Troubles for Tolerance

A little while ago I wrote a post about problems for the idea that it would be good if everyone were equal. That is an easy target for me, because I don’t think that it would be good if everyone were equal, in part because of those problems. Today, then, I want to look at something that’s a bit harder for me. I think that tolerance is a good idea, and I would like to see more of it. There are, however, some problems for the concept.

Let’s start with a rough characterisation of tolerance. A tolerant person allows people to live according to ethical and aspirational systems with which he does not agree. It is a central aspect of tolerance that you allow people to actually put the ethical and aspirational systems into practice; it is not tolerant to pretend to allow them to do something, but to take everything away afterwards, or undo all their work. Someone who removes graffiti as soon as it goes up is not tolerating graffiti.

Now, there is an obvious problem for tolerance, and one that everyone grapples with. This is the person who wants to go around killing people. You can’t tolerate that behaviour, the argument goes, because it infringes on the rights of others. I agree. Tolerance should not extend to tolerance of murder, and the problem of where to draw the line is a difficult one. Here, I want to suggest that it is even more difficult than most people are inclined to think.

It is common for tolerant people to believe that the line over what you should tolerate should be drawn in general terms. It should not deal with the specific details of any actions, such as whether they happen on a Sunday. Rather, it should deal with broad ideas, such as “any action that does not harm anyone who is not a voluntary participant in the action”. This is harder than you might think, but I want to suggest that it may be a complete non-starter. I will consider a couple of very minimally tolerant standards, and argue that accepting either has difficult consequences.

Tolerating Intolerance

First, consider the standard “You should tolerate any ethical system that is more restrictive than your own”. That is, if someone’s ethics forbid him to do things that your ethics permit, but do not permit him to do anything your ethics forbid, then you should tolerate those ethics. Similarly, if his ethics require him to do things that are permitted, but not compulsory, on your ethics, then you should tolerate those ethics.

This standard looks plausible because such a person would, it seems, not do anything that you would rule ethically unacceptable. What could be the problem in tolerating such a system?

Well, we’re looking for problems for liberals like me, so let’s take a concrete one. There are a significant number of people who believe that homosexuality is wrong. As a result, they believe that homosexuals should be criticised for their sexual activities, encouraged to seek counselling, and discouraged from engaging in such activities. Portrayals of such sexuality should be severely restricted in most media.

This really, really basic level of tolerance seems to require us to allow them to campaign for this. Let’s take each one in turn, and compare it to standard liberal ethics.

Criticising people for their behaviour, when you believe it is unethical, is permissible for liberals. They do it quite a lot. And, certainly, it is permissible to criticise these people as well. It is, similarly, permissible to provide counselling for homosexuals who don’t want to be, and even to encourage people you believe to have psychological problems to seek counselling.

It is permissible to campaign for unethical behaviour to be outlawed. Whether it is permissible to actually make the behaviour illegal will depend on the standards we finally establish for tolerance.

Finally, it is permissible to remove ethically dangerous material from the public media. Cigarette advertisements are restricted, and liberals put a lot of pressure on to keep homophobic portrayals out.

From a more general perspective, it is important to see that this is different from racism. Racism is based on someone’s skin colour, which is not regarded as ethically significant. However, a person’s sexual preferences and behaviour are regarded as ethically significant by almost everyone. Rape, for example, is generally regarded as wrong. Paedophilia is regarded as wrong as a preference. Indeed, most liberals would be perfectly happy to go along with all the suggested moves above, were they directed at paedophiles rather than homosexuals.

The key difference is simple. Most liberals believe that paedophilia is actually wrong, and homosexuality actually isn’t. However, the whole point of tolerance is that you allow people to disagree with you about what is wrong, with, in this case, the limitation that they can only be stricter than you.

So, the problem with this formulation is that it seems to require the tolerance of homophobia. More generally, it requires “tolerance of intolerance”. Most liberal ethical systems permit women to learn to read, but do not make it compulsory. Thus, it seems that you should tolerate a system that believes it is wrong for women to learn to read. (Or men, but historically and currently women are the targets of such prohibitions.)

There are two possible ways to get around this. One is to look at broader implications of the liberal ethics, arguing that making homosexuals feel guilty or excluded is not permitted. Therefore, we can tolerate systems that are simply stricter, but the homophobic system actually violates some of the ethical standards of liberalism.

I don’t think this can work. Any ethical system that has an application to more than one person licenses a negative attitude to people who violate its rules. (That’s a posh way of saying that it’s perfectly all right to think that rapists are bad.) If we claim to be tolerating stricter systems, as long as they do not license negative attitudes to people who violate the rules that are not in common with our system, then we are not really tolerating stricter systems of ethics; we are simply tolerating individuals who choose not to do some things that are permitted, but not required, by our ethics. But this is not tolerance. No matter how extreme you are, you have no problem with people choosing not to do things that are permitted, but not required, by your ethics. That is what “permitted, but not required” means.

A second approach is to say that it’s all right for people to think that homosexuals (or literate women) are bad, as long as they never act on those beliefs. The UK recently passed laws on roughly this basis, making it illegal for guest houses to refuse homosexual couples.

Again, I do not think that this is really tolerance. It’s like claiming to tolerate graffiti, while making it illegal to actually spray paint on walls. Claiming that you tolerate people thinking about graffiti is not really convincing.

The reason this is a difficult example for me is that I think that forbidding guest houses to turn homosexuals away is a good thing. So maybe we should reject this as a minimum standard of toleration; it’s harder than we thought. Obviously, we can’t just invert it; ethical systems that permit things we forbid include the ones that permit murder, and that’s just the classic set of problems. So, let’s try something different.

Uncontested Benefits

Consider this rule: Tolerate any system in which all the actions are agreed by the actor, all targets, and you to be good for the target, even when you believe it is ethically required to bestow the benefit on someone else.

Again, this seems fairly minimal. It does not tolerate people going around harming each other, or doing harm by your standards because they think it’s a benefit, or imposing benefits on people who think that they are harms. All you are tolerating is people being good to one person when you think that they ought to be good to someone else.

This is a very weak form of tolerance. It does, for example, permit the state to ban the teaching and promulgation of any and all religions, on the basis that such beliefs are actually harmful. Even if the teacher and the student believe that the teaching is beneficial, the state does not, so it does not fall under this definition. In short, a state could accept this principle, and still not look at all tolerant. On the other hand, it is easy to imagine that any state that was tolerant would, at the very least, accept this principle.

So, what are the problems?

First, there is one I have discussed earlier. This principle rules out any action against private education. The parents, the children, and the critics all agree that the education is beneficial to the children. Indeed, that is why the critics are critical; they say that it gives the children of the wealthy an unfair advantage. The critics argue that the benefit ought to be given to other people. However, that falls directly under this principle. It would be intolerant, more intolerant than banning Christianity, to ban private education.

But then, I’m in favour of private education, so this isn’t very difficult for me. There is a more difficult application of this principle.

Consider a racist, but hard-headed, employer. When he has a vacancy, he always fills it if there is a qualified applicant, but never hires someone who is not qualified. However, if there are qualified black candidates and qualified white candidates, he always hires the best-qualified black candidate. He only hires white people if no qualified black people apply, even if the best-qualified black person is only barely qualified, and the least-qualified white person is supremely suited to the job.

Any laws to stop him doing this fall foul of this principle of tolerance. The job is a benefit. The employer thinks so, the people applying for it clearly think so (anyone who changes his mind can rule himself out), and society at large thinks so, which is why it would like to stop such practices. Society thinks that the job should go to the best-qualified candidate, independent of race, but, again, that is a pure-case description of the principle as described. Society thinks that the benefit ought to be conferred on someone else.

This obviously applies to all forms of discrimination; sexism and nepotism are also covered. So, this is why it is difficult for me. I think that laws to promote equal opportunity based on race and sex are a good thing, and that, where they exist, they have had a good effect on society. But they appear to be so intolerant that they make the relentless persecution of religion look minor, at least in principle.

Now, you could reject the final clause of the principle, and say that you can only choose where to give a benefit when good liberals think it doesn’t matter. However, this, again, is not tolerance. It’s saying that you have to follow liberal ethics.

It’s important to see what the principle doesn’t say. It doesn’t say that affirmative action is wrong. Indeed, it specifically defends affirmative action, because that involves hiring black people even when there is a better-qualified white candidate. It doesn’t stop the government passing laws to require government bodies to hire purely on merit; organisations can set their own rules for hiring. Finally, it does not, by itself, allow anyone to simply deny everyone a benefit rather than giving it to a white person. Thus, while it would allow a hotelier to give a room to a heterosexual couple rather than a homosexual couple, it would not allow the hotelier to simply turn the homosexuals away.

As the last example shows, however, it is quite broad. It allows unequal pay. As long as you pay women at least the minimum wage, you can pay men more for the same work. It allows unfair service. You can always serve white people before black people, even interrupting serving a black customer to serve a white one. For clubs with a limited membership, you can restrict that membership to white people.

In short, it supports much of the settlement in the pre-Civil Rights American South. (Note, pre-Civil Rights, not pre-Civil War. It doesn’t justify slavery.)

I am strongly inclined not to tolerate such a settlement. I think it’s deeply wrong. But to reject it, I have to reject this principle of tolerance as well.

“Tolerate Things You Like”

The problem is that, if I reject both of the principles above, I can’t see any general principle of tolerance that I could accept. The principle mentioned at the beginning, “Tolerating people who do no harm to anyone who is not a voluntary participant”, has the second principle as a special case, and thus faces all of its problems. Tolerating people doing harm to other people is even more difficult; it might suggest that we should have tolerated the lynchings in the South.

I don’t have a solution to this problem. I still think that tolerance is a good principle, but I do not like the consequences of the concrete versions of the principle I have come across. There are two ways to handle this. One is to reject tolerance. It might be just superficially appealing, but the things that are genuinely good about it might be best captured by some other ethical principle. The other is to bite the bullet, and to say that we should tolerate nepotism or racist hiring practices. Right now, I don’t know which way I want to go.