Sam Harris, Popper and Morality

Sam Harris wrote a book called The Moral Landscape (TML), which is about a secular moral philosophy. It sucks. I’m going to contrast his position with Popper’s, which also has some flaws.

Harris claims that we should accept the idea that morality is about the well-being of sentient creatures and depends on “events in the world and states in the human brain” (p. 5). He then claims that if we accept this idea philosophers who claim you can’t derive an ought from an is have missed the point.

Chapter 5 of The Open Society and Its Enemies pdf, Kindle is about why Popper considers it impossible to derive values from facts. Different people can see the same facts and come to different moral judgements and so morality can’t be derived from facts. This is correct as it stands but there is more interesting stuff in the notes.

OSE Chapter 5 Note 18

The hope of getting some argument or theory to share our responsibilities is, I believe, one of the basic motives of ‘scientific’ ethics. ‘Scientific’ ethics is in its absolute barrenness one of the most amazing of social phenomena. What does it aim at? At telling us what we ought to do, i.e. at constructing a code of norms upon a scientific basis, so that we need only look up the index of the code if we are faced with a difficult moral decision? This clearly would be absurd; quite apart from the fact that if it could be achieved, it would destroy all personal responsibility and therefore all ethics.

Popper claims scientific ethics is absurd but doesn’t explain why. I think the problem is that if you claimed to have a scientific derivation of morality that wouldn’t get you anywhere because that just raises questions that can’t be answered by your theory. Are you doing the science correctly? And why would it be right to judge this issue using science? And what counts as science in the first place and why? Also progress would be impossible if it were true. Progress involves inventing new ideas that won’t be on the approved list because those ideas and the practices required to criticise and understand them don’t exist yet. So if there is such a list then all progress of any kind will have to end.

Or would it give scientific criteria of the truth and falsity of moral judgements, i.e. of judgements involving such terms as ‘good’ or ‘bad’? But it is clear that moral judgements are absolutely irrelevant. Only a scandalmonger is interested in judging people or their actions; ‘judge not’ appears to some of us one of the fundamental and much too little appreciated laws of humanitarian ethics. (We may have to disarm and to imprison a criminal in order to prevent him from repeating his crimes, but too much of moral judgement and especially of moral indignation is always a sign of hypocrisy and pharisaism.) Thus an ethics of moral judgements would be not only irrelevant but indeed an immoral affair. The all-importance of moral problems rests, of course, on the fact that we can act with intelligent foresight, and that we can ask ourselves what our aims ought to be, i.e. how we ought to act.

This seems a bit muddled. Are we supposed to judge the criminal’s actions or not? I would say if we’re going to lock people up we should have made sure to judge their actions bad enough that we want to prevent him from acting that way by force.

Popper talks of “too much” moral judgement. I think this refers to something like a real problem but the problem is people making moral condemnations without understanding the issues or the situation the person in question faced. In other words, it’s not a problem of moral judgement but a problem of lack of judgement: that is, a lack of carefully considering the problems involved.

An example of not understanding more problems: many people in the UK say the US ought obviously to ban guns to stop mass shootings at cinemas and that sort of thing. Those people fail to understand the context. It is perfectly possible to use a gun properly, i.e. – don’t shoot people, or only shoot them in self defence. So if you ban guns you deprive people of an item that can be used in a legitimate way, including defending people from being shot up at a cinema. This is a bad idea. There are other relevant moral issues too.

Nearly all moral philosophers who have dealt with the problem of how we ought to act (with the possible exception of Kant) have tried to answer it either by reference to ‘human nature’ (as did even Kant, when he referred to human reason) or to the nature of ‘the good’. The first of these ways leads nowhere, since all actions possible to us are founded upon ‘human nature’, so that the problem of ethics could also be put by asking which elements in human nature I ought to approve and to develop, and which sides I ought to suppress or to control. But the second of these ways also leads nowhere; for given an analysis of ‘the good’ in form of a sentence like: ‘The good is such and such’ for ‘such and such is good’), we would always have to ask: What about it? Why should this concern me? Only if the word ‘good’ is used in an ethical sense, i.e. only if it is used to mean ‘that which I ought to do’, could I derive from the information ‘x is good’ the conclusion that I ought to do x. In other words, if the word ‘good’ is to have any ethical significance at all, it must be defined as ‘that which I (or we) ought to do (or to promote)’. But if it is so defined, then its whole meaning is exhausted by the defining phrase, and it can in every context be replaced by this phrase, i.e. the introduction of the term ‘good’ cannot materially contribute to our problem. (Cp. also note 49 (3) to chapter 11.)

Note 49(3) to Chapter 11 of OSE:

Essentialism and the theory of definition have led to an amazing development in Ethics. The development is one of increasing abstraction and loss of touch with the basis of all ethics—the practical moral problems, to be decided by us here and now. It leads first to the general question, ‘What is good?’ or ‘What is the Good?’; next to ‘What does “Good” mean?’ and next to ‘Can the problem “What does ‘Good’ mean?” be answered?’ or ‘Can “good” be defined?’ G. E. Moore, who raised this last problem in his Principia Ethica, was certainly right in insisting that ‘good’ in the moral sense cannot be defined in ‘naturalistic’ terms. For, indeed, if we could, it would mean something like ‘bitter’ or ‘sweet’ or ‘green’ or ‘red’; and it would be utterly irrelevant from the point of view of morality. Just as we need not attain the bitter, or the sweet, etc., there would be no reason to take any moral interest in a naturalistic ‘good’. But although Moore was right in what is perhaps justly considered his main point, it may be held that an analysis of good or of any other concept or essence can in no way contribute to an ethical theory which bears upon the only relevant basis of all ethics, the immediate moral problem that must be solved here and now. Such an analysis can lead only to the substitution of a verbal problem for a moral one.

The idea that morality is about “the immediate moral problem that must be solved here and now” is at best a bad formulation. You need moral principles to judge what you should do in a particular situation because otherwise all you have is a pile of competing claims and no way to sort them out. The best construction I can put on what Popper said here is that each situation has some unique aspects and we have to think on our feet to apply moral principles to that situation and no definition of the good could contribute to such thought. Presumably that’s what he meant but he didn’t state it clearly.

In TML Sam Harris claims that morality can be derived from facts about human wellbeing and we can understand those facts by looking at how the brain lights up when a person is happy. This invokes all of the bad ideas Popper criticised above. He defines morality as being about wellbeing but this just evades the question of what counts as wellbeing and why. And why is the way a person’s brain lights up the relevant issue? Harris claims that this is because thought is instantiated in the brain, but by that logic the vast bulk of knowledge is about paper or electronic information storage devices since most knowledge is instantiated in those forms.

At any particular time you’re going to have some areas of your life you find unsatisfactory. To make your life better you have to find out why these aspects of your life are unsatisfactory and solve the relevant problems. That’s an epistemological problem, which has nothing to do with brain chemistry.

Indeed, if a person is solving problems then his brain chemistry has to be explained in terms of the morality of how to solve problems. He will be thinking about what is required by epistemology (and other stuff too, like physics or economics or aesthetics or whatever) and his brain chemistry will instantiate the relevant ideas.

Most of TML is spent discussing moral issues without tying them to brain states or explaining them in terms of brain states in any substantive way. Rather, the discussion of brain states gets in the way of openly discussing moral standards in some cases. For example, on p. 94 he claims that looking at the brains might help us understand whether we should choose to throw one person in the path of a moving train to save five more. If the best thing you can think of to do involves shoving somebody under a train you ought to consider that you don’t actually understand the problem too well and shoving somebody under a train under those circumstances would be stupid or criminal.

And as I noted in discussing Popper’s objection to scientific ethics above, if there was some particular scientific theory that provided the answers to all ethical questions, then progress would be impossible. It’s not clear to me how Sam Harris would avoid this. Once we understand how the brain works do we understand all of morality? If so, then all progress will end when that happens.

Just one final note, I think TML is crap not just because it’s so badly wrong but because there is so much in it that is silly. There are so many points where Harris should have said “okay, there’s something badly wrong here, I’m writing stuff that’s just complete crap.” Take, as an example, the bit on p. 71 where he discusses Parfit. According to Parfit if we were to aggregate utility then it might be better to have one person who is really happy to a world where there are billions of people who are just a little bit happy. But we have to aggregate utility, claims Harris, so we have to think about such “paradoxes”. You don’t have to think about aggregation. You have to think about solving problems. If you solve a problem you make things better. If you don’t solve problems you don’t make things better. End of story. And how could you possibly be put in the position of making a decision between the sorts of scenarios Parfit discusses? How could you be in a position to decide whether there will be one person who is very happy or billions who are slightly less happy? You would either have to be a mass murderer or a tyrant who controls whether people get to have children.

TML is abysmal. Read David Deutsch, Ayn Rand, William Godwin, Karl Popper or Thomas Szasz instead.

UPDATE: What I said above is wrong in an important respect. Ethics doesn’t have a basis in problems or anything else including principles, but both problems and principles are important. Ethics can’t be based on problems because problems only come up in the light of a clash between different pieces of moral knowledge. So the knowledge has to be there before you can have the problem. If you’re going to solve the problem rather than just paper over it then the moral knowledge involved should be made explicit enough to be criticised.