The source of preferences and the effectiveness of consent
There are many things we can do that switch the moral facts. One of the most striking examples is consent. Usually, if I consent to you taking my money, I switch the moral facts: where before it was morally impermissible for you to take it, it is now morally permissible; and where before it might have been morally permissible for some third party such as the state to intervene to prevent you from taking it, now it is morally impermissible. But giving explicit consent isn’t the only way in which I can make the moral facts switch like this. Usually, if I promise you that money, I switch the moral facts in the same way; similarly if I request you take it, if I offer it to you, if I agree to doing something that inevitably and foreseeably requires you to take it, or as John Gardner argues, if we have embarked on some joint project together, and that project inevitably and foreseeably requires you to take it. Of course, you might say that, in each of those cases, you do something that thereby gives consent, and so it is in fact consent that’s crucial in all cases. That’s fine; it’s just a terminological point. I just want to be clear at the outset that there are many things we can do beyond giving explicit, stated consent that usually switches the moral facts in the way that giving explicit, stated consent usually does.
Why only usually? Well, even explicit, stated consent does not switch the moral facts if it is coerced, for instance. That is, if someone removes the possibility of not giving consent, or places an enormous cost on not giving consent—physical harm, social exclusion, financial ruin—then the resulting consent does not switch the moral facts. It does not become morally permissible for you to take my money if I only said you could because you said you’d punch me in the nose if I didn’t. But there are other cases in which consent doesn’t effect its usual change on the moral landscape. For instance, if the consent is not coerced, but it flows from preferences that I have only because you’ve brainwashed me into having them, then again we probably want to say it doesn’t make it permissible to take my money. And there are more nuanced cases, which come closer to those that really motivate this post: cases in which the preferences on which the consent is based—the preferences I have that lead me to give consent—are in some sense not my own; they are the result of the unjust society in which I live, perhaps, or the oppressive relationship in which you have ensnared me. If I consent to you taking my money, and my consent is genuinely based on a real preference I have, but I have that real preference because of external pressures and not because it is authentically mine, then again, it seems the consent doesn’t have the effect it usually has—it is not permissible for you to take my money, and it is permissible for a third party to prevent you from taking it.
The task of the post is to say when consent switches the moral facts in the way it usually does; and the same for the other sorts of action I listed at the beginning. What must the preferences on which that consent is based be like, and how must they have been formed, in order for the resulting consent to be effective? And ditto for promising, offering, agreeing, and so on.
It is worth noting another complexity before we begin. Usually, in the ideal cases, consent does two things: (i) it makes permissible an action that was impermissible—the act to which consent is given; and (ii) makes impermissible an action that was permissible—any intervention to prevent the action to which consent is given. But we might think that, in these non-ideal cases, these two switches can come apart: perhaps there are some cases of consent to X that fail to make X permissible but nonetheless make it impermissible for a third party to try to prevent X; and perhaps there are some cases of consent to X that make X permissible but it nonetheless remains permissible to try to prevent it.
What’s more, sometimes, in these non-ideal cases, consent to X might not make X alone permissible, but the consent does change the moral facts so that there are further things one can do such that doing them as well as doing X becomes permissible where doing X or doing this combination wasn’t permissible before. Quill Kukla gives examples in which sexual consent that is based on preferences that were formed in not wholly autonomous ways might permit things to go ahead, but only so long as the other partner provides sufficient scaffolding for the agency of the one whose consent is non-ideal in this way.
So the landscape here is complex, but hopefully we can map some of its contours. It might help to begin with some examples that other philosophers have proposed. Many come from the literature on adaptive preferences or false consciousness, but I think it’s important to look beyond this, since definitions of adaptive preferences can be very narrow, and we want to cast our net wider. I should say up front that I don’t necessarily agree with the judgments about permissibility in all of these cases; and that will be become apparent later, when we look at the finer details. But I think it’s helpful to have some examples of where philosophers have tended to think consent or other acts like it fails to switch moral facts in the usual way.
In her ‘Defending Prostitution’, Carole Pateman writes:
[A]s an examination of consent and rape makes graphically clear, throughout the whole of sexual life domination, subjection, and enforced submission are confused with consent, free association, and the reciprocal fulfillment of mutual desire. The assertion that prostitution is no more than an example of a free contract between equal individuals in the market is another illustration of the presentation of submission as freedom. Feminists have often argued that what is fundamentally at issue in relations between women and men is not sex but power. But, in the present circumstances of our sexual lives, it is not possible to separate power from sex. The expression of sexuality and what it means to be feminine and a woman, or masculine and a man, is developed within, and intricately bound up with, relations of domination and subordination.
I take it that part of Pateman’s point here is that, if women are raised in a society in which they are told that part of what it is to be a woman is to be submissive towards men, and they are, through other means, made to consider that identity very important to them, then consent to sex work might not be genuine—that is, it might not have the power to make the practice of sex work permissible and intervention to prevent it impermissible. Catharine MacKinnon suggests something similar might be true of all heterosexual sex in a highly patriarchal society.
Martha Nussbaum draws on the case of a group of women living in a Bangladeshi village, as described by Martha Chen in A Quiet Revolution:
These women had low status in their community, in every area. They were less well-nourished than males, less educated, less respected. Let us consider their situation where just one question, the question of literacy, is concerned. A desire-based approach […] argues that if these women do not demand more education and a higher rate of literacy, there is no reason at all why government should concern itself with this issue. Polls are taken; women express satisfaction with their educational status; no further effort is made. [T]his approach looks very short-sighted once one considers the weight of the cultural forces pressing these women not to demand more education (and also not to feel that they want more); once one considers, as well, the absence, in their daily lives, of paradigms of what education could do and be in lives similar to theirs, the absence of any experience of the alternative they are asked to consider. Sometimes the combination of ignorance and cultural pressures actually prevents the formation of the desire for education.
So the fact, if it is a fact, that these women would agree to a situation in which they and other girls and women in their community continued to be excluded from literacy education does not make it legitimate for the government to continue to exclude them. This suggests that, because of the way they were formed, their preferences, which include no desire for education, cannot give rise to genuine consent that makes permissible an inegalitarian policy.
Natalie Stoljar and David Enoch describe a case that Serene Khader discusses. Here is Enoch’s presentation; I’ll return to Khader’s below, since it includes crucial further detail.
A young woman grows up in a highly conservative society […] where it is common for women to malnourish or even starve themselves in order to better feed (well beyond need) their husbands and male children. And she chooses to malnourish herself in this way. Gender norms play a role here, of course, but at the end of the day, the woman prefers to have her husband eat more than enough rather than have enough for herself.
As in Nussbaum’s case, we’re invited to conclude that the preferences here are genuine—the woman is not simply saying she prefers this situation; she genuinely does. But the preferences are formed in such a way that we don’t think that consent based on them has the power to switch the moral facts.
And we might include one final one of our own:
Marriage Equality A lesbian woman living in the UK in the mid-1980s, in the era of Section 28 and well before marriage equality was even a genuine possibility, who prefers not to be married, and forms that preference precisely because it looks like it will never be possible for her to marry the person she loves.
Khader’s and Nussbaum’s cases sound initially like classic cases of adaptive preferences, as does our imagined case of the lesbian woman. On the standard definition, a preference in favour of X over Y is adaptive if it is formed precisely because X is available and Y is not. MacKinnon’s case might be like this as well, but it’s less clear. Perhaps Pateman’s case is as well. But again, it’s not clear that the preference on the basis of which the women consent to sex work is formed precisely because other possibilities aren’t available. It seems rather that the preference is a consequence of a whole value system that the person has taken on as their own because it is part of the dominant value system in the country. In any case, the point is that it is not only when the preferences on which consent is based are adaptive that its effectiveness is compromised.
So, when do preferences give rise to genuine consent? A natural answer based on the examples given so far is that a necessary condition for genuine consent is that the preferences on which it is based are not formed in conditions of oppression: that would cover Pateman’s cases, MacKinnon’s, Nussbaum’s, Khader’s, and our example of the lesbian woman who disfavours marriage. And indeed John D. Walker and David Enoch suggest this. Walker’s view is nuanced in just the way I described above: he thinks that consent based on preferences formed in response to oppressive social situations can be effective in some ways, but not in others; it can make it inappropriate for the state to intervene in private affairs involving the consent, but it doesn’t necessarily make it inappropriate for the state to do something that has a bearing on the sort of private situation in which such consent is given. Enoch says:
My suggestion, then, is that an important class of cases of nonautonomous preferences is those that were shaped (in the appropriate way) under the causal influence of unjust conditions, conditions that violate the rights or entitlements of the relevant agent. […] The causal role played by injustice here explains why it is that these preferences (or consent or choice based on them) do not manifest the value of autonomy (as nonalienation), for it shows an important sense in which these preference are not truly the agent’s.
And he goes on to clarify that the injustice in question must be injustice towards the person with the preferences: so, if I live in a society that treats me justly but others unjustly, and I form a preference in response to this situation, that doesn’t prevent it from giving rise to effective consent. And, what’s more, the causal pathway from the injustice to the preference must be of the right sort: “the shaping of the preference has to be sensitive to the unjust circumstances being, well, unjust.” And he gives an example of someone who forms a preference for playing the piano because they are kidnapped and held in a room with only a piano for entertainment—the preference is caused by the injustice, but it can ground effective consent because the shaping is not sensitive to the injustice of the cause.
I don’t think Enoch’s account can be quite right. I raised this point in a previous post, where I offered an example like this:
Lone activist A gay activist campaigning for LGBTQ+ rights, who greatly values addressing political problems collectively. However, he comes to live under a homophobic, unjust regime that criminalises collective action, particularly targeting queer liberation groups. It places restrictions on freedom of association, lobbying, protesting, etc., in ways that affect them disproportionately. As a result, he turns to more individualistic approaches, trying to influence individual people in positions of power. Over time, he comes to value that sort of approach; indeed, he comes to favour it over more communal approaches.
The activist’s preferences are formed under an unjust regime that is unjust to him specifically. And while Enoch’s requirement that the preferences are sensitive to the injustice of the situation under which they’re formed is vague—sensitive in what way?—surely this case satisfies it, if any does. We might even add to the story that the oppressive regime introduced these measures specifically to target LGBTQ+ activists, and their explicit goal was to lead any such activist to favour a more individualistic approach, because the regime believes that is less effective and will threaten them less. And yet I think we wouldn’t worry that any consent given on the basis of the activist’s new preferences was ineffective.
Another case, which I think is trickier, but nonetheless should give us pause:
Pressured Parent Someone is pressured very heavily by their partner, family, and society to become a parent when their initial preferences are strongly against doing so. However, becoming a parent proves a personally transformative experience for them and their preferences change so that they value raising a child greatly. They are now considering having a second child, and they want to do so based on these preferences.
Again, it seems that the new preferences are formed under pressure of injustice, and in a way that is perfectly sensitive to the injustice. And yet, when this person decides that they would like to have a second child, it seems unlikely we would say they aren’t able to consent to that, or that any stated consent they give is ineffective at switching the moral facts.
In this post, I want to argue that, if we are to account for cases like this, we need to look in much more detail at the way in which the preferences are formed. It seems intuitively plausible that people’s preferences are formed and changed by the circumstances in which they find themselves, and oppressive or unjust circumstances are surely no exception. And indeed there is evidence that certain changes in your life are likely to lead you to shift your preferences: Anat Bardi and her collaborators offer evidence that, while reading for a particular degree or training for a particular job is unlikely to shift preferences, moving to a country with significantly different cultural norms is likely to do so. But we think less often about the different ways in which this might happen. What’s the mechanism? Because preferences, like beliefs, are sticky. They tend not to change quickly and for no reason. And, I want to argue, when we do attend to these mechanisms, we see that it makes a difference to the effectiveness of any consent based on those preferences; how they’re formed determines whether that consent switches the moral facts and in what way. This in turn raises a problem, however, for we rarely have access to the mechanism by which a person’s preference was formed; indeed, the person themselves rarely enjoys that access. And so we face a situation of widespread ignorance about the demands of morality. But we’ll come to that in due course.
There are doubtless myriad ways in which preferences are formed. But I’d like to focus on four of them.
Mechanism 1: Sanctions. Preferences can be formed, or changed, because the price of not having them is too high. Suppose, for instance, that you’ll be very heavily sanctioned for behaving in ways that could only be grounded in different preferences. Then of course it’s open to you to have those different preferences and simply to pretend to have the target ones. But that is exhausting; it requires constant vigilance over your own behaviour, which will not be possible for everyone; and even those who are best at it are likely to let their mask slip at some point and then receive the sanctions. And so it is easier and safer to simply acquire the preferences.
It’s plausible that Sanctions is the mechanism at play in many patriarchal societies, such as a Victorian England, where women whose behaviour reveals preferences in tension with the patriarchal norms are often sanctioned through domestic violence, and they are also ostracized by friends and family, and excluded from the possibility of marriage, which can be the only way in which a woman can access money and shelter. It’s also possible that this is at play in Nussbaum’s example of the women with no preference for education.
Mechanism 2: Beliefs. Preferences can be formed, or changed, by gaining beliefs about what is objectively preferable. In many, though certainly not all cases, a belief that X is objectively preferable to Y leads to a preference in favour of X over Y. In any case, it can do. This means, of course, that someone’s preferences can be manipulated by controlling the evidence they receive, the reasoning to which they are exposed, and the training they receive in thinking things through for themselves—indeed, Paulo Freire’s educational proposals are all based around the need to evade or counteract such control.
For at least one woman in the situation that Khader described above, where a woman intentionally and knowingly starves herself to overfeed her husband and children, this is the mechanism by which the preference is formed. It’s likely the mechanism in Pateman’s example as well.
Hanna Papanek describes the networks of beliefs that sustain gender-biased intrahousehold food distribution in a particular community in Java. An elderly Javanese woman recalled being told as a child that women needed to discipline themselves, because they were superior to men who could not control themselves. Papanek adds that this idea could be easily supported by a web of religious beliefs encouraging self-deprivation as a spiritual discipline. Complicated and reflective beliefs may sustain preferences that intuitively appear to be adaptive.
Mechanism 3: Attention. Preferences can be formed, or changed, by having extended contact with the object of the preference, if that extended contact results in you attending to those features of it that lead you to value it (though of course this can also result in you attending to features that lead you to disvalue it, and that would give rise to a preference against it).
I suspect we are all familiar with this way of forming our preferences. It might be the way we become fond of someone who has just entered our lives and with whom circumstances requires us to spend a good deal of time: a co-worker, a neighbour, a fellow volunteer with a local charity. When we first meet, we see some of the more obvious contours of their character: are they outgoing or shy? are they patient or irascible? and so on. And we might form an initial preference concerning them based on that. But, as time goes on, many of those bigger, more obvious features become less important as we have been put in a situation in which we observe and interact with less obvious parts of their character. And that can lead us to change our preferences concerning them, or it can reinforce the ones we already had.
At the beginning of Painting as an Art, Richard Wollheim talks about how he comes to appreciate a work of art, and I think he is describing something like what I’m talking about here:
For I came to recognise that it often took the first hour or so in front of a painting for stray associations or motivated misperceptions to settle down, and it was only then, with the same amount of time or more to spend looking at it, that the picture could be relied upon to disclose itself as it was.
This mechanism is most familiar as a way of coming to appreciate a particular item, be it an object, like a painting, or a person, like a co-worker or neighbour. But I think it is also the mechanism that drives our preferences for different ways of life. I move to a new city or country or village, or I join a new online community, or I find myself in a particular sort of romantic relationship, and I initially notice its bigger, more obvious features; but gradually, as I live with it and in it longer, I come to see the smaller, more detailed features, and I can come to value those and acquire a preference for the way of life I’m inhabiting.
It’s plausible that this is the mechanism in the examples of Lone Activist, Pressured Parent, and Marriage Equality from above. In each case, the protagonist’s circumstances put them into extended contact with particular ways of life: the life of the lone activist, the life of the parent, the unmarried life. And in each case that extended contact reveals to the protagonist detailed features of that way of life that are perhaps not visible from the outside, or are visible but that require extended contact with them to come to value them.
I suspect that this is the source of the way of valuing things to which G. A. Cohen draws attention in his ‘Rescuing Conservatism’. There he points out that we often move from having a preference for having some thing of a certain sort to having a preference for the particular thing of that sort with which we’ve formed an attachment—he gives the example of romantic relationships, but we might also think of jobs, friends, and even quite ordinary objects like cars, to which people become attached. I suspect that Attention is the mechanism by which that change is often effected. But there is another mechanism that is often in play, and I turn to that now.
Mechanism 4: Association. Preferences can be formed, or changed, by associating other things of value with them, or by ensuring that other things will lose value if you don’t have the preferences.
I might value a car because of the wonderful nature reserves it allowed me to visit, the mountains it allowed me to climb, the family it allowed me to see more regularly. And, on the flip side, I might value a romantic relationship in part because, if I were not to value it, that attitude would infect my attitude towards many other things that I currently value greatly, and might even lead me to think of much of my life as lacking value. I spent so much of my life with this person; we shared so many of its most important moments. To come to think of them negatively or even just neutrally can rob that life and those moments of much value.
It’s plausible that this mechanism might be at work in Pressured Parent as well.
Let’s leave it there. We have four mechanisms now. There are certainly more. But thinking about these will help us see how sensitive the effectiveness of consent is to the ways in which the preferences on which it’s based are formed. My contention is this: typically, when preferences are shaped under pressure from an unjust situation, if they are formed via the first two mechanisms—Sanctions and Evidence—that weakens the effectiveness of any consent based on them, but if they are formed via the second two—Attention and Association—that has no impact on the effectiveness of the consent to which they give rise. And this is why, when the Bangladeshi women in the situation Nussbaum draws from Martha Chen’s description consent to their exclusion from education, there is reason to worry that the consent cannot switch the moral facts in the way it usually does. And it is why, when the pressured parent wishes to have a second child, or the activist continues to work alone even after a new government comes to power and removes the restrictions on freedom of association, etc., their consent to things that are necessary to pursue these goals is effective.
Before I wrap up, let me discuss a question that Fabienne Peter raised after I gave a talk in Warwick about this proposal. We know that your consent is not effective when it is coerced—if I consent only because you threaten to punch me if I don’t, it isn’t effective. And many think that your consent is not effective when it is based on beliefs you acquired because of someone deceiving you—if I consent to give you my money because you led me to believe you’d use it to help a charity, but in fact you spend it on your bathroom refurbishment, many think the consent is not effective. Perhaps, when you consent to something based on preferences formed by Sanctions, you are being indirectly coerced into giving that consent. After all, one effective way to make someone do what you want is to cause that behaviour directly; but another way is to cause them to have preferences that will, in turn, cause them to behave in that way. And you might think that, if your preferences are formed under threat of sanctions, they are essentially coerced, and that means that behaviour based on them is indirectly coerced. And similarly when you consent to something based on preferences formed by Evidence. In this case, your consent is not directly caused by a deception, but the preferences on which it’s based are, and that is an indirect form of causing consent by deception.
If this is correct, perhaps we don’t need a distinct account of which preferences can give rise to effective consent? Perhaps we just need an account of how consent itself is directly undermined, and then we just say that it is undermined indirectly when analogous forces apply to the formation of the preferences on which the consent is based?
I’m sure something like this is doing some of the work in driving our intuitions about these cases. But let me delve into this in a little more detail, because I think things are a bit complicated. Consider, for instance, two people who both wish to have a child. The first formed this preference via Sanctions: their society ostracises those who behave as if they have they don’t have this preference, and the most effective way to ensure you behave in a way that avoids sanction is to in fact have the preference. The second formed this preference via Attention: they already have a first child and their time as a parent has made them want to have a second; but they had the first child not because they then wanted to, but because their society ostracises people of their age who can have children but don’t. Now, if it were the case that consent is deprived of its efficacy whenever it is directly coerced, and also, to a lesser extent, when it is indirectly coerced, we’d expect our judgments about the effectiveness of the consent to have a second child that each of these people give to be the same. After all, both are indirectly coerced into having the preferences—the first by the threat of sanctions against contra-preferential behaviour; the second by being coerced into having a child, which leads them to value the experience of parenthood. And we might assume that the chance that the original coercion would lead eventually to the consent is the same in both cases: suppose that 90% of people who feel the pressure of the ostracism in the first case respond by adopting the pro-child preferences, and 90% of people who become parents respond by adopting the pro-child preferences. So if the consent of the person in the first case is ineffective, so should be the consent of the person in the second case. But that isn’t so. We take the consent to be ineffective in the first, and effective in the second. The difference, I claim, is that the mechanism is Sanctions in the first case and Attention in the second.