Sam Harris's Science of Morality: A Case Study of Scientific Atheism

Sam Harris’s The End of Faith was perhaps the first major book in the wave of New Atheist books published after 911. Harris argued for the importance of challenging our so-called private beliefs, since beliefs (mental representations) cause our behaviour and thus have public consequences. He argued also that so-called moderate religion shouldn’t be off-limits to nontheists, since moderates enable more dangerous, fundamentalist religion by contending that since religion has a harmless form, religion itself is never a primary cause of violence. That book defended a commonsense realist philosophy, according to which beliefs are made true or false by the facts, and the facts support atheistic naturalism.

Whatever you might think of his earlier case for a certain philosophical perspective, you should be struck by the shift taken by his more recent book, The Moral Landscape, in which he attempts to show that social conflicts between groups who disagree on moral issues aren’t inevitable, because science has the potential to show us the truth about moral values just as science has done with regard to the rest of nature. Harris uses his case for a science of morality as a weapon against religion, since theists claim that religion (along with philosophy) are valuable in part for providing the only conceivable framework that justifies morality; that is, the theist means to bash science-centered nontheism for the latter’s inability to justify morality. In the process of countering this moral argument for theism, however, Harris throws the baby out with the bathwater. If morality is actually in the purview of science, then neither the philosopher nor the theologian can have anything crucial to say about moral issues, just as a chef or a politician has no authority to speak about biology or physics.

Unfortunately, Harris’s case for scientific morality conforms to the positivist’s pattern of ironically celebrating science with a philosophy that must be kept in the shadows. In Harris’s case, he should have two reasons not to call attention to the philosophical nature of his arguments for scientific morality. First, those arguments would demonstrate that there is a crucial philosophical debate about morality after all, namely the debate about whether morality can be scientifically justified; the nature of this meta-debate, in turn, prevents a fulsome, Scientistic worship of science at the expense of philosophy. Second, his arguments happen to be badly flawed, often resting on evasive verbal tricks or contradicting each other, due presumably to his contempt for philosophy and thus for its ideals of clarity and rigorous logic even in discussions of nonscientific issues. Harris’s case for scientific morality, therefore, illustrates the perils of scientific, as opposed to philosophical, atheism.

Facts and Values

Harris says that he sees no reason to pay much attention to philosophical ethics or meta-ethics, because that philosophical work “increases the amount of boredom in the universe” (Chapt. 1 n.1). In a talk he gave in New York about his book, he says we don’t have to pay attention to such “intellectual backwaters” (see approx. 12 minutes in). His goal, he says in his book, is to write for a wider audience, not just for academia. This complaint with academic philosophy is all-too-convenient, though, since Harris’s book naturally contains many references to scientific research which are surely just as boring to many people. There are popularizers of either science or academic philosophy, who simplify the cutting-edge research and arguments without crudely misrepresenting them or slanting the discussion with amateurish errors. Harris’s many, more technical footnotes at the back of his book, not to mention his undergraduate degree in philosophy from Stanford, show that he’s equipped to engage responsibly with the philosophy of morality. The reason he doesn’t, therefore, may be just that even a half-way rigorous philosophical analysis of his arguments for scientific morality reveals their egregious flaws and thus his true technique for persuasion, which is his implicit reference to the power of technoscience. Once Harris’s philosophical arguments are disposed of, all that remains is the tribal expectation that science will dictate our moral beliefs just as science has come to dictate so many of our others.

Standing in the way of this inductive inference is the philosophical distinction between facts and values. Science is usually thought to discover what the natural facts are and how they work, not to show how we ought to behave or to settle any other normative matter. According to Harris, there’s no such distinction. After all, he says, science itself rests on certain “built-in” epistemic values, such as the value of critical thinking (11). But this postmodern observation is irrelevant to the point of the fact-value distinction, which has to do with logical spaces of justification, or with the giving of reasons to accept a belief, not with causal connections. Sure, a person’s values can cause her to act and thus to bring about certain facts in the world, perhaps even to make a scientific discovery. This doesn’t mean that the person’s values by themselves logically entail a reference to any of those facts, or that a reason to believe some fact obtains follows just from a belief about what ought to be the case. As the postmodern skeptic is often reminded to her chagrin, to reason otherwise is to commit a form of the genetic fallacy, of reducing an empirical belief’s justification to its subjective origin, such as to the believer’s feeling, character, or some other normative factor. To take the hackneyed example, just because Hitler was evil doesn’t mean all of his empirical beliefs were false; again, rationally assessing whether an empirical belief is justified doesn’t end with considering the believer’s values.

Harris commits the same error in his neurological argument against the fact-value distinction. He points to some evidence that beliefs about facts and values respectively arise from similar brain processes. Again, even were this evidence strong, it would be irrelevant to the distinction at issue since it would show only a certain causal connection--in this case, between mere beliefs about facts or values and their neural origins. Just because those beliefs might be processed in similar ways by the brain, doesn’t mean there’s no logical difference between reasons in support of either kind of belief. This is just another form of the genetic fallacy.

What about the other direction of inference, from facts to values? This is the direction made famous by Hume’s argument that you can’t infer a normative statement merely from factual premises. Harris gives short shrift to this, the more relevant aspect of the fact-value distinction. His most direct response, buried in a footnote, is a citation of Dennett’s hand-waving protest that an “ought” just has to be derivable from something like “an appreciation of human nature” or “a sense of what a human being is or might be” or what the person wants. According to Harris’s quotation of Dennett, there’s no fallacy in that sort of derivation (Intro. n.13). But there obviously is. Just ask yourself whether wanting something makes it right. If Hitler wants to kill Jews, does that mean he ought to do so? Sure, facts about our capacities as humans informs us of our moral limits, but this doesn’t mean that the rightness or wrongness of what we do reduces to facts of those limits. A sociopath may be incapable of acting morally, but that limit by itself doesn’t tell us that his actions are immoral. Perhaps his lack of a conscience is like a gift and his manipulation of weaker people conforms to a posthuman moral standard. And were the facts of our nature that we have no real freedom at all, values would be illusory and thus so too would a derivation of normative from factual statements.

In his New York talk, Harris raises Hume’s point about “is” and “ought” statements, and in the very next sentence pretends to answer it with his postmodern point about the value-ladenness of science (see just before the 25 minute mark of the talk linked above). It doesn’t seem to bother Harris that these points are headed in opposite directions, inferentially speaking.

Dennett and thus Harris (who quotes him) want to know from where moral statements are derivable if not from factual ones. Scientistic atheists are stumped, you see, because they’re wannabe hyper-rationalists. They assume that moral beliefs have to be logically inferred or based on evidence. But morality, like theism, might be irrational and thus either might be particularly fitting for us--as smart as we are compared to all other known species--considering that we’re animals nonetheless.

Is happiness the only moral value?

So Harris hardly overcomes the philosophical distinction between facts and values. But put that aside. According to Harris, “whatever can be known about maximizing the well-being of conscious creatures--which is, I will argue, the only thing we can reasonably value--must at some point translate into facts about brains and their interaction with the world at large” (11). Harris’ main argument is that the only intelligible object of value is the welfare or happiness of conscious beings and that morality therefore reduces to facts about what makes those being happy, facts which scientists can discover.

(Harris emphasizes another point, about the origin of values. For example, he says, “consciousness is the only intelligible domain of value,” “consciousness is the basis of human values and morality,” and any other “source” would be “the least interesting thing in the universe” (32). He speaks as if these statements were equivalent, which they’re not, since whether something is interesting has to do with whether it’s the object of value, not with whether it’s the source or sustaining cause of values. In any case, this point about the source of values is once again irrelevant. You might as well say that morality reduces to facts about our solar system, since that system produced and sustains the conscious creatures who evaluate things. The origin or vehicle of morality may be scientifically interesting, but it’s irrelevant to the philosophical, meta-ethical question of whether science can tell us everything we could want to know about morality.)

Harris’s relevant premise, then, is that our welfare is the only possible thing we can value. His meta-ethical defense of this statement proceeds by way of thought experiments, not by any scientific evidence. Harris merely appeals to his reader’s intuition, asking whether it would make any sense to call the most miserable state of affairs morally good. He contrasts The Good Life, in which a person is successful, emotionally fulfilled by social connections and a worthwhile job, and fortunately blessed with a long life free from much pain, with The Bad Life in which an illiterate, impoverished, starving, terrified woman in war-torn Africa watches as her child is raped and killed, and is forced to flee from a gang of soldiers who rape and kill her (15-16). Harris wants to know what someone could possibly value were the difference between those two lives not to matter to her. If someone wanted to steer another person’s life in one or the other direction, as it were, could we conceivably have the same moral opinion of that controlling person, regardless of the direction? Harris wants to say that this difference between happiness and unhappiness is all there is to moral value, that a moral value not captured by the distinction between The Good Life and The Bad Life is inconceivable and impossible (17). And since that distinction has to do with facts that relate--in a scientifically explainable way--to the brains of conscious creatures, if you share Harris’ intuition about the distinction, you’ve got to agree that there can be a science of morality.

Clearly, there’s a difference between happiness and unhappiness and clearly most people would rather be happy than otherwise. Most people prefer pleasure to pain, for well-known natural reasons. The question is only whether this preference is all there is to morality, or indeed whether it’s even relevant to the question of moral value. All that’s logically required to refute Harris’s conceivability argument is to provide an example of a moral value not captured by his point about the majority’s preference for happiness. One such example is found in my existential argument for the aesthetic value of suffering due to knowledge of our absurd and tragic plight as naturally accursed creatures, a value that trumps the more conventional, materialistic preference for personal fulfillment. (See Happiness and Cosmicism.) Ask yourself not just whether we can or actually do prefer a life of personal pleasure to one of misery, but whether we morally ought to, given our tragic knowledge of our animal nature, of the obviousness of atheism, and thus of the absurdity of our theistic delusions. Instead of appealing fallaciously to an intuition’s popularity, as does Harris, consider whether we ought to be happy on Earth even though Earth isn’t Heaven. 

Take even the fortunate soul in Harris’ example of The Good Life. In that thought experiment, the hero’s job is actually to help other people, including children in the developing world by means of a billion-dollar grant. So a happy person needn’t be selfish. But notice how Harris’s own thought experiment self-destructs as soon as you start to examine it. Anyone interested in dedicating her life to helping people much less fortunate than her is bound to be motivated by suffering due to her knowledge of their misery. That suffering will compete with her pleasure, tearing her away from her cocktail parties, rendering her enjoyment of her comparative luxuries petty and awkward. How could she enjoy her success and her friends, knowing that millions of children suffer each moment she’s offered some pleasure or other, and knowing too the ultimate cause and consequence of that grotesque inequality? The cause, of course, is the fact that we’re byproducts of mindless, inhumane natural processes, and the consequence is that we’re thoroughly natural creatures, namely animals whose fates are tied entirely to our bodies, for whom there’s no perfect justice. Harris’s own example reveals that happiness isn’t crucial to moral value, after all, since the morally praiseworthy person living The Good Life is bound to be unhappy! Sure, she may be successful, wealthy, popular, and so on, but she’ll suffer from anxiety rather than be contented. Her angst is what will motivate her to live a more respectable, altruistic life, thus freeing Harris’ argument from the counterintuitive implication that selfish pleasure is all that’s morally valuable.

If that hero’s happiness (in the sense of a set of higher and lower pleasures tied to her brain states) isn’t what makes her life morally valuable, why do we morally praise her? Surely, because she chooses to sacrifice her pleasure to help others, and she chooses this out of a sense of duty to respond well to her suffering from the horror of our existential predicament. Because we’re the spawn of a mindless, inhumane world and not children of any loving god, some people succeed in the struggle for life while others fail, some are lucky while others are not, and many of us live effectively as subhumans exploited by posthuman demigod overlords (globe-trotting oligarchs). The hero can’t bear to live selfishly, because her conscience prevents her from contributing to an aesthetically appalling pattern in which she enjoys luxuries while millions of impoverished children starve. Again, what’s morally praiseworthy about this hero isn’t just that she happens to succeed in helping others, but that she chooses to try even at the inevitable cost of sacrificing her peace of mind and her enjoyment of her privileged lifestyle. Remove this altruistic aspect from Harris’ example of The Good Life, and the intuition is no longer triggered that there’s anything morally right about the person’s pleasures from her fortune, social network, and her long life, that is, about her happiness. As I say elsewhere, not only is happiness not the only morally correct goal, but happiness in the sense of personal contentment is especially unbecoming for creatures in our existential situation. We morally ought to be restless and angst-ridden, given what we now know, thanks indeed largely to Harris’s cherished modern science.

I stress the aesthetic aspect of existential morality, but the philosopher Immanuel Kant provides a third example that refutes Harris’ contention, that is, an example in addition to my case against the moral value of happiness and to the self-destruction of Harris’s example of The Good Life. Kant famously argues for what ethicists call deontology, which is the philosophy that morality is a matter of duty, not of happiness. Naturally, Harris gives short shrift to Kant as well, stating in a footnote that Kant's categorical imperative (that we have a rational duty to act as if our private reason for acting were a universal law, so that we’d expect to reap what we sow), “amounts to a covert form of consequentialism” (Chapt.1, n.10). This is supposedly because, as Mill allegedly showed, Kant’s rational basis for morality works only on the assumption that rationality is generally beneficial and thus maximizes happiness. Even were this so, which Kantians will hardly concede, happiness might be a morally irrelevant byproduct of rationality. In any case, as I’ve argued elsewhere, rationality rather produces angst and alienation than contentment. (See Curse of Reason.) Were everyone good Kantians, always worried about treating each other as ends rather than as means, we’d suffer from our awareness of how often nature treats us all as means rather than as ends.   

In another footnote, Harris criticizes Kant’s notion of treating people as ends, arguing that we’re split between our present and future selves, so that when we act prudently to benefit ourselves in the future, we abuse our present selves as means to that end (Chapt. 2, n.45). But once again, Harris’s own philosophical assumptions undermine his scientistic notion of morality. Harris says that morality is a concern only for conscious creatures. According to the philosopher John Locke’s view of personal identity, a person is essentially the interconnections between a set of mental states, especially memories. A stream of consciousness can be interrupted without losing personal identity as long as memories are retained of the earlier mental states. So a Kantian could reply that present and future states of a self are united by memories. Moreover, according to Kant, we’re ends rather than means because we have freewill in the sense of autonomy, and without that attribute morality is nonsensical in the first place.

Note also that Harris’s example of The Bad Life is morally irrelevant. The victim’s misery is surely not to be preferred, but this doesn’t make it morally wrong, precisely because she’s a colossal victim who isn’t to blame for anything, and even those who immediately persecute her aren’t to blame (the soldiers are “drug-addled” and indoctrinated as children). The true culprits who are morally blameworthy are the oligarchs who control the political and economic relations between countries. But the point is that immorality isn’t just a lack of happiness; rather, it’s the cowardice or folly that causes us to ignore our existential predicament, retreating to theistic or to other delusions, as well as the tastelessness (unoriginality) of reinforcing that predicament by contributing to inhumane dominance hierarchies.

Finally, a word about Harris’s dismissal of objections to his conceivability argument for the equivalence of moral value with the welfare of conscious creatures. In talks and his book, he often belittles his critic's objections by saying that they run up against the self-evidence of Harris’s arguments and therefore don’t need to be taken seriously. For example, he says,

if we don’t want everyone to experience the worst possible misery, we shouldn’t do X. Can we readily conceive someone who might hold altogether different values and want all conscious beings, himself included, reduced to the state of the worst possible misery? I don’t think so. And I don’t think we can intelligibly ask questions like “What if the worst possible misery for everyone is actually good?” Such questions seem analytically confused...if someone persists in speaking this way, I see no reason to take his views seriously. (Chapt.1, n.22)

Here, Harris dismisses a strawman argument, reducing the conclusion of his own conceivability argument to a mere analytic statement that depends on the stipulated meaning of words. The reason it makes no sense to ask whether the worst possible misery for everyone is good is that “worst” includes the meaning bad. But the description of The Ultimate Bad Life, in which misery is maximized, needn’t use that particular loaded label, “worst.” Indeed, we shouldn’t beg the moral question at issue by initially describing this scenario as very bad, but should rather neutrally say that that misery is complete, extreme, permanent, and so on. Then the question becomes whether, say, complete misery for everyone is actually moral. Given existential and Eastern philosophical traditions, this question can hardly be dismissed as easily as an irrelevant one about the loaded use of words. The renunciation of our pleasures may indeed be the moral choice, given our predicament that those who know too much are condemned to live with angst, and that we can either flee to delusion or nobly persevere as tragic heroes.

Of course, the philosophical and indeed the scientific standard is to go out of your way to ensure that you’ve made no error, to work with the critic to formulate the best possible objection to see whether your argument holds up. The alternative is to dogmatically presume that you’ve got a godlike handle on the truth and to belittle objections, like a self-righteous theist. Harris’s evident pride goes before the fall of his argument, and the source of his pride is his commitment to the religion of Scientism, which assures him that somehow science has all the answers--even if you’ve got to employ shoddy philosophical arguments to prove as much.

Just what is Harris's science of morality?

Were Harris’s conclusion just that science can show us how to be happy, assuming happiness is sustained by certain brain states, there would be little to criticize. The problem is that he thinks we can replace the philosophy of ethics with that sort of science, because the equivalence of happiness with moral value is allegedly self-evident. As I’ve shown, his argument for that equivalence is deeply flawed. But put that aside for the moment and consider what exactly Harris wants to show. He says that his “general thesis” isn’t merely “that science can help us get what we want out of life,” but that, in principle, science “can help us understand what we should do and should want--and therefore what other people should do and should want in order to live the best lives possible” (28). But this is a slippery rather than a clarifying statement of his thesis. The weakness in saying that science can merely “help” in this regard trivializes his notion of scientific morality, since even his critics will agree that an “ought” can follow from a combination of “is” and “ought” statements. The main criticism is just that a prescription doesn’t reduce exclusively to descriptions, in the sense of being implied by factual premises without any normative ones.

Harris proves himself to be an even slipperier fish when he compares morality to health, pointing out that medical science proceeds regardless of philosophical worries about the definition of “health,” but also conceding that “Science cannot tell us why, scientifically, we should value health. But once we admit that health is the proper concern of medicine, we can then study and promote it through science” (37). As far as I can tell, this admission is devastating to Harris’ overall argument, since it reduces his thesis to what he explicitly rules out as such, namely to the statement just that science can help us get what we want (i.e. happiness). If science can’t tell us why we should value health, and health is comparable to morality for Harris’s purposes, then science can’t tell us why we should value happiness (the welfare of conscious creatures). In that case, the science of morality doesn't replace the philosophy of ethics, and that science becomes the instrumental business of helping many people achieve their philosophical or theological goal of being happy.

As I’ve said, Harris’s conceivability argument in support of his main premise is itself patently philosophical rather than scientific, so his own case for scientific morality bears out his admission that science doesn’t settle the key philosophical issue of what we should morally value. But the slippery fish squirms yet again, adopting the same ploy as the scientific atheist Jerry Coyne’s, of defining “science” broadly as “our best effort to understand what is going on in this universe,” the boundary between this “science” and “the rest of rational thought” being sometimes impossible to draw (29). With this Quineian holistic view of the relation between rational endeavours like math, philosophy, and science, Harris can credit “science” in the broad sense for the work of philosophy and thus satisfy his scientistic preference for science in the narrow sense. Why not instead use “philosophy” broadly as the name of our best effort to understand the universe, and thus credit philosophy for the work of Newton, Darwin, and Einstein? Again, if you hollow out “science” as a weasel word, you can have a cheap scientific morality, but philosophy will still be needed to settle the normative questions about morality, which are the crucial ones about what goal we ought to pursue, while science in the proper sense will consist of discoveries and explanations that enable us more efficiently to pursue that goal. Harris is clearly all over the map with regard to what exactly a science of morality amounts to: because of his disdain for philosophical rigor, and because of the conflict between his scientism and his philosophical method, Harris is unsure of how he’s entitled to state his conclusion. Thus he equivocates, backslides, and contradicts himself.

There’s more in The Moral Landscape than what I’ve discussed here, but I think the above demolishes Harris’s main argument and exposes the amusing ironies of scientific atheism. The scientific atheist thinks that science rather than philosophy is the best, if not the only useful, weapon against religion. But the scientific atheist typically subscribes to what is effectively the religion of Scientism, and inevitably resorts to philosophy in arguing against theism. Thus, scientific atheism self-destructs twice over.