SCIENCE CANNOT DETERMINE HUMAN VALUES

Abstract

Sam Harris, in his book The Moral Landscape, argues that ‘science can determine human values.’ Against this view, I argue that while secular moral philosophy can certainly help us to determine our values, science – at least as that word is commonly understood – must play a subservient role. To the extent that science can ‘determine’ what we ought to do, it is only by providing us with empirical information, which can then be slotted into a chain of deductive (moral) reasoning. The premises of such reasoning, however, can in no way be derived from the scientific method: they come, instead, from philosophy – and common sense.

Introduction¹

I recently finished a booklet by the neuroscientist and author Sam Harris – on lying² – and I may write about it at a later date. But for the purposes of the present article, I would like to dig up an older Harris book, The Moral Landscape,³ so that I may express my hitherto un-expressed puzzlement about Harris's (now aging) ‘bold new’ claim, presented in this otherwise cogent manuscript, that science can ‘determine human values’.

In his new book (the one about lying) Harris says, in effect, you should never, ever, do it – yet his pretense in The Moral Landscape to be revolutionizing moral philosophy seems to me the very height of dishonesty. What he actually does in this book is what you might call ‘plain old’ secular moral reasoning – as profane philosophers have been doing for a very long time – but he claims instead to be using ‘objective’ science to decide what's right from wrong. That Harris could really have managed to transcend the famous ‘is/ought’ chasm seems unlikely, so let us go ahead and take a look at his argument to see what may be going on.

Discussion

I shall start by saying what the ‘is/ought’ divide is, for the sake of any reader who may be unfamiliar with the concept. It's an old idea, tracing at least to David Hume,⁴ and its gist is that there is no way to reason from facts about the way the world is, to statements about the way the world should be, without smuggling in additional premises. In short: you can't derive values from data. I'll use one example to illustrate and then move on.

Example. It's a fact that rape occurs in nature – among chimpanzees, for instance; and there are some evolutionary arguments to explain its existence in humans and non-humans alike.⁵ But this fact tells us exactly nothing about whether it's OK to rape people. This is because ‘natural’ doesn't entail ‘right’ (just as ‘unnatural’ doesn't necessarily mean ‘wrong’)⁶ – indeed, the correct answer is that it's not OK, and this is a judgment we make at the interface of moral philosophy and common sense: it isn't an output of science.

You get the idea. The domain of science is to describe nature, and then to explain its descriptions in terms of deeper patterns or laws. Science cannot tell us how to live. It cannot tell us right and wrong. If a system of thought claims to be doing those things, it cannot be science. If a scientist tells you she has some statements about how you ought to behave, they cannot be scientific statements, and the lab-coat is no longer speaking as a scientist. Questions about ‘How should we live?’ – for better or worse – fall outside the purview of ‘objective’ empirical research. We have to sort them out, messily, by ourselves.

Now: if there were a way to get from ‘is' to ‘ought’ it would take a work of philosophical genius to lay it out, and (whatever else its merits) Harris's book is not a work of philosophical genius. I can summarize his argument in a few lines:

1. 1. Morality is ‘all about’ improving the well-being of conscious creatures.

2. 2. Facts about the well-being of conscious creatures are accessible to science.

3. 3. Therefore science can tell us what's objectively ‘moral’ – that is, it can tell us whether something increases, or decreases, the well-being of conscious creatures.

Here is the problem. Premise (1) is a philosophical premise. It is not a fact of science, it is not a fact of nature (and if it's supposed to be derivable from science or nature, Harris doesn't show us how). Instead, it's a value judgment. Now, you might think this is a good premise; you might not – and even if you do think it is basically on track, there is still a lot of philosophical work to be done to spell it out. (Exhibit A – how do you define well-being in the first place, and why should we accept your definition?)

What this boils down to, then, is that given a certain philosophical value, premise, or starting-point, science can feed us relevant facts in our sorting-out of how to live. Ok, but so what? That's just what science has always been able to do. This is just ‘plain old’ moral philosophy, minding the facts.

But let us grant Harris his first move. Let us give him his philosophical premise. Maybe he means to say that science is getting sophisticated enough to help us solve certain precise moral puzzles that exist within the overarching philosophical framework we've agreed to (i.e., some version of utilitarianism). Maybe neuroscientists will one day tell us astonishing things about how pain is processed in the brain, for example, and this will allow us to decide the correct moral outcome – again, premises granted – in some particular complex case.

Maybe. But if this is what Harris wants to say, the examples he relies on are weak. Consider his discussion of the Taliban. Harris says that according to science, the Taliban's treatment of women (enforced burqa-wearing, etc.) is objectively morally wrong. Why? Because enforced burqa-wearing (etc.) is not conducive to the well-being of conscious creatures, namely the conscious creatures forced to wear burqas.

I hope we can agree that we didn't need science to tell us that treating women in this way is bad (or at least seriously problematic in a number of different ways): common sense, or, better, secular moral philosophy, will do just fine. And if someone disagrees (say, the Taliban), intoning ‘but science says you're mistaken’ will do little to change their minds – and rightly so. What Harris appears to be doing is trying to hijack the prestige and ‘objectivity’ of the scientific enterprise⁷ to label the behavior of certain groups as categorically WRONG.

In philosophy, of course, there is a big debate about whether certain moral systems are better than others, or whether, indeed, there are ‘objective’ moral facts at all.⁸ This has been going on for a few hundred years. By asserting that all we need to know about morality is that some unspecified version of utilitarianism is correct, and that, further, there are strict facts about what sorts of things maximize utils, Harris adds nothing to the debate. He just sidesteps it.

By the way, Sam Harris came to Oxford a few years ago (when I was a graduate student there) to give a talk about The Moral Landscape. This particular talk was hosted by Richard Dawkins. To kick off the Q&A, Dawkins pressed Harris on just what he was saying that was new. Here is a bit of that conversation:

Dawkins: You're facing the classic problems that moral philosophers have been facing for a long time… You appear to be bringing to those problems a new thought, which is that science, as opposed to just philosophic thinking – reason – could help. Now, moral philosophy is the application of scientific logical reasoning to moral problems. But you are actually bringing your neurobiological expertise to bear, which is a new way of doing it. Can you tell me about that, because I'm not quite clear about how doing neurophysiology adds insight into these moral problems.

Harris: Well, I actually think that the frontier between science and philosophy actually doesn't exist… Philosophy is the womb of the sciences. The moment something becomes experimentally tractable, then the sciences bud off from philosophy. And every science has philosophy built into it. So there is no partition in my mind.

Now we see the source of the problem. By ‘science’, Harris evidently means, ‘philosophy’ … or at least something that's not different from philosophy in a principled way. Now, of course it is the case that science has certain philosophical assumptions built into it;⁹ and of course is the case that ‘scientific logical reasoning’ (as Dawkins put it in his question) is central to both the carrying out and interpreting of empirical research, as well as to armchair philosophy. But when Harris says that ‘science’ can ‘determine human values' he quite clearly expects his reader (or listener) to conjure up an image of those lab-coat wearing investigators I alluded to earlier – otherwise his claim is uninteresting.

To illustrate this point, let me fish out my old brochure for a second and confirm what the title of Harris's Oxford talk was – that radical-sounding title that sold so many tickets – yes, here it is, it's, ‘Who Says Science Has Nothing to Say About Morality?’ If we do a quick update, however, based on Harris's personal definition of science, we get … ‘Who Says Philosophy Has Nothing to Say About Morality?’

The answer is: no one ever said that. Moral philosophy plus facts is not ‘science’ telling us objective moral truths.

Conclusion

I would like to close on a personal note. I was in the audience at Harris's talk, and during the Q&A period I tried to nudge him on two specific points. First, how exactly did his argument get us over the is/ought divide; and second, what can ‘science’ tell us about morality that we didn't already know from common sense (or secular moral reasoning). Our exchange can be seen in a YouTube video of the event,¹⁰ and I'll make just one comment before you watch it (if you do). Notice the first four words of Harris's reply to my question: ‘The moment you grant…’ My point has been that what Harris wants you to grant is a philosophical, not a scientific, premise; hence, his ‘moral landscape’ is not scientifically determined as he claims.¹¹