Feinberg on Offense Notes for April 5

Main points

Well, that was certainly interesting! Thank you very much for your patience and general good nature.

According to Mill’s harm principle, the state is permitted to interfere with an individual’s liberty only if doing so is necessary “to prevent harm to others” (p. 267).

We can quibble about what is included in preventing harm to others. For instance, we can debate about:

But let’s put the quibbles aside, especially when I’m losing.

There is a distinction between the merely offensive and the harmful. It’s not always obvious where the line is to be drawn, but there are plenty of cases that fall on either side of it. And, if you're still worried about how to draw the line between the harmful and the offensive, consider this. Sometimes, people argue for restrictions on liberty only on the grounds that the doing so would prevent offensive behavior; they do not even allege that it is harmful.

So suppose we have a proposal to limit liberty solely on the grounds that doing so would prevent offensive behavior. Mill’s harm principle rules the proposal out out. If the legal restriction passes, the harm principle could be used to criticize it for objectionably limiting liberty.

Feinberg’s contention is that this is too hasty. It is possible for the state to be fully justified in imposing criminal penalties for merely offensive behavior.

The problem

Will put his finger on the problem. Everyone finds something offensive. Does that show that Feinberg’s supplement to the harm principle, the offense principle, completely undermines the harm principle? It might seem to do so by allowing the state to regulate all kinds of behavior. The goal of the harm principle was to limit the state’s ability to regulate behavior by ruling most behavior out of bounds. The offense principle seems to reverse that by allowing the state to regulate almost everything.

Feinberg narrows the offense principle in several ways to limit this effect.

  1. The behavior has to be wrongful, that is, it has to violate someone's rights.
  2. The behavior has to cause universally disliked mental states such as unpleasant sensory experiences, disgust, revulsion, shock, shame, embarrassment, anxiety, and so on.
  3. The offense has to be serious.
  4. Criminal sanctions have to be both effective and necessary in stopping the offensive behavior.
  5. Punishments should be light, along the lines of a parking ticket.
  6. Finally, Feinberg requires that the interests of those who wish to avoid offensive behavior be weighed against the interests of those who wish to engage in it. The factors listed on pp. 282-3 are drawn from similar kinds of weighing in tort law. Each party’s inconvenience must be taken into account, compared with alternatives (taking a different bus or getting a room, respectively). The social value of each party’s request and each party’s motives are taken into account. And so on.


One thing I don’t understand is where the relevant rights come from. Our reading comes from a much larger work. It’s three volumes long, in fact. So I’m sure there’s an answer somewhere. But for our purposes, Feinberg’s use of rights or, equivalently, the idea that behavior might wrong someone, makes it hard to compare his offense principle to Mill’s harm principle.

Mill, after all, explicitly says that he won’t use any assumptions about rights. (Though it’s not clear that he stuck with this).

I suppose Feinberg might say that he has still raised an objection to Mill’s harm principle. He has shown that regulation of offensive behavior can be legitimate. Since Mill rules it all out, that’s all that Feinberg has to show.

What if we’re wrong?

Alex pointed out how what counts as offensive can vary over time. In addition, there are disagreements in our own time about what is offensive. I suspect that few of our fellow citizens would agree with the 19th century punishment in North Carolina for “the unmentionable crime against nature”. But many of them find TUCAN offensive, even when done behind closed doors. Most of us do not and some of us take offense at their offense.

Geez, offense is confusing.

Anyway, Alex’s point was that there’s no way of sorting out what counts as offensive per se. All we have is some conventional stipulation. This point tended to bleed into a second point. This was a defense of Mill along the following lines: since some people take offense at innocent behavior and there is no way of knowing what is “truly” offensive, the state should not be allowed to regulate offensive behavior.

In my opinion, the last point goes too far. Standards that are very important to us now might change in the future. And we might be wrong about just about anything. What is counted as harmful now might be regarded differently in another time. But it doesn’t follow that we shouldn’t regulate any kind of behavior.

When errors seem especially likely and costly to those whose liberty will be limited, we should be cautious. Both conditions apply to offensive behavior, as Feinberg’s examples on p. 280 show. So we should limit the state’s ability to punish offensive behavior. There are two ways of doing that. One is to prohibit the state from punishing merely offensive behavior at all. The other is to limit the state from punishing merely offensive behavior to a small range of clear cases.

We could say that we’re better off safe than sorry and allow everything Mill’s harm principle does. But I think we’d be even more sorry as a result. I certainly don’t ever want to have to take a bus ride like the one Story 23.

This page was written by Michael Green for Philosophy of Law, Philosophy 34, Spring 2007.
Name of website