• Science Has Nothing To Say About Morality


    In a short excerpt from an interview, published yesterday, Michael Shermer claims that science can somehow weigh in on issues of morality. I disagree – science has nothing at all to say about morality.

    I’ll begin by providing a scenario that invokes a relatively easy moral question:

    You are late for a very important interview for a very well-paid job. If you miss it, you’ll certainly not get the job. You are driving to the interview and come to a fork in the road. As you are about to take the right turning, you witness a pedestrian being hit by a car in the other direction. The car flees the scene, and there are no other people or cars around. The hit pedestrian looks to have life-threatening injuries.

    The problem is this: if you help the injured person you won’t get the job. If you leave them then they are more likely to die. Should you sacrifice your job prospects to save the injured pedestrian?

    I take it that practically everyone will say that the moral thing to do is to sacrifice the job. To ignore the injured person would be immoral. Of course, we may not all actually help them, but I’m fairly sure that even those who don’t would concede that they aren’t performing the most moral action.


    Now I’ll take two key quotes from the interview that I think best illustrate Shermer’s position. The first:

    Religion and philosophy have had their say for thousands of years what’s moral and not moral. I’m just saying let’s add to our quiver the arrow of science to bring to it studies and data and experiments and research to see what works and what doesn’t.


    I’m puzzled by “see what works and what doesn’t”.  In the case of the injured pedestrian, what is it that “works”? Ignoring the pedestrian “works” best to get me the job. Helping the pedestrian “works” to give them the best chance of survival. What we mean by “works” adds nothing at all to the moral question – it may only be of importance once we’ve already decided what the answer to the moral question is.

    What can science say about what the most moral choice is, in our example? Nothing at all, as scientific questions concern how the natural world is, or was, or even will be, and not how the world ought to be or how we ought to act. This is the famous is/ought distinction.

    So how do we decide what is the most moral action to take? Religion? Well that actually is one possibility. If a tenet of a particular religion is that we should help people in need, over our own selfish interests, then that might be a good moral reason for adherents to help the injured pedestrian. Of course, there are well-documented difficulties with this, but then again there are with lots of theories in moral philosophy. Is the reason that adherents of this religion act morally that they are trying to appease their god? If so, are they really acting morally, or are they doing what “works” with regards to their own standing with that deity, heavenly rewards, etc.?

    What about secular philosophy? Well let’s just take one view, utilitarianism. On this view we look at the consequences of our actions – notably the happiness or suffering they will cause. The injured pedestrian is suffering, and so the suffering prevented by helping them will outweigh the happiness we are denied by losing out on the lucrative job (yes there are difficulties with this analysis, but I’m just trying to give the basic idea). Shermer at this point might say that the scientific method can be employed to test exactly how much happiness is gained by well-paid jobs, or how much suffering is prevented by assisting victims of a hit and run incident. But the moral work is done by defending a particular view of ethics, such as a certain flavour of utilitarianism, and the only way to do so is with philosophical arguments. The scientist who argues for utilitarianism (say) is doing philosophy, not science.


    Here is the other quote from Shermer:

    …religious morality has no means of testing whether its claims are true or not. Take something like female genital mutilation. Can we do better than to say it’s wrong in America but it’s OK in African countries? I think yes, we can base our moral values in scientific reasons, starting with the individual. [Female genital mutilation] is wrong because it robs the individual of dignity, autonomy and the core of her nature.


    Shermer talks of dignity, autonomy and the ‘core’ of a woman’s nature. But those aren’t “scientific reasons” at all. What’s the scientific test for ‘dignity’? What does science say is the ‘core of a woman’s nature’? The best attempt of the latter question might be some biological description. But what does that offer us? Appealing to nature to decide moral questions (sometimes mistakenly called the ‘naturalistic fallacy’) is well-known to be a rather indefensible position.

    Like our adopting utilitarianism to answer the moral question in the injured pedestrian example, Shermer’s claims, first that FGM undermines dignity, autonomy, etc. and second that whatever does that is immoral, are philosophical claims. Try to defend a definition of ‘dignity’ without providing any philosophical arguments at all. Try to devise a scientific test that shows that whatever robs an individual of their dignity is immoral. I submit that it is a priori impossible to do so.


    Also see this post by Brian Earp, talking about Sam Harris’ position (which is very similar to Shermer’s).


    Category: Ethics

    Article by: Notung

    I started as a music student, studying at university and music college, and playing trombone for various orchestras. While at music college, I became interested in philosophy, and eventually went on to complete an MA in Philosophy in 2012. An atheist for as long as I could think for myself, a skeptic, and a political lefty, my main philosophical interests include epistemology, ethics, logic and the philosophy of religion. The purpose of Notung (named after the name of the sword in Wagner’s Der Ring des Nibelungen) is to concentrate on these issues, examining them as critically as possible.

    One Pingback/Trackback

    • Even in your example, science (or at least the facts) is critical to the choices you might make. Can you help the woman, is it possible? How do you know? From reason and experience. A person can at times save another’s life in those situations. That’s not emotion, that’s a fact. The person may die without help. That’s another fact. If you do not know these facts, you can not make an appropriate moral judgment about the situation.

      I think that is Shermer’s point, or part of it. Many moral questions are utterly meaningless without knowledge. Is drug X ethical to prescribe? Well, to know that one needs to know all side effects and main effects. Is one progressive tax scheme better than another? Freon and trans-fats were considered ethical goods when they were first used because CFC’s were non-reactive in the environment and therefore not considered pollution and trans-fats were plant based, and therefore much healthier than those nasty saturated animals fats. Only *knowledge* changed them from good choices to horrible ones.

    • Yep, I agree with that, but I don’t think it contradicts my position. I suppose we might say that there’s a moral ‘component’ and an epistemic ‘component’, and you need both when deciding how to act. What I’m trying to say is that science doesn’t inject any morality (if I can speak of it as if it’s a sort of ‘stuff’!) into our decision.

      Take abortion for instance. If we want to know whether a fetus feels pain during the abortion process, we can devise some sort of scientific test, whether this is behavioural or neurological. But that doesn’t actually tell us what we ought to do. We need some philosophically argued-for ‘moral premise’ such as “whatever causes pain is immoral” (crude I know). Then that, coupled with our scientific knowledge about fetuses will tell us how we ‘ought’ to act. But the scientific facts on their own are just brute facts; they aren’t moral claims in any sense. The moral ‘component’ comes purely from philosophical reasoning. The question of whether or not fetuses feel pain is independent from the question of whether we ought not cause pain. The former is a question of is, the latter ought.

      I hope that explains my position a little better. Of course knowledge of empirical facts is necessary act morally (cf. W.K. Clifford’s famous argument). But what causes or helps people to act morally isn’t the same as what is moral.

      So I still contend that science has nothing to say about morality, even though it is necessary for morality. Is this inconsistent? I say no, since the necessary science is interpreted through a ‘moral framework’, which is wholly provided by philosophy. Philosophy ‘tells’ us (in a manner of speaking) what is moral, and ’employs’ science to fill in the other premises that allow us to make a moral claim about the real world.

    • RussellBlackford

      I agree with you that we can’t go straight from an is to an ought without some premise about what we are trying to achieve (or using some kind of logical trick that won’t help in the real world, such as arguing from factually inconsistent premises). It might be a moral premise or it might simply be a value or a desire (“I want to reduce suffering”). But that doesn’t mean science, insofar as it can tell us something about how what we are trying to achieve can actually be achieved, has nothing to say about how we ought to act. It simply doesn’t have everything to say. In the end, the facts revealed by science cannot motivate by themselves.

      I’m not sure why you express it as “science has nothing to say about morality”. I largely (I have some reservations) agree with the position you’re developing, but that seems like an odd way for you to label it.

    • I think he may be misusing some of the terms; he is certainly using philosophical reasoning in the interview. I take it as rebuttal of the age-old religionist bigotry that religion owns morality and that science is impotent to play any role in ethical considerations. It turns out that empiricism, reason, and knowledge, some of the fruits of science, are massive multipliers of ethical and philosophical progress.

      That is a point worth making, that without religion we can just as ethical. Without science, we’d be ignorant fools, thrust backward in time.

    • I label it that way mainly because I think it’s quite punchy and controversial – and yet I do feel that it is true. I think that scientific considerations inject exactly zero normativity into moral questions, and it is normativity that gives morality its character.

      My title might be read as ‘scientific knowledge should be disregarded when answering moral questions’ but that isn’t what I’m claiming. It’s more to do with asking ‘what exactly is it that science is contributing to moral questions’? Science says lots about how the world is, and nothing about morality. Philosophy can argue for what is moral. Further philosophical inquiry brings these two (necessary) components together to construct an argument for why such-and-such an action is or isn’t moral.

    • Oh, no question about that. His position is a philosophical one, and I agree with him about religion.

      I see science becoming especially useful for issues in Bioethics, since we need to know exactly what we’re talking about when we say ‘we ought to do such-and-such’, and these issues are usually highly nuanced.

      However in my dying pedestrian example, I don’t think we need ‘science’ (in the usual sense, unless we define it very broadly to include all rational empirical beliefs) to determine that the suffering of the victim is greater than the happiness of my future in a well-paid job.

      But point taken – he’s having a pop mainly at religion, not philosophy.

    • Coel

      What can science say about what the most moral choice is, in our example? Nothing at all …

      I disagree there. Science can tell you that there is no “most moral choice” because there is no objective or absolute standard of morals. Science can tell you that all there is is people’s feelings and opinions about what is moral. Thus asking “what the most moral choice is” in the abstract is not a sensible question, it literally has no meaning.

      Now, “What do most people consider the more moral choice” in your example is meaningful and does have an answer, and science can tell you that answer. Thus I disagree with you, as I explain more fully here.

    • I don’t agree with the first part – I don’t think science tells us that at all – I think that position is a philosophical one. After all, what scientific test could determine whether there is such thing as objective morality? (By the way, I don’t believe in objective morality either).

      I suppose science could answer what in fact people think about morality (in the form of a poll, I guess), but that isn’t answering a moral claim. Even if statements about morality are meaningless (or translate into a preference), as the logical positivists held, that would still mean that science can’t determine ‘what is moral’ since ‘what is moral’ isn’t meaningful.

    • Coel

      what scientific test could determine whether there is such thing as objective morality

      Science explains what morality is, why we have morality. It is a set of attitudes and feelings programmed into us by evolution to facilitate our highly social and cooperative way of life. That understanding tells us that there is nothing absolute or objective about morals (in the same way that aesthetic opinions are subjective, not absolute).

      From there the onus is on anyone believing in absolute morality to establish an absolute morality and explain what it means. No-one has.

      science can’t determine ‘what is moral’

      That’s because there is no abstract absolute morality, and thus claims about it have no answer. Science’s answer, that there is no objective absolute moral scale, and thus no answer to such questions, is an entirely valid and complete answer. The fact that it doesn’t mesh with human intuition is a defect of human intuition, not a limitation on science.

      Even if statements about morality are meaningless

      Only statements which imply or refer to an absolute moral standard are meaningless (since there is no such standard), statements about morality that recognise that morals are feelings and opinions of humans are entirely meaningful.

    • But even if (and I concede this is what we’ll find) we analyse the brain of every person when they are making a moral choice and find that it is wholly a product of feeling and preference, that doesn’t mean that’s what morality is, unless we’re defining it as that. But that’s not how most philosophers and indeed most people define it. ‘What is right?’ is a logically different question to ‘what is the cause of my thinking this is right?’. Perhaps the former question (which can involve either objective morality or subjective morality) is meaningless, but either way, science can’t answer it. Science can answer the latter one, but I don’t take that to be what ‘morality’ is. Of course, if we do we open ourselves up to tricky questions like ‘what if everyone thinks fascism is good and moral?’ and so on.

      Subjective morality doesn’t mean that morality is identical with the cause of our thinking such-and-such an action is the moral one – rather it means that morality is somehow mind-dependent.

      And so science can’t tell us what is moral whether morality is subjective or objective.

      Anyway, I’m enjoying the discussion!

    • Coel

      even if … we analyse the brain of
      every person when they are making a moral choice and find that it is
      wholly a product of feeling and preference, that doesn’t mean that’s
      what morality is, unless we’re defining it as that.

      Let’s make an analogy with vitalism or notions of a dualistic soul. A vitalist could argue: “even if we analyse life and find that materialism is entirely adequate to explain everything about life, that doesn’t mean that that’s what life *is*, unless we’re defining it that way, vitalism could *still* be correct”.

      To me that response is empty: through science we understand enough about life to dispense with vitalism and declare it false, despite having no absolute proof of the non-existence of elan vital.

      I submit that we’re in the same position over morality. Through science we now understand enough about why we have moral sentiments that we can dispense with notions of “absolute/objective” morality, just as we can dispense with vitalism.

      There really are no good arguments any more for an objective morality (I even have no conception of what “absolute” morality would even mean). *Everything* about morals just makes so much more sense when you dispense with the idea of morality being objective and absolute.

      Once we understand that our moral system is just a system cobbled together by evolution for its utility, just like our aesthetic system or our immune system, then puzzles that have baffled philosophers for aeons just disappear. There is nothing about our immune system or our aesthetic system that we cannot study scientifically, and ditto for our moral system.

      The only reason that our intuition leads us to *think* that our moral judgements reflect an absolute standard is because evolution has programmed us to think that way, because it makes us feel that our morals matter more and this it makes our moral system work more effectively.

      Once we accept that human intuition is fairly fallible we’re left with absolutely no reason to suppose that there is something objective about morals, nor any Absolute Shouldness Scale.

    • Russell, what’s so wrong with the additional premise that we value human flourishing? That seems a reasonable one since, after all, we want happiness, holistic happiness. And why is it a requirement that an ethical commitment to human flourishing should be embraced by everyone before it can be accepted at all? People disagree. So? The sociopaths, the Freudians with their death wish, the selfish, and the criminals don’t count, since after all, they are hurting themselves first, and as such cannot really be holistically happy.

    • Take for example human freedom and democracy devoid of religious demands for a theocracy. It’s the wave of the future. When people can vote with their feet they go to a democratic country. I would have to say the evidence from this human propensity is in. We want freedom of conscience, of speech and the right to engage in the political debates. Is this scientific evidence? Why not?

    • “‘what exactly is it that science is contributing to moral questions’?”

      I think it provides all the knowledge you need in order to achieve a moral end. You want to diminish suffering? Fine, science can give you anesthesia, tell you where and when to apply it and the dose. It tells you when a fetus starts feeling pain (about week 25, as far as I know), it tells you you probably shouldn’t run above the speed limit nor crash into another car when you’re driving someone to the hospital, the tools of the scientific method allow you to choose the best route to get faster and safer to the OR.

      All of this is knowledge and it ehances your performance when reaching for moral goals. And science can be described as reliable knowledge.

      It is contributing a lot.

    • I agree with that, but my point isn’t that science is useless – in fact it is invaluable. As I said in another comment, scientific knowledge is necessary to make effective moral choices. My point is that none of the normativity inherent in moral questions can come from science, by their very nature.

    • A vitalist could argue: “even if we analyse life and find that materialism is entirely adequate to explain everything about life, that doesn’t mean that that’s what life *is*, unless we’re defining it that way, vitalism could *still* be correct”.

      Actually, although I disagree with vitalism I think I’ll bite the bullet and say that your vitalist could in principle be right, although given the way ‘life’ is used by philosophers and scientists I’d say they were wrong. However when we talk about moral claims in philosophy, we are talking about normative claims. “Jane believes that we ought not kill” isn’t a normative (prescriptive) claim at all, but a positive (descriptive) one (and the kind of claim you are calling a ‘moral claim’), while “we ought not kill” is a normative claim, not a descriptive one. My view is that the normativity in moral claims is not imparted in any way by science.

      However if I was to define morality as you do, I’d certainly agree that can all be known by science.

    • Coel

      My view is that the normativity in moral claims is not imparted in any way by science.

      I entirely agree with you. And the reason why it cannot be imparted by science is that “normativity in moral claims” is a meaningless and incoherent concept. Just because we have an intuition about absolute morality does not make it a valid concept.

    • I featured this on my blog. I hope I’ve done you justice. It’s a great piece and it has me thinking a lot about cutting my own bias toward utilitarianism.

    • RussellBlackford

      John, answering all those questions would require a book (in fact, I hope to write that book some day!). Just to answer a couple of them, for now, “human flourishing” is a rather fuzzy concept, so it will give us only fuzzy guidance. There’s nothing wrong with that. You and I probably largely agree and what it is and both value it. But what if some other rational being doesn’t value it – perhaps a rational Martian? In that case, this being is not making a mistake about the world, it is simply not guided by the same value as we are.
      In the upshot – sure, there are values that people can more or less agree on, even though they are fuzzy, and these values can be action-guiding. I don’t think I’ve ever argued against that position. My argument with Sam Harris, which I think you’re referring to, is that he doesn’t seem to acknowledge fuzziness and he doesn’t seem to understand that morality can be non-arbitrary without being strictly (in a much stronger sense than is used by lawyers) objective.

      I realise that doesn’t deal with all your questions. Maybe another time! But meanwhile, I think it’s important that we recognise that morality does not have to be either fully determinate or objective in the strongest sense to be non-arbitrary and useful. So often, we’re presented with the false dilemma that morality is either entirely determinate and strictly objective or it’s all completely arbitrary, crudely relativist, or some such thing.

      (And, of course, I don’t think our other value judgments, such as judgments of non-moral goodness, or of beauty, or of literary merit, are fully determinate and strictly objective. Most people would probably concede this. It does not follow that they are useless or arbitrary.)

    • Thanks very much!

      I’ve commented on your post.

      I’m not criticising utilitarianism – I’m saying that utilitarianism (and other positions) are philosophical, not scientific ones. I think a lot of those who claim that science tells us things about morality assume something like utilitarianism, use science to find out some relevant non-moral facts, then claim that science was really doing the moral work. From what I can gather Sam Harris does that, but I may be wrong since I haven’t read his book (which I don’t consider a high priority).

    • Sounds complicated Russell. A few years back before Sam Harris’s book, I had asked some of the same questions to Stephen Law. He responded similar to you saying it requires a book. Damn it all, when will one of you write it. 😉

      Still waiting…

    • Michael R

      Science can measures states of happiness and therefore, in theory, it can say which is the most moral (desirable) course of action for an average individual or group in a particular situation. But in practice, who knows how far such science can take us? It’s hard to imagine it offering practical answers.

    • Coel

      Science can measures states of happiness and therefore, in theory, it can say which is the most moral (desirable) course of action …

      In saying that you have assumed that the “most moral” course is that which increases happiness, and that is begging the entire question.

    • Pingback: The science of homosexuality | Incredulous()

    • Michael R

      Coel, on your blog you said: “The only anchor for “morality” is in the feelings and emotions of humans”. I agree, but I don’t see how that differs from happiness. The only thing science can (potentially) measure is states of emotional fulfillment. How is that different from happiness? It’s the same thing to me.

    • Coel

      I agree, but you can’t say: “science … can say which is the most moral” in the abstract, because there is no such thing as an abstract “most moral”. You can only makes statements about what people opine to be more moral.

      In the same way you can’t say what “tastes best” in the abstract because dung beetles and humans have different tastes, so you can only talk about what “X considers to taste best” where you need to specify who X is.

    • Michael R

      “You can only makes statements about what people opine to be more moral”. Yes, and that’s what I intended to say by qualifying moral as “desirable … for an average individual”.

    • Nina

      Actually, that does wholly contradict your position, and the title of this piece.

    • What about my explanation as to how science being used in moral decisions does not mean that science is saying anything about morality would you say I got wrong?