Preference utilitarianism differs from traditional utilitarianism in that it doesn’t enjoin us to maximize any sort of commodity such as pleasure or happiness. The rather murky concept of “utility” has no place in preference utilitarianism.
Instead, as Peter Singer lucidly put it, preference utilitarianism is the “minimum moral position”. To act morally, an agent should simply respect other agents’ preferences in the same way as he trivially respects his own preferences. To put it another way, preference utilitarianism enjoins us to “respect agency in general”.
Whatever we do, we do it because our beliefs and desires cause our action. Respecting agency means giving due deference to the beliefs and desires that cause an agent to act. At the moment of acting, given the constraints of circumstances, limited options, limited time, and so on, what we actually do is what we most want to do, given what we believe. If our wants (desires) were aimed at different goals or had different strengths, or if our beliefs differed in content or degree of entrenchment, we might act differently — but we would still do what we prefer to do. As long as we are genuinely acting, our preferences always “win”.
But let’s take care to consider “scope” here. If I hand over my wallet to a mugger rather than risk death, within the narrow context of being mugged I do what I “prefer”. But considered from the wider perspective of me going about my business, I would prefer not to be mugged at all. So obviously, although we always do what we prefer in a trivial sense looked at from the narrowest perspective, we don’t always manage to do what we want in a larger sense. In other words, our preferences are often thwarted; we are often unfree. Preference utilitarianism says that to act morally, we should as far as possible prevent the thwarting of others agents’ preferences, considered from the widest perspective, ideally taking account of the entirety of the agent’s beliefs and desires.
Like other forms of genuine utilitarianism, preference utilitarianism considers each act individually, in its particular circumstances, rather than promoting a general prescription or rule. However, a fairly close approximation to such a rule is the so-called Golden Rule expressed in Matthew 7:12:
all things whatsoever ye would that men should do to you, do ye even so to them
In other words, treat others as you would like to be treated yourself. Respect the agency of others as you respect your own agency.
But what if an agent acts with false beliefs? Wouldn’t we want others to prevent us walking out in front of an approaching bus? Or to whip out of our reach the glass of acid that we mistakenly think is a cool beer?
Indeed we would, and we should do the same for others, as long as — and this is absolutely vital — we do not overrule their overall agency in so doing. By “overruling their agency” I mean neglecting to give their beliefs and desires due deference as their own beliefs and desires, and not, please note, as true beliefs or as acceptable desires. When I grab someone to prevent them walking out in front of a bus, I assume that they have a whole bunch of other mental states such as a desire to not to be run over by a bus, a desire to reach the other side of the road uninjured, that they believe what they want can be found on the other side of the road, and so on. By overruling their belief that there is no immediate danger in stepping out into the road, I respect the many more other beliefs and desires that cause them to cross the road in the first place. Overall, I respect their agency more by overruling one belief while respecting the larger whole of the rest of their beliefs and desires.
Much the same goes for keeping a glass of acid out of the reach of a thirsty beer-drinker. It is reasonably safe to assume that we respect his overall beliefs and desires more by thwarting his narrower desire to drink the contents of this particular glass.
I’d like to emphasize, again, that these infringements are justified by respect for the agent’s overall agency considered as a larger whole, and not because they are caused by false beliefs.
As with the example of getting mugged, above, much depends on “scope” here. Within the narrow context of stepping out onto the road, or of drinking the contents of this glass, it might look like prevention means agency is disrespected. But from the wider perspective of the agent’s overall beliefs and desires, we respect agency more — we give due deference to the agent’s own beliefs and desires — by preventing this or that particular act.
In these spur-of-the-moment cases of preventing action, we rely on our psychological assumptions being correct. We assume that most people don’t want to be hit by a bus, don’t want to drink acid, and so on. We assume they believe buses and acid can kill. As with any assumption, we might conceivably be wrong, but in any case, we can easily check afterwards: we simply bring the bus or the acid to the agent’s attention. In most cases, the agent will thank us for keeping an eye out for his safety. But it might turn out that he has unusual beliefs or desires. He might doubt the presence of the bus, or strenuously resist the idea that the glass contains anything dangerous. He might be practicing his daredevil skills, or hoping to take his daily dose of vitamin C. Or he might be suicidal. In such cases, I think the preference utilitarian should respect overall agency. He should try to consider the agent’s entire system of beliefs and desires, and respect them as much as possible. If the agent persists in his beliefs and insists on his course of action, we should respect that. It might entail letting him go ahead and kill himself — or allowing him to take great risks in performing some sort of experiment in living.
As an aid to thinking about such scenarios, we can use the Golden Rule above as an approximation, and ask: What we would want others to prevent us doing ourselves? I think most of us would want others to pull us out of the way of a bus, or to grab the acid before we can drink it. We’d thank them for it. But we would not want them to overrule our overall agency, nor would we thank them for doing so. No one has the moral or intellectual superiority to legitimately thwart an act simply because they think it’s caused by an improper desire or a false belief.