Science and Morals: Can morality be deduced from the facts of science?

You can’t beat science.  “One by one, the great questions of philosophy, including ‘Who are we?’ and ‘Where did we come from?’ are being answered to different degrees of solidity. So, gradually, science is simply taking over the big questions created by philosophy. Philosophy consists largely of the history of failed models of the brain.” So much for philosophy! Thus spake eminent biologist and chronicler of sociobiology E. O. Wilson, in a 2009 interview[1] where he also said “If the empiricist world view is correct, ought is just shorthand for one kind of factual statement, a word that denotes what society first chose (or was coerced) to do, and then codified.” So, morality can be deduced from science, according to Wilson.

Wilson’s confidence in the omnipotence of science, his belief in scientific imperialism, is shared by vocal members of the so-called New Atheists. Richard Dawkins, another well-known biologist, has notoriously said that belief in anything that cannot be scientifically proved, i.e., faith, “is one of the world’s great evils, comparable to the smallpox virus but harder to eradicate…”

The New Atheists are moral people. But as I will show, they are wrong to think that their morality, or any morality, can be derived from science.


Dawkins deems faith “evil precisely because it requires no justification and brooks no argument”. Faith in this sense seems to include non-religious as well as religious beliefs. All people believe things that they cannot prove, many of which Dawkins would surely allow as good: the virtues of generosity, kindness, courage and so on. But Dawkins seems to be especially critical of faith that has a religious basis: belief in God, in the specifics of religious stories, and religious prescriptions that violate the (often-unstated) morals of 21st century Western intellectuals, such as the rights of women, sexual freedom, freedom of belief, the innocence of abortion, the evils of punishment, and so on.

Dawkins concedes that “it’s very difficult to come to an absolute definition of what’s moral and what is not.” [talk, 2012] He does not claim that morals can be deduced from science. But he does say he has given up what he calls “the hectoring myth that science can say nothing about morals.”  Evidently science has something to say. But he is reluctant to say just what it is.

Dawkins sidesteps saying what morality is, by suggesting how it might come about. In other words, he retreats from conclusion to process. The process involves consensus: “Be good for the reason that you’ve decided together with other people the society we want to live in: a decent humane society. Not one based on absolutism, not one based on holy books and not one based on … looking over your shoulder to the divine spy camera in the sky.”  But where is the guarantee that everyone in the group that is “together deciding” on a morality will have relinquished all faith, all allegiance to holy books and all belief in an omnipotent god? Dawkins has dodged the question.

Not so the most forthright and committed of the New Atheists, Sam Harris. In his book The Moral Landscape: How Science Can Determine Human Values, Harris directly confronts the issue and comes down on the side of science.  He solves the ethical problem by arguing that “questions about values — about meaning, morality, and life’s larger purpose — are really questions about the well-being of conscious creatures.”  “Values are a certain kind of fact” he argues. Harris also points to felt experience as a signal of value: “[T]here’s no notion, no version of human morality and human values that I’ve ever come across that is not at some point reducible to a concern about conscious experience and its possible changes.”  For Harris, “the good” is about the feelings and “flourishing” of individuals.


“[C]oncern about conscious experience” — human feelings — is integral to Sam Harris’s scientific take on human morality. But feelings are not a reliable guide to truth, moral or otherwise, if only because many scientific, value-free, statements nevertheless elicit strong emotional reactions. For example, in the Origin of Species — which is subtitled The Preservation of Favoured Races in the Struggle for Life — Charles Darwin describes many examples of competition: “the more vigorous … gradually kill the less vigorous”, etc. Some critics of Darwin have reacted emotionally to the word struggle and the implication that some individuals and races survive at the expense of others.  The idea that some animal, plant and human varieties — races — are ‘superior’, in the sense that they will prevail in the ‘struggle’, makes Darwin  “obviously racist” in the eyes of one author[2]. Indeed, any reference to human individual differences, especially in relation to ‘race’, will elicit passionate feelings in many readers, no matter how ‘scientific’ the context or disinterested the account.  Facts are neutral; the human reaction to them very often is not. People find it very difficult indeed to separate the factual from the emotional.

This is why philosopher David Hume, perhaps the most perceptive figure of the 18th century Enlightenment, famously separated “ought”, the dictates of morality, from “is”, the facts of science.  Reason is value-neutral, Hume argued:

It is not contrary to reason to prefer the destruction of the whole world to the scratching of my finger. It is not contrary to reason for me to chuse my total ruin, to prevent the least uneasiness of an Indian or person wholly unknown to me[3].

He goes on to point out that

Since a passion [motive, desire] can never, in any sense, be called unreasonable, but when founded on a false supposition, or when it chuses means insufficient for the designed end, it is impossible, that reason and passion can ever oppose each other, or dispute for the government of the will and actions. The moment we perceive the falshood of any supposition, or the insufficiency of any means, our passions yield to our reason without any opposition. I may desire any fruit as of an excellent relish; but whenever you convince me of my mistake, my longing ceases. I may will the performance of certain actions as means of obtaining any desired good; but as my willing of these actions is only secondary, and founded on the supposition, that they are causes of the proposed effect; as soon as I discover the falshood of that supposition, they must become indifferent to me.

In other words, reason is just the link between passion (will, motivation) and action: “Reason is, and ought only to be the slave of the passions, and can never pretend to any other office than to serve and obey them.” Without passion, the facts established by reason are impotent. The findings of science are neither moral or immoral, according to Hume. Hume’s distinction between “is” and “ought” is not a distinction between doing science and doing religion. It is a distinction between being and acting.

Sam Harris admits that this is “the received opinion in intellectual circles” but begs to disagree. He makes three points:

  1. 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;
  2. the very idea of “objective” knowledge (i.e., knowledge acquired through honest observation and reasoning) has values built into it, as every effort we make to discuss facts depends upon principles that we must first value (e.g., logical consistency, reliance on evidence, parsimony, etc.);
  3. beliefs about facts and beliefs about values seem to arise from similar processes at the level of the brain[4]

Point 3 is something of a red herring, in the sense that any difference of behavior is caused by, and thus will be reflected in, some brain activity. We don’t yet understand enough about how the brain works to make much of the apparent similarity that Harris describes.

Point 1 assumes what it purports to support: that the “well-being of conscious creatures” is our highest good. Not everyone will agree.

Point 2, that the pursuit of science involves values, is of course, correct. The reason is that, as Hume argued, any action requires some kind of motivation, some kind of value. Hence, the fact that doing science requires scientists to believe in “logical consistency, reliance on evidence, parsimony, etc.”, not to mention honesty and curiosity, does not invalidate Hume.  Neither does the fact that pursuing science requires faith in a fixed, hence discoverable, nature. The stability of natural law is not self-evident, like a syllogism or simple arithmetic. In order to seek, a scientist must believe there is something to be found.

Yes, to do science requires values; but the facts thus obtained are not themselves values. The facts that men are on average taller than women, or that African-Americans have lower average IQ than white Americans, are equally value-neutral. But, human nature being what it is, the second fact is likely to elicit much stronger emotions than the first, even though both are just facts. Neither one impels us to action, unless we feel, as a value, that race or gender differences are a bad thing.

‘Science-based’ ethics: Human flourishing

If science cannot provide us with an ethics, how about those ethical systems that pretend to be science-based? They cannot be based on science, so how should we judge them?

There are at least two supposedly science-based ethical systems on offer. One is Sam Harris’s “human flourishing” idea, which rests on the well-being of individuals. The other is based on evolution. I’ve already alluded to some problems with Harris’s proposal. Here are a couple more. “Values are a certain kind of fact” says Harris in a 2010 TED talk. Perhaps, but values are not scientific facts, because they cannot be tested. We can show that one course of action leads to better results than another. But “better” is always a judgment of value not a provable fact. Harris provides a number of apparent counter-examples, but solves them all by resorting to his well-being idea. And, as commentator Sean Carroll points out,  who says that personal well-being is the highest good anyway?

The other alternative is evolution and natural selection.  Wilson and radical behaviorist B. F. Skinner have both suggested[5] that evolutionary epistemology in some form allows “is” to be transformed into “ought”.  In his provocatively titled 1971 bestseller Beyond Freedom and Dignity Skinner said:

“Questions of this sort…are said. . .to involve ‘value judgments’—to raise questions…not about what man can do but about what he ought to do. It is usually implied that the answers are out of the reach of science…It would be a mistake for the behavioral scientist to agree.”

The hypothesis that what ought to be (in the moral sense) can be inferred from what is was termed the naturalistic fallacy by English philosopher G.E. Moore (1873-1958). Obviously, Skinner did not believe it to be fallacy, and neither does E.O. Wilson: “I find it hard to believe that had Kant, Moore, and [John] Rawls known modern biology and experimental psychology they would have reasoned as they did…. Moral reasoning, I believe, is at every level intrinsically consilient with the natural sciences[6]” and “The empiricist argument, then, is that by exploring the biological roots of moral behavior, and explaining their material origins and biases, we should be able to fashion a wiser and more enduring ethical consensus than has gone before.”  In sum: “Ought is the product of a material process.” Note the reference to “material origins and biases”, which again points to a confusion between process and outcome: Understanding the historical process that led to a belief can justify a scientific claim, but not a moral one.

So what lesson does Wilson draw from science? Unlike Harris and (as we will see) Skinner, Wilson is not specific. His view is consequentialist, we judge “moral instincts…according to their consequences.” There are two problems with this. First, over what time period should we look? Should we judge the consequences today, this week, a hundred years from now? How should a good consequence now be weighed against a bad sequel 10 years down the road? How well can we predict remote consequences? And second, how do we tell good consequences from bad; in other words, what is “the good”?  Wilson does not answer this question directly. We can infer what he thinks is good from the things he calls bad: he dislikes xenophobia and what he calls “paleolithic egalitarian and tribalistic instincts”.  He advocates more research, assuming that the better we understand what human moral sentiments are, the better we will know what they should be.  Wilson is a natural scientist. The path he recommends may tell us why we do what we do; it can never tell us what we should do

B. F. Skinner defines “the good” in two ways. One is merely descriptive: “good” is just whatever society “reinforces” — rewards — or punishes. His more fundamental definition goes to the heart of evolution, survival of the culture and the species. “The ultimate sources [of values] are to be found in the evolution of the species and the evolution of the culture.” Perhaps “survival” is a value everyone can agree on. The problem is deciding just what will promote survival and what will endanger it. If “survival” is to be our guide we must be able to predict, at least in broad outline, the course of biological and cultural evolution.

Survival as the ultimate value

The assumption that evolutionary history is predictable is closely related to the doctrine of historicism, espoused most famously by Karl Marx. It was devastatingly criticized by Karl Popper, who wrote: “Marx may be excused for holding the mistaken belief that there is a ‘natural law of historical development’; for some of the best scientists of his time…believed in the possibility of discovering a law of evolution. But there can be no empirical ‘law of evolution[7]’”

There are also practical difficulties. First, looking to “survival” for answers to ethical questions will often point to conclusions that conflict with values that are now deeply held. Are we to abandon them? Second, there are very many cultural and genetic “fitness” questions that simply cannot be decided at all: the problem with “survival” as a value is that it provides little or no practical guidance in difficult cases.

A few examples should suffice to show that deciding on evolutionary “good” and “bad” is at least as difficult as predicting stock movements. For example, alcohol is a poison. Hence, cultures that use alcohol must be less “fit” (in the Darwinian sense) than cultures that do not. But are they? There might be hidden benefits to one or the other that we cannot now foresee. The Puritan consensus was that alcohol was an unmitigated evil. The social benefits associated with moderate drinking were assumed to be outweighed by its bad effects. Yet alcohol ingestion is a custom common to the majority of cultures, and now it turns out that there might even be health benefits to moderate drinking, so the evolutionary balance sheet on alcohol is not yet closed.

Another example: alcohol might be controversial, but smoking is certainly bad—isn’t it? This is not so clear either. Some smokers (by no means all) die from lung cancer and emphysema, usually in unpleasant ways, which is unquestionably bad for “human flourishing” as well as individual survival. However, smoking-induced illnesses generally do not kill until their victims reach their fifties and sixties, after their productive life is almost over and before they become a burden to their children and to society. It is an evolutionary truism that life history is determined by adaptive considerations, and a short but productive life is often “fitter,” in a natural-selection sense, than a longer and less productive one.

Perhaps a society that encourages smoking—which yields a generally short but productive life—will be more successful in the long run than one that discourages smoking and has to put up with a lot of unproductive old people? Should we perhaps encourage smoking? There are some data to support the idea. Several studies have shown that the lifetime health-care costs for smokers are actually lower than for non-smokers (public-health rhetoric to the contrary). Whether or not reduced financial cost corresponds to evolutionary advantage is of course not known, but an inverse relation between cost and “fitness” is perhaps more likely than not.

Argument from evolutionary survival very quickly comes up against many traditional beliefs. Even obvious virtues like safety and the emancipation of women, not to mention tolerance for anti-progenitive sexual abnormalities, might be questioned by a thoroughgoing evolutionary ethicist. Is it really adaptive to outfit 3-year-olds on tricycles with crash helmets so they grow up timid and unadventurous, or to fit our cars with air bags and seat belts so that the reckless and inept are protected from the consequences of their actions? And does it make evolutionary sense to encourage the brightest young women to delay, and thus limit, childbirth so they can spend the prime of their lives as physicists and investment bankers rather than mothers? Lee Kuan Yew, President of Singapore, thought a few years ago that it did not. He was pilloried for providing maternal incentives to well-educated women. But surely a conscientious evolutionary ethicist should applaud him?

The problem of what really conduces to “fitness” — of a culture or a race — has become especially acute with advances in medicine. Should parents be allowed to control the sex and other characteristics of their children? Should human cloning (which may have already happened) be permitted? What extraordinary measures are justified to keep a sick person alive? Kidney transplants, yes. Heart transplants, yes, perhaps—but what if the patient is already old or has other ailments? When should a sick person be allowed to die? What is the “optimal lifespan”? We know that lifespan is a subject to natural selection[8], so there must be an optimal—in the sense of most favorable to the continuation of the species—lifespan. What is it?  What if it is shorter than the current average in the West?

Politics are not immune from evolutionary optimality. What is the best political system? Most Americans assume that hierarchy is bad, and the American Constitution enshrines democracy and the rights of the individual. However, the most stable (i.e., evolutionarily successful) societies we know were not democratic and egalitarian but hierarchical and authoritarian. The ancient Egyptian culture survived substantially unchanged for thousands of years. The Greeks, the inventors of democracy, survived as a culture only for two or three centuries and were defeated by the undemocratic Romans, who lasted three or four times as long. The oldest extant democracy is less than 300 years old. In the animal kingdom, the termites, ants and bees, with built-in hierarchies, have outlasted countless more individualistic species[9].

The attempt to base values on evolutionary success very soon raises questions about traditional beliefs, albeit in an inconclusive way. The problem with “survival of the culture” as a value is that it requires reliable knowledge of the future. While some customs are clearly maladaptive under most imaginable circumstances, others are more contingent. The problem is that most of the prescriptions of traditional morality fall in the latter class. We simply do not know, belief by belief, custom by custom, rule by rule, whether or not our culture would, in the long run, be better off with or without them.

It is certain that some cultures will survive longer than others. It seems very likely, moreover, that the ones that survive will have many beliefs that were in fact essential to their survival. But the importance of at least some of those beliefs could not have been foreseen, even in principle. This is the fatal flaw in Skinner’s belief and E.O. Wilson’s claim that the fact of evolution allows all morality to be reduced to science. A comprehensive evolutionary ethics is impossible. Scientific imperialism is simply false.

Faith returns

Furthermore (and this will not please your average atheistical social scientist), the argument that demolishes evolutionary ethics also provides a rational basis for faith—although not, I hasten to add, for any faith in particular. The reason is not that a faith is true in the scientific or any other sense. The reason is that for a society to function at all, rules seem to be necessary, even in cases like the examples I have given, in which certainty is (and perhaps must be) lacking. We deter smoking, outlaw some drugs, emancipate women, tolerate the non-reproductive and preserve life at almost any cost, even though the evolutionary consequences of these decisions are unknown and probably unknowable. If rules must exist—even for situations in which science provides no clear basis for choosing them—then some other basis for choice is necessary. That basis is, by definition, a matter of faith.

The evolutionary approach to the problem of values promises more than it can ever deliver. Harris, Dawkins, Skinner, E.O. Wilson and most other scientific imperialists are confident of their own values and believe them to derive from science. Hence, they think the ethical problem easier than it is — which allows them to try to persuade us that values which are so obvious to them in fact flow from science.  It appears that their aim is not so much to understand the world, as to change it.

* * *

The issue should have been settled by David Hume in 1740: the facts of science provide no basis for values.  Yet, like some kind of recurrent meme, the idea that science is omnipotent and will sooner or later solve the problem of values seems to resurrect with every generation.

The science-based criterion most likely to achieve consensus, survival of the species or the culture, is impossible in practice because evolution is unpredictable. Embarrassingly, many practices that seem to favor survival are opposed to accepted Western values.

Cultures depend on practices and beliefs. We do not know which of them in fact promote survival and which do not. We do know that without at least some of them, no culture can long survive. We must have faith in some unprovable things, but science cannot tell us what they should be.

John Staddon is James B. Duke Professor of Psychology and Professor of Biology, Emeritus, at Duke University.  His most recent book is Scientific Method: How science works, fails to work or pretends to work. (2017)Routledge.”

Notes and references

[1] Junod, T. (2009, January 5). E. O. Wilson: What I’ve learned. Esquire. Retrieved from The quote also appears in his book Consilience (1998).

[2] Leon Zitzer (2017) A short but full book on Darwin’s racism (Universe), which is a summary of a much longer book accusing Darwin of racism. On the contrary, authors Adrian Desmond and James Moore (Darwin’s Sacred Cause, Allen Lane, 2009), trace Darwin’s whole evolutionary project to his hatred of slavery… go figure.

[3] Hume, David. David Hume Collection: A Treatise of Human Nature. (1738-40) Kindle Edition.

[4] Harris, Sam. The Moral Landscape: How Science Can Determine Human Values (p. 11). Free Press. Kindle Edition.

[5] Staddon, J. E. R.(2004) Scientific imperialism and behaviorist epistemology. Behavior and Philosophy, 32, 231-242.

[6] Wilson, E.O. (1998). Consilience: The unity of knowledge. New York: Alfred Knopf.

[7] Popper, K. R. (1950). The open society and its enemies. Princeton, NJ: Princeton University Press. Popper, K. R. (1962). Conjectures and refutations: The growth of scientific knowledge. New York: Basic Books.

[8] Stearns S.C The evolution of life histories. 1992, Oxford, UK: Oxford University Press.

[9] Russian anarchist Peter Kropotkin (1842-1921) thought bees a fine model for a human utopia.

14 responses to “Science and Morals: Can morality be deduced from the facts of science?

  1. Pingback: Values, Even Secular Ones, Depend on Faith: A Reply to Jerry Coyne - Quillette

Leave a Reply

Your email address will not be published. Required fields are marked *