Why I think Sam Harris is wrong about morality | The Righteous Mind

Several commenters  have said I should not just critique the excessive certainty of the New Atheists. I should respond directly to Sam Harris’s Moral Landscape Challenge. I should say why I think the argument he makes about a science of morality are wrong. (Harris argues that what is right and wrong can be determined scientifically, just as we can determine truths in the natural sciences). Fair enough. So this morning I submitted the following text as my entry in his challenge.

via Why I think Sam Harris is wrong about morality | The Righteous Mind.

I didn’t realize that Jonathan Haidt had a blog!  Since his latest post is relevant to our discussions in the last couple of weeks, I thought I’d highlight it here.

I learned a new term I’ll have to remember, since it expresses a concept I’ve only clumsily been able to articulate before.  There are anthropocentric facts, such as sugar is sweet, and there are non-anthropocentric facts, such as metal is conductive.

I’d also point out that anthropocentric facts can be further divided into facts of human cognition, such as market prices, and facts of human evolution, such as sugar being sweet.  I think it’s important to understand that moral facts have grounding in human evolution, not just human cognition.  But as Haidt said, they are still anthropocentric facts.

25 thoughts on “Why I think Sam Harris is wrong about morality | The Righteous Mind

  1. Hi SAP,

    I entered the Moral Landscape Challange myself. Haidt’s essay is well written but I see some arguments against it.

    Haidt’s problem 1 is vulnerable to Harris’s analogy to medicine. We can’t objectively measure health but we don’t have a problem with a science of medicine. I think Haidt is correct here but without answering Harris’s usual response to such criticisms Harris will be unmoved.

    Haidt’s problem 2 is, I think, a little off base on two fronts. Firstly, I’m not sure that Harris would agree that he has claimed that moral values are non-anthropocentric facts. In fact, i think he would agree that they are anthropocentric, but he would disagree with Haidt about whether science can determine anthropocentric facts. Secondly, I don’t think moral values are even anthropocentric facts. I don’t think moral values are facts of any kind, save perhaps facts about what people in specific cultures view about morality. Haidt’s anthropocentric fact idea seeks to establish within moral relativism a kind of moral objectivity which I don’t think exists.


    1. Hi Disagreeable,
      My issue with Harris’s medical analogy is that it doesn’t tell us what he implies that it does. What medical study ever studies “health”? None that I know of. They instead focus on some specific condition or set of conditions that can be defined. Which condition to study is a value judgment, and often a controversial one. (Homosexuality, alcoholism, and attention deficit disorder spring to mind.)

      If sciences can’t determine anthropocentric facts, then it seems like we’d have to throw out psychology, sociology, anthropology, economics, and any other social science. Certainly these fields can’t produce the reliable conclusions of the natural sciences, but they still produce useful insights. To conclude otherwise is to ignore a lot of history.

      I can definitely understand where you’re coming from on moral facts. I went through a stage of thinking the same thing. But pure cultural relativism assumes that we start as blank slates, and there is a lot of scientific work demonstrating that we are born with just as many, if not more, instincts than most other mammals.

      Best of luck on your essay! I know Harris plans to publish the best one and award the author $2000 ($20,000 if you convince him).


      1. Hi SAP,

        Your views on medicine are close to my own. My argument was that medicine is the study of how to manipulate human biology to attain certain goals, but cannot tell us which goals are desirable, e.g. whether we ought to choose longevity or quality of life, given a trade-off. But if Haidt doesn’t address this then his point is redundant as it’s been done to death and Harris is well aware of this argument.

        My point about anthropocentrism was that Harris would surely agree with you that science can determine anthropocentric facts, so he will not be impressed by an argument that morality is anthropocentric.

        I’ve read the Blank Slate and I’m under no illusion that we are born as blank slates. I don’t see how that leads to Haidt’s objective anthropocentric morality. I think that many of our moral instincts are undesirable, in particular the urge to sanctify, to be pure, and to hate the outgroup. I’m a moral relativist but not in the normative sense. I own my moral preferences and will fight for them if necessary. I judge others according to these preferences, but I don’t pretend that they are objectively correct.


        1. I suspect Harris will have already heard all the arguments that will be presented to him. Of course, whether he has understood them, or chooses to understand them, is a different matter.

          While Haidt does say that anthropocentric facts can be measured, he’s careful to clarify that he’s talking specific things, and that philosophical discussion is necessary to determine what should be measured. It’s actually the fact / value distinction, the is / ought divide, something I’m sure Harris will here about in a lot of the essays. Harris, of course, is definitely aware of this issue, but impatiently dismisses it as confusion. Exactly who’s confused is a matter of debate.

          Scanning Haidt’s essay again this morning, I don’t see him claiming that moral facts are objective anthropocentric facts, but the narrower “emergent culture-specific anthropocentric truths.” His wording is careful here, to indicate that morals are relative, but not in a sense of existing in a vacuum, but arise from lower level biological facts.

          I agree with you. I’m effectively a descriptive relativist, but not a normative one, although my understanding of how culture influences values makes me very cautious about judging others.


          1. Hi SAP,

            I think where I take issue with Haidt’s description of values is his comparison to the value of gold or silver. I think these really do have objective anthropocentric value, at least in some sense. It really is true that gold is worth more than silver. It is clear what this means given a basic grasp of economics, even if we understand that this is not intrinsic to the metals but an emergent property of human civilization on planet earth.

            It is not as clear to me what it means to say that “abortion is wrong”. If it works similarly to market value, it must be some way of aggregating the views of all humans, something like “the majority of humans feel that abortion is wrong”.

            But I categorically reject that what is moral depends upon upon majority views. What the majority hold to be moral is sometimes odious. I prefer to deny any kind of objectivity at all and instead hold fast to my own moral principles.


          2. I think you’ve centered on the main rub. Each of us has a visceral feel of what is right and wrong. Luckily, there is a huge overlap among most of us, making societies feasible. But none of us exactly matches the overall consensus. We can try to advocate people to our viewpoint, but what brought us to that viewpoint won’t always bring them to it.

            In a diverse society, how we respond to this situation, I think, is almost as important as our actual sense of morality. How we handle our meta-morality, that is, morality about morality, is important. What variances can we tolerate among others, and what is intolerable?

            You say that you don’t like the market value approach. Intuitively I agree. Unfortunately, I think we’re stuck with it. Moral propositions can only be advocated for (or imposed), not proven.


          3. I agree with you, so I think you may slightly misunderstand my position.

            I think the market value approach is the best practical solution for how to manage disparate viewpoints without descending into violence. My problem is that I do not accept that I am wrong in any sense if I disagree with the consensus. I do accept that I am objectively wrong if I believe that silver is worth more than gold.

            But belief in democracy has limits. Were I granted absolute power, I would feel a moral obligation to impose my most strongly held views on others whether or not these are majority views. If I’m not willing to do that, then I don’t think I really hold the views I claim I do. The kinds of things I would do would be to grant equal rights to LGBT people and women everywhere on earth, to give religious freedom to all and to ban the teaching of creationism as science. I would seriously consider doing something about phasing out meat-eating.

            (I’m not actually a vegetarian because of practical issues but I am leaning towards thinking that meat-eating is probably immoral.)


          4. It does sound like we’re mostly on the same page. I’m not sure I’d agree though that just because I wouldn’t be willing to impose my personal morality on others, that I don’t really hold it. I have a visceral aversion to ridiculing anyone, so much so that it usually makes me uncomfortable to watch anyone else do it, but I wouldn’t necessarily impose that stricture on anyone else.


  2. I may agree with division of facts into anthropocentric and non-anthropocentric. But division of anthropocentric facts into “facts of cognition” and “facts of evolution” seems like a stretch. First of all, Haidt erases the limit between facts and opinions. Second, you use the name “cognition” whereas he uses the name “emergent culture-specific anthropocentric truths”. It’s not quite the same.

    I’m not sure if I would accept opinions as “facts”. Actually, drawing the line between opinions and facts might be a better way to refute Harris.


    1. It’s not clear to me that Haidt does erase the fact / opinion divide, unless you consider “sugar is sweet” to only be an opinion.

      I agree that my “facts of cognition” and “emergent culture-specific anthropocentric truths” aren’t quite the same thing, although for most discussions they are close enough to effectively be the same. The point I was trying to make, and still apparently being clumsy about, is that there are anthropocentric facts that we decide on, and anthropocentric facts evolution has handed us. Morality has foundations in the latter, although it’s heavily shaped by the former.


      1. “Sugar is sweet” or “sky is blue” can be called “antrhopocentric facts”. But I am not so sure about judgments of humor or deliciousness or sexiness and, to extend, happiness, well-being, etc. Often, musicians or artists are “better” not by virtue of being “better”, but by virtue of heavy marketing or being noticed by the right people in the right time. Popularity is a poor measure of quality.

        I can also see how statements like “God exists” can be smuggled into the category of facts this way. We need a better distinction criterion between facts and opinions, don’t you think?


        1. I agree that not all judgments are facts. I’m a Doctor Who fan, but that is a preference, an opinion of mine that Doctor Who is a good show. It’s a fact that I’m a Doctor Who fan, but not a fact that Doctor Who is good (although it feels like one to me 🙂 The fact of my preference, my opinion, and what causes it, can be studied, but not the rightness or wrongness of the opinion itself.

          You actually highlight something I think about from time to time. What is a fact? It’s not a consensus opinion since the consensus is often wrong. It’s not an authoritative opinion, because authorities are also often wrong. Is it an expert opinion? But what makes an expert? Maybe a fact is an assertion that, our our judgment, it is unlikely to be proven wrong. It’s hard to eliminate judgment from this equation.


          1. The fact of my preference, my opinion, and what causes it, can be studied, but not the rightness or wrongness of the opinion itself.

            In the same way, we can study what causes certain acts to be considered morally wrong or write based on evolution or neuroscience — I can agree with Harris here. But we cannot judge moral rightness or wrongness using science. E.g., there are studies showing that religion reduces the suicide rates. But studies cannot show that suicide is “bad”.

            There is “something” in Haidt’s distinction between anthropocentric facts, but I think, this “something” needs more work. As-is, the idea does not go beyond intuitive feeling for me.

            Maybe a fact is an assertion that, in our judgment, is unlikely to be proven wrong.

            Then Harris’s “moral landscape” is a fact. 🙂 (for him). What is a fact is a great question to explore. I’m not clear about it myself.


          2. Another thought. The “facts” about state of our mind are different from the facts of matter. Consider statements of intention, for example. It may be considered to be a fact that I intend to hit “Post Comment” button when I finish writing this. But I can change this “fact” on a mere whim. Even if you read this post which can be construed as evidence that I did hit the “Post Comment” button, after all, it cannot serve as evidence that I had intention to do so, because I could have clicked on the button by accident (if a concept of an accident makes sense at all if we accept deterministic world view). Although, courts may consider ways of proving or disproving people’s intentions, I don’t think, these ways can be called scientific.

            Harris seems to be confused about many things. But he does not think so. It’s funny that being wrong does not feel any different from being right. Following a wrong path in a forest does not feel any different from following the path you want to follow if you are sure that you follow the right path.


          3. Maybe a better labeling of these different domains of fact are social, biological, and natural. Recognizing of course that social facts emerge from biological ones, which in turn emerge from natural ones.

            In my experience, those who feel certainty about something are usually just not aware of how uncertain their position is. I perceive levels of certainty. Scientific facts have a high, but not perfect, certainty. Philosophical conclusions are lower on the scale. Personal subjective assumptions are the lowest of all. Part of wisdom seems to be recognizing where on that scale a particular notion that you hold resides.


          4. The term “social fact” crossed my mind when I thought about different kinds of anthropocentric facts. I think, “social fact” (gold is more valuable than silver) and “biological facts” (sugar is sweet, sky is blue) may be a good way to categorize these concepts.

            But it’s a mere convention. Even though we may adopt this convention, it does not mean that social facts cannot be proved by science — we merely define social facts as such which cannot be proved by science. It’s easy to prove someone wrong “by definition”. But Harris needs to agree to our definitions to be convinced, of course.

            Certainty and uncertainty is a separate discussion. I’d like to reflect on it some time later. Decision making in uncertain situations is a very fascinating topic. “The only constant thing is change.” Paraphrasing Heraclitus, “the only certain thing is uncertainty.”


          5. I actually didn’t mean to imply that a social fact categorically wasn’t determinable by science. I think we’re agreed that what values are held, and what leads to them being held, is scientific, but whether those values ought to be held, isn’t. I think Harris’s confusion is his belief that the is/ought divide is confusion.


          6. Yes, I think, we are in agreement here.

            I think Harris’s confusion is his belief that the is/ought divide is confusion.

            That’s the thing. Seeing where confusion lies is prerequisite to getting out of it. Just as acknowledging ignorance is the first step to knowledge. But when a person is confused about whether he is confused, I cannot help but quote the Bible:

            Do you see a person wise in their own eyes? There is more hope for a fool than for them.



  3. Forgive me if I’ve completely missed the point (I suspect I have), but …

    From what I gather, Sam Harris argues that morality is about “the well-being of conscious creatures” which he says can be measured. This sounds to me like happiness. Others have criticised the problems in measuring happiness (short term vs long term) even assuming that you can measure it reliably. But morality is about many things, not just happiness, which might be relatively low in importance. Concepts like justice, fairness and honesty feature prominently, and I don’t know how they can be measured.

    Reading a powerful book or watching an interesting TV show can make people feel sad. Are such works of art immoral in Sam Harris’ universe?


    1. He specifically says “well-being” to circumvent those criticisms. There surely are ways that sad TV shows add value to our lives. Whatever those ways are, Harris would define well-being so as to take account of them.

      He doesn’t advocate one particular definition of well-being, in particular because this is so difficult (or perhaps impossible). His point is that, whatever well-being is, it is a property measured over time of the mind of a conscious creature, and on naturalism this must be an empirical fact that can in principle be measured.

      I think he’s right for any particular definition of well-being, but I am not optimistic that we will ever find a definition that all agree with.


  4. So 1) we would need to agree objectively on a definition of well-being, 2) we would need a way to measure it objectively, and 3) we would need some way of predicting in advance which action would most increase this measure?

    I wouldn’t bet a lot of money on that being a successful route to a moral code.


Your thoughts?

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.