying-rather-than-not-being-subject-to-Badguy’s-threat. And this (together with some details I get to in the next subsection) explains why Badguy doesn’t get to keep the money.
3.2 The Account Detailed
I suggest, then, that consent be always understood as contrastive. I don’t claim, of course, that contrastive content has to be there, explicit, in the consent-giver’s mind. Rather, I claim that the best understanding of what is going on when one consents is contrastive — in a way similar to how we routinely utilize implicit beliefs in our psychological explanations, or in which we think of preferences as essentially contrastive. Consent, in its nature, is contrastive, even if this nature is not always explicit and obvious.18
Often, as we saw with regard to Coercion, a given instance of consent may be valid with one contrast class and invalid with another. Indeed, the whole point of going contrastive is precisely to allow for such possibilities. In order to have an answer to the question of whether the consent is valid (whether, say, Badguy gets to keep the money in Coercion), we need a way of determining what, for some specific purpose, the morally relevant contrast is.
What determines the morally relevant contrast, I now want to suggest, is the normative relationship between the relevant parties and, in particular, the options the consent-giver is entitled to and against whom. In Coercion, Goodguy is entitled against Badguy that Badguy not threaten him in the first place. In other words, Goodguy is entitled (against Badguy) to the option of keeping-his-money-and-not-be-subject-to-Badguy’s-violence. So the relevant contrast is precisely this one, and the question we must ask is whether Goodguy consented to paying-rather-than-keeping-the-money-and-not-be-subject-to-violence. But of course, it is clear that Goodguy never consented to that; he never consented with this as the relevant contrast. (He did consent to paying-rather-than-being-subject-to-Badguy’s-violence, but seeing that he is antecedently entitled to more from Badguy, this is not the morally relevant contrast.)
As already noted, talk of which options the consent-giver is antecedently entitled to is not new. At least since Nozick (1969), it is central to the understanding of coercive offers. The innovation here, then, lies not in the emphasis on that question, but rather in the structure a contrastive understanding of consent gives to the way that question is answered.19 The very same instance of consent may be — depending on the relevant contrast — valid in one way but not another; and the prior entitlement to options determines the morally relevant contrast for specific purposes. The best way to see this is to go through the cases again and see how the suggested account plays out.
3.3 The Cases Again
We’ve already seen how a contrastive understanding of consent nicely delivers on Coercion: Badguy doesn’t acquire a right to the money, because Goodguy never consented to paying rather than keeping his money and also avoiding harm, and this, for the purposes of determining the status of the money transfer, is the morally relevant contrast. But Goodguy did consent — fully, authentically, autonomously consent — to paying rather than suffering harm at Badguy’s hands. This consent, with this contrast, is fully his, and it reflects his preferences prior to the interaction with Badguy.20
How about Bodyguard? Goodguy consents to Bodyguard getting the money rather than Goodguy keeping his money but remaining unprotected. That consent is perfectly Goodguy’s own, it accurately reflects his preferences, and it is not affected by Badguy’s threat — that threat affects the relevance of this preference and consent, but not the relative order of the options. Of course, Goodguy doesn’t consent to Bodyguard getting the money rather than Goodguy not being under a threat to begin with. Which of these is the morally relevant contrast for which purpose? This depends, as above, on what contrast Goodguy is (pre-interaction) entitled to, and against whom. Clearly, Goodguy has a right against Badguy to have available the option of keeping his money and not being harmed, which is why if then Goodguy sues Badguy for Bodyguard’s fee, Badguy can’t defend himself by claiming that Goodguy consented. But Goodguy does not have a right against Bodyguard that Badguy not threaten him.21 In Goodguy’s relationship with Bodyguard, then, Badguy’s threat is just a part of the background circumstances. So the contrast in Goodguy’s consent — to give Bodyguard the money rather than to keep the money and be vulnerable to Badguy’s harming him — is a morally relevant one. So Bodyguard is entitled to the money. The same analysis applies, it seems to me, to Bystander:22 Goodguy is not entitled against Bystander to not be harmed by Badguy. In the interaction with Bystander, then, Badguy’s threat recedes to the status of background circumstances, and so Goodguy’s consent to giving the money to Bystander rather than suffering harm at Badguy’s hands is the morally relevant one.
As I said at the outset, I think that our — certainly my — intuitions about Bystander are not as clear and robust as they are about Coercion and Bodyguard. So I think we should be willing to let the best theory of third-party coercion dictate what we end up saying about Bystander. The theory suggested here — which delivers the right verdicts in the paradigmatic Coercion and Bodyguard cases, and seems on reflection to do so for the right reasons — classifies Bystander with Bodyguard rather than with Coercion. Here, too, it seems to me that it does so for what on reflection seem good reasons. But a worry remains, first, because Bystander doesn’t feel quite as unproblematic as Bodyguard does, and second, because there’s a fairly robust intuition that if the relevant third party knows about Badguy’s coercive threat, this makes a huge difference in Bystander, but not in Bodyguard. So I want to show how I can accommodate this intuition (and thereby also vindicate the feeling that some important difference between Bystander and Bodyguard remains).23 In Bodyguard, Goodguy is not entitled (against Bodyguard) to Bodyguard’s services (or so we’ve been assuming as a part of the stipulated case). This doesn’t change when Bodyguard knows of Badguy’s unjust and coercive threat. In Bystander, though, the situation with the relevant entitlements is different: In the case we started with, when Bystander doesn’t know of Badguy’s coercive threat, Goodguy has no relevant prior entitlement against Bystander, and then the analysis from the previous paragraph applies. But if Bystander knows of Badguy’s coercive threat, this changes things: Now another contrastive option becomes salient. Bystander can — and can be expected to — take the money from Goodguy just in order to help him avoid Badguy’s violence, and with the intention to return it to Goodguy after Badguy no longer poses a credible threat to Goodguy’s bodily integrity.24 And assuming that holding on to the money is not burdensome for Bystander, this is what he ought to do; indeed, this is what he owes it to Goodguy to do. In other words, in this case, while it remains true that Goodguy is not entitled against Bystander not to be threatened by Badguy to begin with, Goodguy is entitled against Bystander to be protected in this way from Badguy’s aggression. And of course, when Goodguy consents to giving Bystander the money, he doesn’t consent to giving the money for him to keep rather than for him to hold on to temporarily until Badguy’s threat is no longer relevant. And, as I’ve just argued, when Bystander knows of the threat, Goodguy is entitled against Bystander to this contrast. So this is, on the analysis here offered, the morally relevant contrast. And this means that Bystander does not, in this case, get entitlement of the money.
What about Anonymous? Goodguy’s consent to let that guy have the money rather than be harmed by Badguy is perfectly valid, as above. Of course, Goodguy never consents to pay that guy rather than keep his money and not be harmed either. And what’s the morally relevant contrast? This depends on who that guy is. Seeing that Goodguy has a right against Badguy to keep his money and avoid harm, if that guy just is Badguy, Goodguy’s right holds against that guy as well, so that the relevant contrast is the second one, and that guy standing there is not entitled to the money. If, however, that guy ends up being an uninvolved bystander, then the case just is, well, Bystander. These are, it seems to me, precisely the right things to say about Anonymous.
In all third-party cases, it is a robust intuition that if the consent-receiver (Bodyguard or Bystander, for instance) is in cahoots with the one doing the coercion (Badguy), the consent is invalid. And the above story, utilizing the idea of contrastive consent, can easily explain why. If Bystander, for instance, knows about Badguy’s threat, benefits from it, does nothing to eliminate it (or to return the money), or indeed encourages Badguy to issue the threat — in such cases, Goodguy does have a right against Bystander to have available the option of keeping his money and also not being harmed. And, as we already know, with this as the relevant contrast, Goodguy has never validly consented to paying Bystander. Similarly for Bodyguard: If Bodyguard is in cahoots with Badguy, then Goodguy’s entitlement to the option of keeping his money and not being harmed — an entitlement he clearly holds against Badguy — extends against Bodyguard as well. This determines the morally relevant contrast, and explains why — with the relevant contrast — Goodguy never validly consented to Bodyguard getting the money. And this explains why Bodyguard is not entitled to the money.25 But in cases in which there is no similar connection between Bodyguard and Badguy, Bodyguard is entitled to treat the threat as a part of the background circumstances. In such a case, the situation is — as far as the normative relationship between Goodguy and Bodyguard — no different from a situation in which Bodyguard is hired to protect Goodguy from the effects of a natural disaster.26
In all these cases, the contrastive nature of consent plays an indispensable role in the suggested explanation: Only once we understand consent contrastively do we see the relevance of the more fine-grained questions here — who is entitled to what options as against whom — as a way of determining the contrast that is morally relevant for a specific question about consent’s validity.27
Recall the “Either consent does its thing or it doesn’t” line of thought from the beginning of the paper. The story just told respects it in one way, and rejects it in another. If the thought is that even when caused by Badguy’s threat, Goodguy either validly consented or he didn’t, then I reject it, because it uses too coarse-grained a specification of the content of the consent. We should add in the contrastive clauses, and then check to see which contrast is the morally relevant one in which interaction. But, once we’ve done that, then the “either consent does its thing or it doesn’t” line nicely applies:28 Given a specific contrast, the consent is either valid, or it isn’t — without any “vis-à-vis” in there.
3.4 Relationality Again
On the emerging picture, then, the validity of consent is not itself directional; for each specified contrastive content, either the consent is valid, or it isn’t. There is no need for a validity-vis-à-vis. But some relationality kicks in at the second stage in determining which of the contrasts is relevant for which interaction. Because this moral relevance is determined by the normative relations between the relevant parties (whether, for instance, Goodguy has a right against Bystander for the availability of some option), what upshot the relevant consent has also depends on these specific normative relations. And there, “vis-à-vis” makes a comeback. This is why Goodguy’s consent in Bodyguard establishes Goodguy’s duty to pay Bodyguard, but in no way affects Badguy’s liability (say, to a requirement to compensate Goodguy for his expense on hiring Bodyguard).
Aren’t we back, though, with an account that is extremely close to the relational account I was less than enthusiastic about above?29 Depending on how exactly a relational account is understood, we may be. But the emerging picture significantly improves on existing relational accounts. There is an important difference between (other) relational accounts and my account in terms of where relationality is invoked. In the relational accounts discussed above, it was consent’s validity that was taken to be relational. This meant that such accounts needed a primitive valid-vis-à-vis predicate — a solution that is non-parsimonious, that leaves the most important things unexplained (how is it that validity can be vis-à-vis?), and that doesn’t sit well with the “either consent does its thing or it doesn’t” intuition. In my account, though, there is no need for a primitive valid-vis-à-vis predicate. The only relationality the account takes as input is that of the relevant prior entitlements. This way of doing things better respects the “either consent does its thing or it doesn’t” intuition, it is more elegant and parsimonious (we already knew, of course, that entitlements may be relational, so utilizing this mechanism does not incur a further theoretical price), and it leaves no unexplained mysteries.30
Does this mean that at the end of the day my account is a competitor of relational accounts, or an instance thereof? Not much depends on such taxonomical questions. The more important questions are explanatory and evaluative. In my account, validity-vis-à-vis does not do any explanatory work. All the explanatory burden is shouldered by the notion of validity we all know and love, together with the contrastive nature of consent and the relational nature of the relevant prior commitments (and we have independent reasons to include all of these in our normative resources). If you want, you can construct a notion of validity-vis-à-vis out of those. I have no objection to this way of putting things, and so to classifying in this way my view as a particular instance of relational accounts. But it’s important to see the explanatory ways in which it differs from more standard ones — and, if I’m right, for the better.31
3.5 Other Payoffs?
What I already said suffices, I hope, to show how plausible it is to think of consent contrastively, and relatedly, how theoretically productive it can be to do so in the context of third-party coercion. But the major point relied on — that consent is best understood contrastively — is much wider in scope. If it is true, then, we might expect to see the theoretical payoffs of such an understanding of consent arising elsewhere as well. At least, seeing such payoffs in other contexts may help to ward off worries about this being an ad hoc solution to the problem of third-party coercion. Without pretending that the discussion below is at all exhaustive, here are hints at how theoretically productive it is to think of consent as contrastive in a couple other contexts.32
What are we to say of contexts in which someone — a victim of oppression, for instance — consents to some arrangement that, despite being oppressive, may still be the best available option for her given the oppressive circumstances she finds herself in?33 Suppose that a woman who lives in a highly patriarchal, polygamous community consents to being some man’s third wife. Is her consent valid? Obviously, much more can be said about such “bargaining with the patriarchy” cases,34 and it’s not as if all of them are alike. But understanding consent contrastively allows us to say the following: The woman in the example consented to becoming-the-man’s-third-wife-rather-than-other-options-available-to-her-in-her-community (options which may be very bad), but she’s of course never consented to becoming-the-man’s-third-wife-rather-than-living-in-a-more-egalitarian-soceity (with all the other options this would open for her). So her consent (with the first contrast) in no way implies that there’s nothing wrong or unjust or oppressive about the set of options she finds herself with (she hasn’t consented to that), and the normative upshots of her consent depend, as above, on the morally relevant contrast in the relevant context, and in particular, on what she is entitled to and against whom (see, in this context, the discussion of men who are collaborators with the patriarchy, and what women are entitled to against them, in the next section). This way of analyzing bargaining-with-the-patriarchy cases — which falls out of the account developed in this paper — seems to me intuitively plausible and sensitive to the right features of the cases.
How should we think of — and what should we do about — sweatshops in which vulnerable people work in terrible conditions and for very little pay, perhaps as a way of producing products for people in richer parts of the world? On the one hand, there is clearly something deeply wrong about such a practice. On the other hand, in many of these cases, those working at sweatshops have no better options, and have given their consent to the terrible conditions they work in. I suggest, then, that we should start by thinking of their consent in contrastive terms. They — or anyway some of them — validly consent to working-in-those-conditions-rather-than-being-unemployed (with the rest of the background conditions held constant). Of course, they haven’t consented to working-in-those-conditions-rather-than-living-in-a-somewhat-more-just-international-economy. Is it morally permissible for you, say, to buy the products they are taking part in producing? The analysis in this paper suggests that this partly depends on whether or not they have a right against you for a better system. If they do not — even if they have such a right against others — their consent (which, in this case, also accurately reflects their interests) seems to matter in the usual consent-kind-of-way. In the opposite direction, the more you should be seen as working in cahoots with the system that so treats them, the less you can hide behind their consent. The situation is complex, and many more considerations apply (about long-term effects on the industry, about likely intermediate stages on the way to systemic improvement, about the relation between you and the society of which you are a part, about the workers’ interests even independently of their consent, and more). But as far as the significance of their consent is concerned, I believe that the analysis above delivers the right results.
4. Consent to Sex
Let’s recap, then. When we are tempted to say that someone consents to something, we should fill in the contrastive details: What are they consenting to rather than what else? It’s possible, of course, that consent to one and the same thing may be valid when it’s rather-than-some-other-thing but invalid when it’s rather-than-some-third-thing. What renders consent (with its more fine-grained contrastive content) valid are factors that lie upstream from it about which I said nothing in this paper — things like what caused the consent, whether it genuinely reflects the choice of the consent-giver, whether it is sufficiently informed, and so on.35 And prior entitlements determine which of the many possible contrasts are the morally relevant ones for which interactions.
So far, I’ve been using examples that have to do with the transfer of property. But much of the relevant literature focuses on consent to sex. The reason I proceeded in a different direction is that it seems to me advisable on methodological grounds to start with simpler cases, then extend our understanding to the more complicated ones. And when it comes to consent — including under third-party coercion — I think consent to sex is a much more complicated and less paradigmatic case than consent to such things as paying a bodyguard.36 Still, it’s important to see how the analysis here offered does apply to consent-to-sex cases. The discussion that follows is an attempt to show — in a very preliminary way — how this is so. It goes without saying that this is not even an attempt at a full story regarding third-party coercion as applied to consent to sex. The discussion in this section can also be seen as bolstering the case preliminarily made in section 3.5 above, showing the payoffs of understanding consent contrastively across wider contexts.
Now, it’s not obvious that permissible sex is entirely about consent.37 But when it is about consent — and sometimes, perhaps when we ask about the most minimal standard of morally acceptable sex, it is — then the discussion in this paper may apply. Let me conclude, then, with the following implications — just sketched here — to some of the relevant cases in the literature:
Pimp:38 Pimp threatens Victim to beat her up unless she has sex with Stranger for money and gives that money to Pimp. Victim approaches Stranger suggesting sex for money, and Stranger accepts.
In Pimp, Victim validly consents to having sex with Stranger rather than being beaten up by Pimp. She does not, of course, consent to having sex with Stranger rather than not being subjected to Pimp’s threat to begin with. If Stranger knows nothing of Pimp’s threat and is in no way cooperating with him (and if the case does not have other complicating factors), then Victim is not entitled against Stranger not to be subject to Pimp’s threat. In that case, Stranger’s action is morally permissible. (Of course, the stipulated conditions — where Stranger doesn’t know of Pimp’s threat — may be unrealistic in many real-life cases.)
Pimp Without Pay:39 Pimp threatens Victim to beat her up unless she has sex with Stranger. Pimp does not expect to get the money, and if Victim enters a room with Stranger and stays there for a while, will conclude that she has had sex with Stranger. Victim consents to having sex with Stranger in order to avoid the beating.
The important difference between Pimp and Pimp Without Pay is that if Stranger knows that being in the room suffices to avoid Pimp’s threat, and nevertheless has sex with Victim (perhaps she doesn’t know that), this is impermissible and he can’t rely on Victim’s consent. Utilizing the contrastive nature of consent can accommodate this: In Pimp Without Pay, Victim has validly consented to having sex with Stranger rather than being beaten up by Pimp, but she has not validly consented to having sex with Stranger rather than neither having sex with him nor being beaten up (by staying in the room with Stranger). And if Stranger knows this is, as far as Pimp’s threat, an available option, then Victim has a right against Stranger to this option. And then this is the relevant contrast, so that Stranger nevertheless having sex with Victim is impermissible. In fact, in such a case, it is very tempting to think of Stranger as, to an extent, working together with Pimp.
Heterosexual Sex Under Patriarchy:40 Man and Woman live in a patriarchal society, which has shaped them in the way societies shape their members. Woman consents to having sex with Man. A part of what explains her consent — and it is always difficult to say what part — is caused by the patriarchal order (perhaps the way it has shaped her desires, or her alternative options, or her self-conception).
Woman’s consent was at least partly caused by the injustices involved in the patriarchal system, and so — perhaps depending on what we should say about structural injustice — the case resembles some third-party coercion cases. I suggest, then, that in many such cases, Woman validly consents to having-sex-with-Man-rather-than-not (holding the patriarchal system fixed).41 Woman has never consented, of course, to having-sex-with-Man-rather-than-living-under-an-alternative-non-patriarchal-system (with all the options, including the sexual options, this would involve).42 In many cases, though, she does not have a right against (that specific) Man to have that option available to her. So her consent — with the first contrast above — makes it permissible for Man to have sex with her.
This of course is a schematic example — as indeed are the other ones in this section — and real-life cases are going to be much more complicated and intricate. Further analytic and explanatory resources will have to be employed — for instance, regarding the scalar nature of consent. But it is a good feature of the suggested analysis that it generates both the right sensitivity to further factors, and interesting further results. Here is one: On my suggestion, the extent to which third-party coercion undermines the validity of the relevant consent (with the relevant contrast clause) depends on whether, and to what extent, the consent-receiver is in cahoots with the issuer of the threat. Applied to our case, this means that we need to ask to what extent (that specific) Man is in cahoots with the patriarchy. And roughly, the more he is, the less permissible his having sex with Woman, not because her consent to have-sex-with-him-rather-than-not in the patriarchal system is invalid, but rather because she’s entitled — against him, among others — to improvements in the direction of a non-patriarchal system.43
Notes
- The title of this section is also a part of the title of Millum (2014). My presentation of the puzzle is somewhat different, but not, I think, in ways that make a serious substantive difference. ⮭
- This example was given to me by Larry Alexander in correspondence that forced me to think much more carefully about these matters than I had previously. ⮭
- See Nozick (1969), Berman (2002), and the references there. Here is a typical example. If a drug company lets you have the life-saving medicine you need only if you pay them, and you agree to pay, their offer does not seem (necessarily) coercive, and your consent is valid. If, however, I then steal the medicine, and will only return it to you if you pay me, and you agree to pay, my offer is coercive, and your consent is at the very least flawed (and if you think it’s flawed in the previous case, too, surely you agree at least that it’s more seriously flawed in this one). What best explains the intuitive difference between the cases is that in the first case you are not, before the transaction, entitled to the medicine (arguably, the drug company is), whereas in the second case you are. ⮭
- For this as a central motivation for their accounts of third-party coercion, see, for instance, Gerver (2021) and Tadros (2021). ⮭
- This is the context, for instance, of the recent discussion of third-party coercion in the symposium in Ethics 131 (2021). ⮭
- For some examples and discussion, see my “Hypothetical Consent” (2017). ⮭
- The example is taken from my (2017, 15). ⮭
- This term is central to Dougherty (2021b). Dempsey (2023) argues both that the moral implications of consent are a matter of degree, and that the legal ones are not — they are either-or. ⮭
- For an emphasis on the need to take seriously the agency of those making decisions under oppression (for instance, of women under patriarchy), see, for instance, Gerver (2021) and Tadros (2021). The point in the text suffices, it seems to me, to reject Millum’s (2014, 115) claim that the problem with threats is that the issuer of the threat takes control over her victim’s decision. It is less clear how the observation in the text reflects on views that tie the conditions for the validity of consent closely to the condition of moral responsibility (see, e.g., Hurd (1996) and White (2017); and see Liberto (2021) for criticism). Initially, it seems the answer is clear: If Goodguy can be proud of the invalid consent (to give Badguy the money), the conditions of moral responsibility and of valid consent cannot be too close to each other. But on the account I’m going to end up offering, we need to be more careful about the content of the relevant consent by going contrastive. Then, there’s room in logical space for a view that denies the validity of one consent (with one contrast class), but affirms the responsibility of another (with another contrast class). ⮭
- Perhaps one kind of debunking explanation is going to try and make sense of the relevant intuitions in normative terms other than those of permissibility, perhaps even ones to which consent is relevant. At one point in what follows — in discussing a variation on Bystander where Bystander knows about the threat — this will become relevant. ⮭
- See Gerver (2021), Kiener (2022), and perhaps, to an extent, Alexander (1996). ⮭
- Alexander’s (1996, especially 166) account of consent puts rights and waiving of rights center stage. ⮭
- In conversation, Pär Sundström suggested another possible way to go here: to insist that the relevant relationality is after all in content, but in content de re rather than de dicto. This seems to get the right result regarding Anonymous. I’m not sure what exactly I think about this or what follows from this interesting suggestion. Let me just say, first, that I think understanding the relevant relationality here in terms of force rather than content seems independently plausible, and second, that it’s not clear to me what could motivate the view that takes the relevant relationality to be in content de re except for the need to accommodate the case. ⮭
- Gerver (2021) may be read as trying to fill in this explanatory gap by starting with judgments about the permissibility of the relevant actions then deriving an account of the validity of the relevant instance of consent from that. This way of thinking about the relation between the validity of consent and the permissibility of the consented-to action is inconsistent, though, with taking consent seriously as the exercise of a genuine normative power. The validity of the relevant instance of consent is explanatorily prior to the permissibility of the consented-to action. ⮭
- Kiener’s (2022) account is a good example here, partly because I think he gets the cases right, and better than the alternative relational accounts he rejects (see his discussion of and references to Owens, Dougherty, and Gerver (Kiener 2022, 378–382)). Still, Kiener is very much vulnerable to the explanatory problem highlighted in the text. See, for instance, his discussion of the Envelope case (384), where he asks whether a consent-recipient’s later action can invalidate an earlier consent. The need for a further explanatory layer — supplied by my account below but missing from relational accounts in the literature — is especially clear here. ⮭
- Tadros (2020, 243–251) discusses contrastive consent, but he thinks of it as a particular instance of consent, not as a feature of all cases of consent. For a detailed discussion (but without using this term) in the context of a discussion of (among other things) third-party coercion, see Liberto (2021). Alexander (1996, 171) hints at the direction of going contrastive. Contrastive consent is central to Pummer (2022), though in a different context. Note that Pummer does not seem to think, as I do, that all cases of consent are contrastive. Still, and despite some details on which we may differ, I am indebted to Pummer’s discussion. And I emphasize the significance of the contrastive nature of consent — in the context of a discussion of oppression and indeed of false consciousness — in my “False Consciousness for Liberals, Part I” (2020). ⮭
- Notice that I don’t claim that permissive consent amounts to an expression of a preference. Obviously, I can grant my consent to your driving my car without having any preference that you do so. The point in the text is more restricted: Consent is somehow rather closely related to preference, and so that fact that preferences are contrastive lends some initial support to the thought that consent is, too. ⮭
- For structurally similar moves regarding knowledge and causation, see Schaffer (2005a, 2005b). Ezequiel Monti raised the question (in conversation) whether other normative powers (commands? requests?) should, for similar reasons perhaps, be seen as contrastive. I am not sure what to think about this. I suspect that they sometimes are, but perhaps not always and as necessarily as consent (so I argue in the text) is. ⮭
- Here’s an interesting complication: Is consent, understood contrastively, transitive? If one consents to A-rather-than-B, and to B-rather-than-C, does this mean one has already consented to A-rather-than-C? Or perhaps that one is at least under rational pressure to consent to A-rather-than-C? I suspect the answers to these questions are No and Yes, respectively. And I suspect that in order to make sense of this, we need a distinction between two values of autonomy — the ones I elsewhere call non-alienation and sovereignty. The negative answer to the first question shows that often the significance of consent is best understood in terms of sovereignty, and the positive answer to the second is grounded in the significance of non-alienation. For a discussion of this distinction, see my “Autonomy as Non-Alienation, Autonomy as Sovereignty, and Politics” (2022). For a discussion of transitivity and contrastive reasons, see Snedegar (2014). And for the somewhat structurally similar topic of “trafficking in knowledge ascription”, see Schaffer (2005a, 239). It seems to me that consent is less vulnerable to these worries than knowledge is (so that, at least in this respect, a contrastive understanding of consent is more plausible than that of knowledge). ⮭
- In my “False Consciousness for Liberals” (2020) I argue that when a choice — for instance, to give consent — is motivated by preferences in the causal history of which injustice plays an appropriate role, that choicefails to manifest one of the values of autonomy (the one I call autonomy as non-alienation). Notice, then, that the point in the text shows that the preferences guiding the consent to pay rather than to be harmed is not caused by the injustice of Badguy’s threat. The threat is relevant in making this preference in some way relevant, not in shaping it. ⮭
- I return to cases where this is not so below. ⮭
- But suppose Bystander puts the money under his bed, and later that day Goodguy shows up, explaining that he only gave Bystander the money because of the threat, and asks for his money back. It seems clear that Bystander should give him the money. (See the “Cash Transfer” case in Millum (2014)). Doesn’t this show, then, that something is wrong with the initial consent to Bystander getting the money? I think not: I agree that Bystander should, in this case, give Goodguy the money back. But I think that the situation is similar to a situation in which Bystander just finds a stash of money — in some circumstances, it would be perfectly permissible for Bystander to take the money; still, if it then becomes clear whom the money belonged to, Bystander ought to return the money. In other words, here I think that the intuition that Bystander should return the money is correct, but it doesn’t show that the consent was invalid to begin with — rather, it is rooted in some other normative considerations, not having to do with the permissibility of initially taking the money (or with the consent grounding it). ⮭
- For pressing me on this and on related worries, I thank Dani Attas, Alex Guerrero, Frances Kamm, Mike Otsuka, Saul Smilansky, and Larry Temkin. ⮭
- Thus understood, the case is similar to the case Pimp without Pay (Millum 2014), which I discuss below. ⮭
- And of course, there may be intermediate and borderline cases: cases in which Bodyguard is not exactly working with Badguy, but doesn’t really put an effort into eliminating Badguy’s threats (they are, after all, great for business), and so on. In such cases, thinking of consent in a scalar way will again be useful. Tadros (2021) thinks that a part of what determines whether the consent-receiver’s action is permissible are his intentions. The account here is, to an extent, an alternative: What matters directly is not the consent-receiver’s intentions, but rather which contrastive options the consent-giver is entitled against the consent-receiver to have, and this depends on whether the consent-receiver is working together with the coercer. Intentions are relevant on this picture only if they are relevant to the question whether they should be seen as acting together, and even then, indirectly. Millum’s (2014) suggestion that the permissibility of the action depends on whether the issuer of the threat and the recipient of the consent can think of the relevant decisions as joint decisions is closer to my suggestion. It’s not identical, though; the (admittedly vague) condition in terms of working together — or being in cahoots — seems to me wider in relevant ways than the suggestion in terms of a joint decision: Bodyguard and Badguy can be in cahoots, for example, even if at no point did a joint decision take place. For a critical discussion of Millum’s suggestion, see Kiener (2022, 370). ⮭
- In a different context, I defend the thesis I call “Merely Circumstances”; that is, roughly, the thought that in taking into account other people’s expected moral violations, I should think of them as merely circumstances, making a difference in the same way other circumstances do, so that their being moral violations drops out of the picture. (See my “Against Utopianism” (2018). See also Tadros (2016).) The point in the text here is different, but it nicely coheres with that thesis. ⮭
- We can now also see that understanding consent contrastively can help explain consent’s scalarity. (I thank Jonathan Schaffer for helping me see this.) For instance, the strength of the entitlement to the relevant contrast may be a matter of degree — perhaps this is so when the third party is, to different degrees, in cahoots with the coercer. Perhaps this is so when someone like Bystander is, to different degrees, aware of Badguy’s threat or of its unjust nature. And there may be other scalar dimensions as well, which — on the view defended here — are inherited by the normative force of the relevant (contrastive) consent. ⮭
- Kiener (2022, 372) rejects what he calls “The Influence-Focus View”, which is very close to the thought I capture in the text with the phrase “either consent does its thing or it doesn’t”. But I go a longer way towards accommodating its underlying intuition by showing how, with the content suitably contrastive, that thought does hold. ⮭
- This worry may be especially worrying regarding Gerver’s (2021) recent relationality-of-consent account that also, like my own, emphasizes the question of which options Goodguy is entitled to and against whom. My two-point response below, then, is, to a considerable extent, my explanation of how my account is still importantly different from Gerver’s (who doesn’t make any use of the contrastive nature of consent). ⮭
- I think that with regard to this point of where does the relationality come in, Liberto’s (2021) account is in line with relationality-of-consent accounts rather than with my account, and this despite the fact that her discussion, just like mine, focuses on consent’s contrastive nature. ⮭
- I thank an anonymous referee at this journal for helpful comments here. ⮭
- Nathan Howard suggested to me that the contrastive nature of consent can explain the need for informed consent, and perhaps some of the details of what informed consent comes to: For arguably, one cannot agree to option A rather than option B unless one knows the relevant facts about option B. ⮭
- See my discussion of “bargaining with the patriarchy” cases in my “False Consciousness for Liberals” (2020), and the references there. ⮭
- See Stoljar’s (2014, 231) discussion of “bargaining with patriarchy”, and the references there (the phrase comes from Uma Narayan). ⮭
- I say a bit about all this in my (2020). ⮭
- At least one reason why consent-to-sex cases are less paradigmatic is that it’s not at all clear that the permissibility of sexual interactions is best — much less, only — understood in terms of consent. If more than mere consent is needed for permissible sex, then intuitions about (im)permissible sex may not translate into observations about the (in)validity of consent. For some discussion of suggested other criteria for the permissibility of sex, see Kukla (2021). ⮭
- Again, see Kukla (2021) and the references there. ⮭
- Millum (2014, 117) and the references there. ⮭
- Based on Millum (2014, 119) and on Gerver’s (2021, 260) “Coerced Sex” example. The case is also analogous to Dougherty’s (2021a) “Third Party Acquiescence”. ⮭
- The relevant literature on this is, of course, huge. For a helpful survey, see Kukla (2021, 277–280). For a recent discussion of third-party coercion centrally motivated by such examples, see Tadros (2021). ⮭
- Does she, though? What if she consents because of desires that were partly shaped by the patriarchal order? This raises questions not discussed in this paper, about what lies upstream from consent — in this case, what desires or preferences cause it, and indeed what causes them. I discuss these in my “False Consciousness for Liberals” (2020). Especially relevant here is the discussion of the possibility of making a problematically shaped desire or preference one’s own (2020, 190–191). ⮭
- Again, see my discussion of bargaining-with-the-patriarchy cases (2020, and the references there). ⮭
- For helpful discussions and comments on previous versions, I thank Larry Alexander, Scott Anderson, Dani Attas, Michelle Dempsey, Ezequiel Monti, Ittay Nissan-Rozen, Saul Smilansky, Ryan Stringer, Victor Tadros, Alec Walen, two referees and the editor of Philosophers’ Imprint, and audiences at the Rutgers colloquium and the Legal Philosophy Trident Seminar Series. This paper was also given as one of the Burman Lectures at Umeå in May 2023, and I thank the participants for the excellent discussion that followed. This research was supported by the Israel Science Foundation (Grant No. 1236/21). ⮭
References
Larry Alexander (1996), “The Moral Magic of Consent (II)”, Legal Theory 2: 165–174.
Mitchell N. Berman (2002), “The Normative Functions of Coercion Claims”, Legal Theory 8: 45–89.
Michelle Madden Dempsey (2023), “The Normative Force of Consent in Moral, Political, and Legal Perspective”, in Tatjana Hörnle (ed.), Sexual Assault: Law Reform in a Comparative Perspective. Oxford University Press.
Tom Dougherty (2021a), “Why Does Duress Undermine Consent?”, Noûs 55: 317–333.
Tom Dougherty (2021b), “Sexual Misconduct on a Scale: Gravity, Coercion, and Consent”, Ethics 131: 319–344.
David Enoch (2017), “Hypothetical Consent and the Value(s) of Autonomy”, Ethics 128: 6–36.
David Enoch (2018), “Against Utopianism: Noncompliance and Multiple Agents”, Philosophers’ Imprint 18.
David Enoch (2020), “False Consciousness for Liberals, Part I: Consent, Autonomy, and Adaptive Preferences”, Philosophical Review 129: 159–210.
David Enoch (2022), “Autonomy as Non-Alienation, Autonomy as Sovereignty, and Politics”, The Journal of Political Philosophy 30: 143–165.
Mollie Gerver (2021), “Consent and Third-Party Coercion”, Ethics 131: 246–269.
Heidi Hurd (1996), “The Moral Magic of Consent”, Legal Theory 2: 121–46.
Maximilian Kiener (2022), “Consenting Under Third-Party Coercion”, The Journal of Moral Philosophy 19: 361–389.
Quill R. Kukla (2021), “A Nonideal Theory of Sexual Consent”, Ethics 131: 270–292.
Hallie Liberto (2021), “Coercion, Consent, and the Mechanistic Question”, Ethics 131: 210–245.
Joseph Millum (2014), “Consent Under Pressure: The Puzzle of Third Party Coercion”, Ethical Theory and Moral Practice 17: 113–127.
Robert Nozick (1969), “Coercion”, in Sidney Morgenbesser, Patrick Suppes, and Morton White (eds.), Philosophy, Science, and Method: Essays in Honor of Ernest Nagel: 440–472. St. Martin’s Press.
Theron Pummer (2022), “Contrastive Consent and Secondary Permissibility”, Philosophy and Phenomenological Research 106: 677–691.
Jonathan Schaffer (2005a), “Contrastive Knowledge”, in Tamar Szabo Gendler and John Hawthorne (eds.), Oxford Studies in Epistemology Volume 1: 235–271. Oxford University Press.
Jonathan Schaffer (2005b), “Contrastive Causation”, The Philosophical Review 114: 327–358.
Justin Snedegar (2014), “Contrastive Reasons and Promotion”, Ethics 125: 39–63.
Stoljar, Natalie. 2014. “Autonomy and Adaptive Preference Formation”, In Andrea Veltman and Mark Piper (eds.), Autonomy, Oppression, and Gender: 227–252. Oxford University Press.
Victor Tadros (2016), “Permissibility in a World of Wrongdoing”, Philosophy and Public Affairs 44: 101–132.
Victor Tadros (2020), To Do, To Die, To Reason Why: Individual Ethics in War, Oxford: Oxford University Press.
Victor Tadros (2021), “Consent to Sex in an Unjust World”, Ethics 131: 293–318.
Stephen J. White (2017), “On the Moral Objection to Coercion”, Philosophy and Public Affairs 45: 199–231.