Why not be immoral?

This semester I have been sitting in on a course on Ethics. Never having previously studied it on a formal basis, I thought it would be worth going along to the lectures in order to see if there are important issues that I am missing. From that perspective, the course has not been a disappointment: I had hitherto overlooked the thorny issue of why people should act morally.

From another perspective, the course has been a disappointment - the proposed answers to this, and the arguments for them, have been remarkably weak. There have been two answers proposed for this question:

1. It is in one's interest to act morally
2. A so-called "rationalist" argument, which I shall outline below

The argument from self-interest seems in on sense to be question begging. Why exactly would one act immorally if it were not in one's interest? That said, perhaps we could argue that there is more to it than this. Humans are not good at judging risk, so perhaps a case is to be made that we tend to err in our judgements of our self-interest, and that acting morally is a good corrective.

This clearly doesn't work as an argument in relation to utilitarianism, where one can frequently be required to sacrifice one's own wellbeing for the greater increase in wellbeing of others. It's hard to see how donating vast amounts of money to the third world is in one's own interest, and yet according to one of the most powerful essays on applied ethics of the last century it is pretty much obligatory. Similarly, any account of virtue ethics which includes a component of "civic virtue" - the propensity to sacrifice one's own interests for the good of one's community - will pretty much by definition fail to consistently serve one's own self-interest. Deontological systems don't necessarily fail to consistently advance one's own self-interest, but then again it would be extraordinarily rare to use a purely deontological system to establish one's view of the Good: usually, deontology acts as a side-constraint upon another system, whether that is virtue ethics, utilitarianism, enlightened egoism, or any other of the myriad moral systems we have developed.

Here is a rough reconstruction of the rationalist argument, taken directly from the lecture slides:

1. What is Right is what is good.
2. What is good is what we ought to do.
3. What we ought to do is what we should do.
4. What we should do is what we have reason to do.

Conclusion: We have reason to do what is right.

There are a couple of crucial problems with this argument. First, there is a problem which was discussed in the lecture: what does it mean to have reason to do something? It means that it is rational to do it. What does it mean for it to be rational to do something? Here lies a thorny debate between those who believe it means that one should want to do something regardless of one's interests, and those who believe it implies a path between doing the thing and achieving one's goals. Since I view the second view as obviously the correct one, the argument fails here. However, I would go further: I would charge that this argument is deliberately misleading and sophist.

One of the key tenets of Analytic Philosophy is being precise and clear in one's meaning. The above argument is not at all clear currently. There is a simple test to see whether an argument relies on the abuse of words: that is, to precisely define every word which is not absolutely obvious in meaning, and then to substitute in the definition for the word, wherever it appears in the argument. I propose the following definitions:

Ought: either that which someone has a duty to do, or that which is suberogatory.

Should: that which it would be advisable for a person to do, given their current set of desires and probable future desires.

I won't go over the whole argument like this; instead, I shall focus upon premise 3. Using the definitions I have provided, it becomes clear that this premise is flawed, and relies upon assuming a desire to act morally. We can try redefining the terms to get around this, but we are inevitably going to face this disconnect somewhere between premises 2 and 4. In ordinary language usage "ought" and "should" feel very much alike due to the ambiguity of the word "should"; however, when rendered in precise, logical language the difference in meaning becomes clear.

I'm not content to suggest that these arguments are not obviously true, though: I want to knock them down completely. For this reason I propose a simple test for the strength of any particular argument as to why a person should act morally:

Suppose a global dictator is to be appointed. This person is to be chosen completely at random out of the set of mentally competent adults of at least average intelligence, and is to have absolute power over the rest of humanity. There shall be no possible way of resisting the dictator's will or of deposing them. Would you rely on this argument in order to keep that person behaving benevolently?

I doubt anyone would be comfortable assuming that it was in a dictator's own self-interest to act entirely morally and honourably (unless they believe in an afterlife, and we prefer to avoid making that kind of assumption). And you can go on all you like about it being "rational" to not maintain a harem of supermodels and to avoid nepotism, but I doubt the average dictator gives more thought to maintaining a veneer of philosophical "rationality" than he does to propagating his genes. Indeed, I doubt that there is any real reason for this absolute dictator to hold back once he is in this position; instead, one would have to act beforehand so as to restrict his power and free will.

There is an objection to this line of argument I can imagine being raised, but which misses the point. This objection is that there is a massive difference between acting morally when one is a near-omnipotent dictator, and when one is an ordinary member of a human society. If a normal person takes up stealing, then they will be arrested and fined or imprisoned. If they start insulting the people around them, then they will face social rejection.

The problem with this objection is that in these cases, it's not morality that we are responding to: it is social incentives. The dictator example deliberately removes all of these other restraints; relying on social incentives is at best a flawed method of achieving good behaviour, for it does not punish many kinds of bad behaviour (e.g. persistence of various hazing rituals, state brutality, and if you believe in it then rape culture) while punishing all kinds of things which are morally neutral or even virtuous (e.g. public cross-dressing, professing and acting based on scepticism of ostensibly well-intentioned but ultimately ineffective or worse). However, this does point in one useful direction, because we have the power to change social incentives.

I suspect that it is not in the interests of the individual to act morally. However, it is in the interests of the individual that other individuals act morally. Therefore, ceteris paribus it is in the interests of the individual to promote social incentives which align with promoting moral behaviour.