Sam Harris is wrong about science and morality

By Brian Earp (Follow Brian on Twitter by clicking here.)

WATCH MY EXCHANGE WITH SAM HARRIS AT OXFORDON YOUTUBE HERE.

I just finished a booklet by “New Atheist” Sam Harris — on lying — and I plan to write about it in the coming days. But I want to dig up an older Harris book, The Moral Landscape. Why? Because it still makes me grimace.

I say “still” because I read the book months ago. I just haven’t yet vented my bafflement. Permit me to gripe, then, about Harris’ (aging) “bold new” claim — presented in his book — that science can “determine human values” or “tell us what’s objectively true about morality” or “give us answers about right and wrong” or however else you package this fiction.

In his new book (the one about lying) Harris says, in effect, you should never, ever, do it — yet his pretense in The Moral Landscape to be revolutionizing moral philosophy seems to me the very height of dishonesty. What he actually does in his book is plain old secular moral reasoning — and not very well — but he claims he’s using science to decide right from wrong. That Harris could be naive enough to think he’s really bridged the famous “is/ought” chasm seems incredible, and so I submit that he’s exaggerating* to sell books. Shame on him.

*A previous version of this post had the word “lying” here, but I was told that my rhetorical flourish might be interpreted as libel. I hope “exaggerating” is sufficiently safe. Now onward to my argument:

I’ll start by saying what the “is/ought” divide is, in case you haven’t heard of this before. It’s an old idea, tracing at least to David Hume, and its gist is that there is no way to reason from facts about the way the world is, to statements about the way the world should be. You can’t derive values from data. I’ll use one example to illustrate and then move on.

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

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

Now: if there were a way to get from “is” to “ought” it would take a work of philosophical genius to lay it out, and Harris’ book is not a work of philosophical genius. I can summarize his argument in a few lines:

1. Morality is “all about” improving the well-being of conscious creatures.

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

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

Here’s the problem. Premise (1) is a philosophical premise. It’s not a fact of science, it’s not a fact of nature, it’s not derivable from science, it’s not derivable from nature: it’s a value judgment. You might think this is a good premise; you might not – and even if you think it’s basically on track, there’s a lot of philosophical work to be done to spell it out. (Exhibit A – how do you define well-being in the first place, “scientifically” or otherwise?)

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

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

Maybe. But if this is what Harris wants to say, the examples he comes up with are weak. Such as? How about the Taliban. Harris says that according to science, the Taliban’s treatment of woman (enforced burqa-wearing, etc.) is objectively morally wrong. Why? Because enforced burqa-wearing is not conducive to the well-being of conscious creatures, namely the conscious creatures forced to wear burqas.

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

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

By the way, Sam Harris came to Oxford several months ago to give a talk about The Moral Landscape called, “Who says science has nothing to say about morality?” This particular talk was hosted by Richard Dawkins. To kick off the Q&A, Dawkins pressed Harris on just what he was saying that was new. Here’s a bit of that conversation:

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

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

So by “science” Harris evidently means, “philosophy” … or at least something that’s not different from philosophy in a principled way. Let me check my brochure for a second and confirm what the title of his talk was — the radical-sounding title that sold so many tickets — yes, here it is, it’s, “Who says science has nothing to say about morality?” If we do a quick update based on Harris’ personal definition of science, we get … “Who says philosophy has nothing to say about morality?”

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

I’ll close on a personal note. I was in the audience at Harris’ Oxford talk, and during the Q&A I nudged him on two points. First, how exactly did his argument get us over the is/ought divide; and second, what can “science” tell us about morality that we didn’t know from common sense. Our exchange can be seen in the video below, and I’ll make just one comment before you watch it. Notice the first four words of Harris’ reply to my question: “The moment you grant …” My point has been that what Harris wants you to grant is a philosophical, not a scientific, premise; hence, his “moral landscape” is not scientifically determined as he claims.

Here’s the link (please forgive my animated gesticulation; I was frustrated, and trying to be polite).

Follow Brian on Twitter by clicking here.

 

OTHER GOOD CRITIQUES OF HARRIS’ ARGUMENT:

Simon Rippon has a critique here; Massimo Pigliucci has one hereRussell Blackford has a pretty good one here; and the best one I’ve seen is by Whitley Kaufman in Neuroethics here.

 

  • Facebook
  • Twitter
  • Reddit

71 Responses to Sam Harris is wrong about science and morality

  • Fel says:

    Brian Earp, you've done it again! Loved the post.

  • Adam says:

    Brian – good article indeed. It frustrates me so much that Sam Harris thinks he is doing serious work in moral philosophy, and that others are continually agreeing with him.

    On his hypocrisy and lies, 'the moment he grants' himself the ability to choose whatever values he wants to ground his theories, he becomes as bad as the theistic folks that he has been railing against for years. Hypocrite? Indeed. Liar? That seems to indicate some sense of self awareness about how bad his theoretic foundations are. And it seems to me that he has not yet grasped enough basic ethical theory to develop enough self awareness to be properly called a liar. You say that he studied philosophy at Stanford, but as the old adage goes, you can lead a horse to water but you can't make it think.

    Final note: I think my hands would be gesticulating even more if i had to listen to him spout his wisdom for an hour and then had the opportunity to ask him directly how he is getting away with such weak work.

    • Brian Earp says:

      Thanks, Adam. I'm looking at some of the endorsements of his book by Steven Pinker, Richard Dawkins, and others – and I'm thinking, really? Can these men be so easily duped? I guess when you have Harvard and Oxford professors as your friends, they'll write nice things about your book, even if it's impotent philosophically. I have to state again: some of the specific claims Harris makes are very interesting, and he is right to remind us about some bad things that are happening in the world. Some of the studies he relates are interesting. And so on. He's just embarrassingly unsuccessful at establishing the meta-ethical point he's got inscribed all over the book jacket – and, as I wrote, I think he knows it. He must. For someone who makes his career chastising others for their lack of integrity and use of poor logic, then, the hypocrisy here truly rankles. That said – he's a good writer. So there's that.

  • Matt Sharp says:

    Agreed with some of your criticisms of Harris, assuming you've accurately represented his position.

    That said, I still have sympathy with his claims, and ethical naturalism generally.

    Whilst premise (1) may be a philosophical claim, how about this:

    'If all other things are equal, then an increase in suffering is bad'.

    It seems to me that if the experience of suffering wasn't bad, then it wouldn't be suffering. However, since people may willingly put themselves through suffering (for example) in order to experience future gains, the 'if all other things are equal' antecedent is required.

    If we accept that, then the rest of Harris' argument follows. Suffering can (or may one day) be measurable via neural imaging. You may want to argue that common sense will do just fine, and in many cases it will. But if, say, you want to do cross-cultural and cross-temporal studies of well-being (as many countries are starting to do), then it is necessary to have an objective baseline to ensure the apparent meaning of the terms that make up well-being are not simply being interpreted differently across languages and through time. i.e you need to see whether, when a person in China reports experiencing pleasure of 6/10, it corresponds to when a person in the UK gives the same score.

    • Anthony Drinkwater says:

      Let us try to get to the logical limit of this argument :
      If we find through MRI or any other technique, that the Chinese " objectively" support suffering more than the English, will this justify differential treatment ?
      The implication of Harris' thought seems to be that if this were the case we should do less to ensure the well-being of the Chinese.
      And that this would be ethical …..
      ?? !!

      • Brian Earp says:

        I think this is on point, Anthony. Harris at no points spells out how fMRI scans (etc.) can give us information about well-being that's any more useful thn just asking people, using common sense, and so on. I had thought that would be the only refuge for his book to say something new or worthwhile, and he didn't go there. When pressed – as in the video clip I've linked to – he says things like, "Well, yes, there IS an objective answer in completed science – even if we can never reach it – about what's a better way to act in situation X; but we DO know certain ways are WORSE, for example, throwing acid on someone's face." … I bought his book (after seeing his talk) for the express purpose of trying to find an example of an interesting moral puzzle solvable by the methods of science. I found none.

      • Matt Sharp says:

        What MRI (or other brain scanning technique) could show is that the Chinese have a different conception of what suffering *is*. It could simply be that there is no word in Chinese that has exactly the same meaning as 'suffering' does in English. In fact, there is quite a bit of psychological evidence which suggests this might be the case; when East Asian people are told to 'think about themselves' there is often slightly different activity to when Westerners do the same thing (and there are also differences with many other similar tasks). It seems likely that this is because East Asian philosophy and culture has been built around a more interdependent concept of 'the self' (or even complete absence of the self as with some views in Buddhism), whereas Western philosophy has been very much individualistic, promoting the idea of personal identity.

        If this is the case, then we cannot compare happiness or suffering on *exactly* the same scale, and an action that promotes well-being in the West may not achieve quite the same degree of success in China. Though there will be a large degree of overlap (after all, we still have the same basic physiological needs), I do think this would justify differential treatment.

    • Brian Earp says:

      Hi Matt — sure, "all else equal, try to reduce suffering" — that's pretty basic, and most would agree with you. But it's nothing new. It hardly gets us past common sense. And that's the point I was trying to make about Harris' book. It's not that what he's saying, on utilitarian premises, is all that unreasonable. It's that he's claiming that utilitarianism — or some version of it — is "true according to science" … It's the philosophical pretense I meant to call attention to in this post. And I think it matters. Harris is going around now giving talks, saying that "according to science" X, Y, and Z is "objectively morally wrong" … This is dangerous in my view. It's scientistic authoritarianism. The examples he always comes up with are things that every Westerner would agree are bad things: burning people with acid, and so on. Yes, we agree that does not conduce to well being. Yes, nobody wants to have acid poured on her face. If we can't get past these cherry-picked examples, though, and say anything whatsoever nuanced about science's ability to resolve actual moral puzzles, then we (or: Sam Harris) should be a lot more modest about what we're contributing to the debate.

      • Matt Sharp says:

        Fair point, well put.

      • Kevin says:

        ZZZzzz, cry more Mr. Earp.

        You're arguing over the definition of science, but you don't realize it because your douchiness apparently requires most of your attention. The point Sam is making is that one cannot understand anything about the human condition through anything other than science. Even if you say, "well what about philosophy and common sense, those things are not science! RAGE!!!", Sam would say as a tool for objectively understanding human well-being, philosophy and so called "common sense" bring nothing to the table that science doesn't. Your question, " 'what can “science” tell us about morality that we didn’t know from common sense'" is meaningless because "common sense" is not a valid realm for understanding anything (In fact, it's woefully inadequate borderline made-up concept).

        There is a difference from answers in practice, and answers in principle. All Sam is suggesting is that science can in principle give us answers to questions of human well being that so called "common sense" can only speculate on. Certainly most things haven't be measured with the tools of science, and many things may seem impossible to measure, but as a measurement tool "common sense" is just as good as taking a guess.

        • admin says:

          This is a friendly reminder that we welcome debate and dissent, but our comments policy requires that you keep it civil. Thanks.

    • Simon Rippon says:

      "‘If all other things are equal, then an increase in suffering is bad’.
      It seems to me that if the experience of suffering wasn’t bad, then it wouldn’t be suffering."

      Matt, is that supposed to be claim about the definition of suffering? If so, then it will never be "measurable by neural imaging" – even if we assume you've managed to get over the hurdle of explaining how neuroimaging is supposed to measure *conscious* states, rather than mere correlates of them.

      Suppose I invent a word: "baddog", which I define to mean a dog that is bad. Then I can't hire even the most highly qualified vet or biologist to count the baddogs – unless he also uses his own moral judgment to determine which dogs are bad.

      • Simon Rippon says:

        Sorry, what I wrote was a bit ambiguous. I meant to say: then it follows that *suffering* will never be "measurable by neural imaging"

        • Matt Sharp says:

          I'm probably being stupid here, but I don't follow.

          Suffering, as I understand it, is necessarily an undesirable brain state. If it's not undesirable, then it is not suffering.

          If someone claims they are suffering, and they are being honest about it, then we can scan their brain to see which areas are active when they claim they are in that emotional state. Clearly you're right that this simply shows us correlates of suffering rather than the actual conscious states, but if the correlates are incredibly consistent I don't see the issue.

    • keddaw says:

      <blockquote>‘If all other things are equal, then an increase in suffering is bad’.</blockquote>

      Matt, unfortunately <i>ceteris paribus</i> cannot be applied here, in order for there to be an increase in suffering something else must change to cause it. If we take the situation before life began in the universe there was no suffering, are we worse off (as a universe) because life now exists which can, and does, experience suffering? Does the joy or, to use Harris' words, flourishing that those creatures experience outweigh it? Is there an objective way to measure? Of course not, it's all subjective. Even if we can use science to determine how our subjective views are related (do I actually prefer mint ice cream to raspberry ripple?) it cannot tell if I like mint ice cream more than I dislike stubbing my toe.

  • Simon Rippon says:

    Nice post Brian, I agree. In fact some similar points about Harris were made on this blog when his book came out:

    http://blog.practicalethics.ox.ac.uk/2010/10/sam-harris-the-naturalistic-fallacy-and-the-slipperiness-of-well-being/

    I suspect a clickable link won't work, but I'll give it a go:
    <a href="http://blog.practicalethics.ox.ac.uk/2010/10/sam-harris-the-naturalistic-fallacy-and-the-slipperiness-of-well-being/">click here</a>

    • Brian Earp says:

      Simon – thanks for sharing your earlier post. You go into much greater detail about the sort of philosophical work that would be required to make the argument Harris pretends to make; and we are in perfect agreement that he doesn't do this work: in fact, he doesn't attempt it. I'm not bothered so much by some of the specific claims Harris makes — such as: throwing acid in a girl's face for trying to learn how to read is not nice — but rather by his extraordinary disregard for basic reasoning, conceptual clarity, etc., coupled with his fantastical claims to be saying something new. Secular moral philosophy (badly done) is still secular moral philosophy: calling it "science" doesn't gain you an inch of objectivity, and I fear it makes progress on these important questions much harder, by muddying the waters. Again, I appreciate your link, and your carefully reasoned critique — I realize I'm joining the parade rather late in the day!

      • Simon Rippon says:

        Thanks. As you've shown, different people can make essentially the same critique in quite different ways. I think it's helpful to do that … and with respect to a publicity-seeker like Harris, it's especially helfpul to do it in ways accessible to non-philosopers. So well done!

        I suppose I only disagree with you when you say that Harris does "plain old secular moral reasoning" in his book – I think he actually does very little moral reasoning, because as you point out, he simply cherry picks easy moral examples and then claims, without any recognizable argument, that they are supposed to teach us a general lesson. He claims that value consists in conscious states, but doesn't even *mention* the extremely important, extremely famous, 30-year-old challenge to that view known as Nozick's Experience Machine, for goodness sake. (How's that for honesty?)

        Certainly you would find much more and much better secular moral reasoning in virtually any introductory ethics book. "Elements of Moral Philosophy by James Rachels, for example, would be a good place for interested readers to look.

        • Brian Earp says:

          Simon – would you be willing to spell out how it is that Nozick's experience machine challenges that view? I'm curious to learn!

          • Simon Rippon says:

            How does Nozick's experience machine challenge the view that value consists only in conscious states, such as pleasure?

            There's a wonderfully clear explanation in Jean Kazez's book, "The Weight of Things", the relevant pages of which you can read here:
            http://books.google.co.uk/books?id=gK_U8rcFMeAC&lpg=PA51&pg=PA51#v=onepage&q&f=false

          • Matt Sharp says:

            As a response to the Experience Machine, may I suggest taking a look at a recent piece of experimental philosophy by Filipe de Brigard:

            http://www.unc.edu/~brigard/Xmach.pdf

            He surveyed a number of his students and found that it wasn't clear that, if they found out their current life was in an experience machine (or The Matrix), they would prefer to abandon it and return to 'reality'. He suggests many people's reactions to Nozick's experience machine can be explained a psychological 'status quo' bias: people want to stick with what they have, regardless of whether it's real or not.

            It can also be argued that even *if* people claim to value reality more than just conscious states, all this shows is that people place value in having the conscious state of experiencing reality; it doesn't show that reality per se is valuable.

          • Brian Earp says:

            Simon — just finished reading the passages in the book you sent — beautifully written, and they very carefully capture the intuition you spoke of. Thank you again for passing this on.

        • Brian Earp says:

          Thank you Simon and Matt for those resources. Matt – I think this new work on the experience machine is crucial: the idea that the status quo bias may explain much of people's preferences in the original thought experiment seems potent.

          • Simon Rippon says:

            Matt and Brian – DeBrigard's work experimental work on the experience machine is pretty clearly fatally flawed, I think. He claims that status quo bias explains reluctance to unplug in his surveys, and could similarly be what dissuades us from plugging in. But there's an extremely important disanalogy between Nozick's original pluggin in scenario and the unplugging scenario he tests in his surveys. Leaving the experience machine would produce the experience of a *sudden and wrenching disconnect* with one's virtual relationships (Oh damn, I was wrong all along – my whole life has been for nought!!!), whereas entering Nozick's experience machine need have no similar effect. My actual desires include desires for the wellbeing of my loved ones and the fulfillment of my plans, for example, and I would look forward to these desires seeming satisfied if I entered the experience machine. In Nozick's scenario, my plugged in life would develop organically from my present one, only better. So there's no similar disconnect here, and for that reason there are no grounds for thinking that status quo bias would dissuade me from plugging in, while it very clearly could dissuade me from unplugging.
            In fact, I don't see why experimental methods actually lend anything at all to De Brigard's argument. There are no reports of any surprising findings about differences in intuitions between groups in it. Rather, we have a number of thought experiments *each yielding fairly uniform reflective intuitions*, and we want to figure out what the best explanation of these intuitions is. This work could be conducted from the armchair by lone philosophers (though of course with input from empirical psychology identifying general psychological sources of bias). Experimental methods like De Brigard's are only of interest when people's intuitions significantly differ.

            Matt, I'm not sure what you mean by the claim that people value "the conscious state of experiencing reality". If you mean that people value the conscious state of *seeming* to be in touch with reality, this can of course be provided by the experience machine itself. So this can't be the explanation of their choosing not to plug in. Maybe you mean that people value a mental state with externalist content – that is, a state that by definition they can only have if they are indeed experiencing reality (like "the knowledge that snow is white", which you can only have if snow is indeed white). Then you could indeed argue that their valuing *that*, rather than their valuing reality per se, is what explains their reluctance toplug in. Assuming (controversially!) that what you've identified is properly described as a "mental state", rather than as, something like a relation between mental states of the world, you could then argue that the experience machine does not challenge the thesis that only mental states are valued. But this seems to me a whole lot less plausible than the claim that people value both mental states of certain kinds, and realities of certain kinds, and these things together explain their reluctance to plug in to the machine. And, anyway, you can trivially show that people value reality being a certain way irrespective of their conscious states by using further, similar examples.

          • Matt Sharp says:

            Simon, thanks for the reply. Some very good points that I'll have to think about.

  • Trevor says:

    Hi Brian,

    Great post. I have to say that I had many of the same reactions while reading The Moral Landscape. I was hoping for some sort of moral revalation that proved, scientifically, beyond a reasonable doubt, that a certain world-view and sense of morality was objectively superior to another. What I was left reading was a paragon of circular reasoning and an exercise in frustration. "fMRIs can measure suffering". Okay…"When humans suffer, their brains register the suffering and we can measure it." Right. "When people have acid thrown in their faces, their brains register that as suffering." Well, sure. "Therefore, throwing acid in people's faces is bad." Well, OK.

    Again, as you say, it's less science than it is common sense. And that's really nothing new. To call a western version of common sense "objective" and back it up with speculative future-science that may not even ultimately contribute to the argument is a specious argument.

    I'm not sure I would call him a "liar" per se, but I would certainly call his book inaccurate and a little misleading. There were some good, hearty re-hashings of good old-fashioned utilitarian hedonism on a macro scale, a school of thought with which I happen to sympathize. But his whole argument rested on a pillar (or, as you said it, cornerstone) of subjectivity on which Harris scribbled the words "objective science".

  • David Brax says:

    I totally agree about Harris, and it's a shame that he is the person associated with this claim, since there are much better attempts out there. It doesn't take philosophical genius, though, just the same sort of caveats and hedgings that any argument in philosophy (even Hume's law) ultimately depends on. I made such an attempt as part of my dissertation "Hedonism as the Explanation of Value". It's basically reductionist Cornell-Realism with a bit of scientific (some neuro-) psychology to back it up.

    • Brian Earp says:

      David,
      Do you have your dissertation posted somewhere online? Is there a link? … Can you summarize the argument? I'm curious indeed.
      Best wishes,
      Brian

  • Jean Kazez says:

    Enjoyable post!

  • Scote says:

    Great post. While I like Sam Harris to a certain degree, many of his arguments don't seem to stand up to scrutiny, and his whole un-defined idea/muddle of "greatest good" just seems like an attempt at hiding the necessary philosophical value judgments under the rug so he can claim his morality is science based rather than value based.

    I do take issue with one thing you wrote, though:
    "I hope you’ll agree that we didn’t need science to tell us that abusing people is bad: common sense will do just fine. "

    No, common sense will not do fine. The "common sense" of the Taliban tells them it is fine to force women to wear burkas. Common sense is highly subjective.

    • Brian Earp says:

      Hi Scote – thanks for your comments. I'm not sure if you're responding to an earlier version of this post, but I tried to address this issue: I wrote: 'And if someone disagrees, say, the Taliban, intoning “but science says you’re mistaken” will do little to change their minds.' My point is that I trust that my readers — among whom I expect are relatively few Taliban members — will find enforced burqa-wearing to be commonsensically a bad thing. If someone DOESN'T have this position to start with, there's nothing "science" can do to address the moral disagreement.

      • Scote says:

        "If someone DOESN’T have this position to start with, there’s nothing "science" can do to address the moral disagreement."

        Ah, that I don't disagree with.

  • Rosemary Lyndall Wemm says:

    I am surprised that Harris does not properly evaluate the place of cognitive-emotive thinking on the perception of pain or the role that hormones and other neuro-active chemicals have on the perception of pain. Any psychologist or neuro-scientist working in the area of mental health should be aware that what is painful or pleasurable to someone at one point in their life may be perceived differently at another point in their lives. So which reading do we take as accurate?

    • Brian Earp says:

      Dear Rosemary,
      Thank you for your contribution. I'm as surprised as you are. The level of detail you bring up is really the only domain in which an argument like Harris' could begin to be productive – that is, given a certain moral worldview, how can science help us resolve genuine, specific moral puzzles that exist within that framework, where a careful analysis of pain/well-being at the level of brain states (and in conjunction with subjective reports) can give us a strong picture of what course of action to take. But his discussion just doesn't go there, wildly swinging from common sense to dubious meta-ethical assertions.
      Best,
      Brian

      • Rosemary Lyndall Wemm says:

        I think the difference between me and Harris is that I have trained and worked as a clinician in a number of different settings and a number of different specialities on clinical and neuro-psychology. I suspect that Harris has had a "sheltered" clinical life, or none at all.

  • While I'm partially sympathetic to the criticisms, and I also realize Harris does none of this sort of investigation either, but what if we determined (as I've guessed for a long time is the case) that humans are "naturally" consequentialists (not utilitarians, mind!) and that learning to be deontological (or something else) is a matter, in part, of somehow putting the natural predisposition aside. (Of course, the reverse would be interesting as well, for the same reason, but I suspect it isn't true.) Or more complicatedly, that there are psychological dispositions to one or the other which vary in strength. Yes, this couldn't tell us which viewpoint was "correct" in some strong moral realist way, but it could tell us how likely it was that a person (or humans generally) could adopt a given ethical viewpoint. If one adopts something like the "ought implies can" principle, then it seems to follow we'd have an argument against the particular "ethical theory" in question, at least to a limited degree. (Needless to say, some people will still say we should adopt it even if it is psychologically impossible for some of us to adopt it, but … )

    • Brian Earp says:

      Hi Keith — Thanks for your input. There's a lot of work in psychology on this, primarily by Joshua Greene and his colleagues, and they show (if you accept their interpretation of the data) kind of the opposite of your intuition: that is, across a range of moral dilemmas, people seem instinctively to choose the "deontological" answer (and this answer is associated with automatic/ emotional processes), whereas it is controlled, effortful, later-to-evolve brain processes that drive utilitarian thinking. (I'm not sure I agree that the data show this, but that's a different story). Either way, people are clearly able to choose options (or reasons in ways) which might be thought of as both deontological and utilitarian, and I'm not sure that a statistical shading of ease going one way or another gives us anything to work with to answer, "how should we live?" Clearly lots of morally respectable behaviors go against the grain of what feels 'natural' or what we may have an evolved predisposition to do …
      Best,
      Brian

    • G. Owen Schaefer says:

      Keith –

      Could you explain how, exactly, the evidence you're speaking of would show certain ethical theories are impossible to adopt? The fact that some theories are more prominent or, perhaps, easier to adopt than others would not show this, nor would evidence concerning whether we were naturally predisposed towards certain theories. The only evidence I could think of in this regard would be to show that *no one*, in fact, holds a particular ethical theory, but that certainly doesn't seem true for deontology or consequentialism (or really any theory that's been seriously proposed).

      Or perhaps you're thinking of the fact that *no one* acts perfectly in accord with certain moral theories, thus casting doubt on certain perfectionist accounts of morality? (ones that say perfect action is not only good, but morally required) Even then, though, I would wonder whether it is proper to infer from no one acting in a certain perfectionist manner to the impossibility of such action; prospectively, our ability to manipulate moral behavior (see Matt Baum's post on this subject above) could at least in theory bring about such perfectionism. Even if we never get the balance quite right and do not manage to engineer perfect moral behavior/disposition, it would at least remain possible and the prefectionist could justifiably point to such behavior as something within the realm of human ability.

  • Stephen Parkes says:

    Okay, I keep being told this is a duplicate comment, but haven't seen my comment appear, so will try one more time:

    I haven't read the book, but is it the case that Harris is very concerned with moral relativism? I mean, look at his examples: the Taliban, and the throwing of acid in a girl's face because she tries to learn to read. Within the Taliban's own value system, forcing women to wear burqas at the threat of being beaten, and stoning gays to death, are morally okay. So moral relativists and subjectivists generally struggle to criticise such behavior because they usually hold a position like "who are we (in our different society/culture) to impose our values on other groups of people".

    An answer to this would be if morality was related to something objective about the world, that would hold true for any group of people. I guess Harris is thinking "what could be more objective than science?" If Harris can show that objective science gives us moral instruction, then that settles that.

    However, there are already arguments (compelling ones, in my view) that have been made against extreme moral/cultural relativism that don't rely on accepting that morals come direct from science.

    • Brian Earp says:

      Dear Stephen —

      You're exactly right. This is Harris' approach: he points out that religions make absolutist moral claims, and secularists have nothing to respond with, because they're stuck with weak-tea relativism. Well then (Harris thinks), what could be more objective than science to ground moral judgments as against the religious folks? … Sure, that WOULD be a nice solution, if it were possible. But it's not. Or if it is, Harris doesn't show how. So while relativism might not be very satisfying — while it feels better to say to the Taliban, "You're WRONG, according to science" — we're left with the grayer position of criticizing them from within a philosophical framework whose cogency we have to show through argument and other forms of persuasive appeal.

      Best,
      Brian

  • Scott says:

    I would love to read a serious negative review of The Moral Landscape. Unfortunately this isn't it. It's full of philosophical cant I learned as a freshman and does nothing to address the arguments Harris uses to support his overarching argument you sketch above. Also, I don't know where people get the idea that Harris' believes himself to be doing original philosophical work, let alone that it is anything approaching revolutionary. For heaven's sake, he cites the work of philosophers who make the same of similar arguments!

    Can't we do better?

  • @blamer says:

    Brian's take-down misrepresents TML.

    Ethicists can legitimise their speculations (of what 'ought') by aligning their claims to what the sciences have discovered about brains, psychology, social groups, etc (the 'is').

    Hence science informs our discussions about human values, at least in philosophy departments.

    This is profound news outside of academia! We see "common sense" leading to vastly less harmonious ways of thinking about emotions, culture, and morality.

    • Brian Earp says:

      Hello,

      "Ethicists can legitimise their speculations (of what ‘ought’) by aligning their claims to what the sciences have discovered about brains, psychology, social groups, etc (the ‘is’)."

      … How can they do this? This is exactly what's under discussion. 1) How can they "legitimize" their speculations; 2) what does it mean to "align their claims" to facts of science?

      Thanks,
      Brian

      • @blamer says:

        I confess your 14-Oct '10 post (the scientistic argument) hasn't helped me to understand why the sciences cannot expand to investigate ethics. I have in mind the social sciences like political science, not simply hard science drawing strong conclusions.

        As to your question, do we not increase the legitimacy of arguments about animal treatment by invoking contemporary scientific claims about pain and stress?

        TML is arguing that – far from being ethically irrelevant – scientific discoveries across the disciplines are already confluent with learning how we 'ought' to behave in the world.

        • Simon Rippon says:

          No, @blamer, The Moral Landscape claims to argue for much more than what you suggest. Nobody even borderline sane in the history of the world has ever denied that scientific discoveries are *relevant* to morality, and can *help us* decide what we ought to do. That's just blatantly obvious!

          Want to know if it's morally OK to prescribe thalidomide to pregnant women? Science can help! Want to know if there's a moral imperative to reduce our Co2 output? Science can help! Want to know if it's OK to point that thing at someone and pull the trigger? Science can help!

          But here's Hume's old insight: You're *also* going to need some kind of moral premise to reach the answers to these questions (for example: you morally ought not to cause future children to be disabled, or we morally ought not perform actions that will lead to famine and devastation). And nobody has ever given a plausible scientific argument for such a premise. That's the is-ought gap.

          Harris, on the other hand, claims the is-ought gap is just so much bunk. Harris, in TML (subtitled: "How Science can Determine Human Values"): "My claim is that there are right and wrong answers to moral questions, just as there are right and wrong answers to questions of physics, and such answers may one day fall within reach of the maturing sciences of the mind." (p.28).

          If you cannot understand the argument of the other Practical Ethics post linked above in the comments, "Sam Harris, the Naturalistic Fallacy and the Slipperiness of Wellbeing," I'd encourage you to read also the discussion that follows it. And then if you still have (specific) questions, I'm happy to try to answer them here.

          • @blamer says:

            Thanks for your reply, it has helped clarify. As too your exchange with Greg under that 14-Oct ’10 post (link above).

            The nub seems to be Hume's is-ought. Without dismantling that, we cannot have a branch of science that focuses on ethics.

            To be cheeky I'll sign-off with Dennett: "<i> If 'ought' cannot be derived form 'is', just what </i>can<i> it be derived from?" (p.196)

          • Simon Rippon says:

            Thanks @blamer. In answer to Dennett's rhetorical question: "Ought" can be derived from "ought", of course! Nobody, by the way, ever asked what we can derive "is" from, if we can't derive it from "ought"!

  • Scott says:

    Thanks Brian. I'll look at it. By the way, I'm glad Simon made an appearance. He provided an example of what I think is the real issue here by describing Harris as a "publicity seeker." Harris' ability to successfully seek publicity is what really raises people's hackles. On the one hand you have Simon implicitly acknowledging that Harris' goal is to engage a wide audience, but behaving grouchily when he does it well — i.e. not writing as if he were planning to publish in an academic journal. @blamer has it exactly right, Sam isn't speaking to academia!

    • Simon Rippon says:

      Yes, that's right Scott, it's not truth, honesty, or quality I care about: I simply dislike anyone addressing the non-academic public about ethics. Glad you came to Practical Ethics News to share your profound insight.

      • Scott says:

        Get over yourself Simon. You delivered the opening ad hominem. Those who really are concerned with the quality of Harris' arguments ought to address what he actually says, and with the proper charity; I haven't seen much of that myself. I don't know another academic field so anxious about communicating with the public than is philosophy. Imagine what another physicist or cosmologist could do to Brian Greene's books if they chose to ignore his intended audience . . .

        • Simon Rippon says:

          Right, Sam Harris is exactly like Brian Greene – if Brian Greene (an eminent physicist and cosmologist) wrote books about the biology of evolution, and said in these books that he had chosen to ignore what academic biologists said about evolution (as well as their criticism of his own theories about evolution) because he had reached his conclusions independently of the biology literature, and because he thought the views of academic biologists and the conceptual distinctions they made were too complicated for the general public, and their terminology was boring. (See "The Moral Landscape", fn. 1)

        • Simon Rippon says:

          And Scott, as you say: "Those who really are concerned with the quality of Harris’ arguments ought to address what he actually says, and with the proper charity" – that's your own principle, so why don't you apply it to your remarks on *our* writing?

  • Matthew Bryant says:

    So when Harris says he thinks the distinction between science and philosophy is a false dichotomy, your response is to sarcastically restate the false dichotomy without saying why you think he's wrong…

    Why would you have a problem with data backing up a philosophically derived statement about ethics? I don't understand why you disagree with Harris.

    • Brian Earp says:

      Harris is right that there's no formal dividing line between philosophy and science: but he's using a play on words to make it sound like he's doing something radical. "Science" usually means the empirical method, i.e., that branch of investigation that establishes facts, and devises theories to explain those facts in terms of deeper patterns. And science on THAT conception is impotent to dictate values, moral or otherwise. What's so exciting-sounding about Harris' talk is that it purports to show that science in THAT sense actually CAN do something that people have been saying since Hume it cannot. But then when you press him about what exactly he means by science, it turns out that what he's talking about is something that's NOT strictly separable from reasoning or philosophy, thereby deflating the source of intrigue in the title of his talk. Because no one has ever argued that "science" (meaning fact-sensitive reasoning) has "nothing to say about morality." Anyone who says "science has nothing to say about morality" is using the word science to mean the first definition I gave at the beginning of this reply.

    • Brian Earp says:

      Also, Matthew — I have *no* problem with using data to back up a philosophically-derived value or premise. I must state this clearly: you have completely mis-read my post, and you have missed exactly my point. What Harris' book ACTUALLY does is what we're talking about right now: it starts with a philosophical premise, and then it shows how science can feed facts into the reasoning process that flows from that premise. The point I'm making is that this procedure — using some combination of facts and philosophy to engage in moral reasoning — is a very old tradition going back hundreds of years. The problem is that Harris does not admit what he's doing. He CLAIMS that science can "determine human values" – when in fact all he shows is that, GIVEN certain values, science can offer relevant facts to sort out what to do. Those are drastically different things. Harris (or his publisher) has adorned his book jacket with claims of having bridged the "is/ought" divide, and it's this false claim for which I mean to take Harris to task. Doing philosophy while minding facts is not the problem.

  • Nick says:

    The main group Harris is attacking is those that claim the ought should come from their personal holy book.

    I think what he's saying is that eventually hard science will have better, more precise answers for all questions of moral philosophy.

    Very similar to what Einstein did to Newtonian physics, I don't think Harris is claiming to be the Einstein of morals, but that morality and psychology are more closer to the Newton stage, and to not think there never will be an Einstein of psychology is wrong.

    • Nick says:

      The last part of the last sentence should read, "and to think there never will be an Einstein of psychology is wrong."

  • Allie says:

    I think the unfortunate thing here is that while philosophers are happy to spend 1,000 years disagreeing on the technical meaning of a few words, religious fundamentalists rush to fill in the gap while academics bicker among themselves.

    Economics? Health? Psychology? All centered around nebulously defined "oughts", as are many of the soft sciences. There is a vague sense at the center of those disciplines that we are shooting for financial, physical, or psychological well-being as the end goal. Harris, as I understand him, simply says we should do the same thing when it comes to creating a soft science we can call "Morality".

    I can see how some disagreement on technical definitions and semantics might arise, but again, I think that misses the bigger picture by a long shot. Surely we can do a much better job of talking about standards for human well-being while we hammer out the details of what title to use.

    • keddaw says:

      Economics is not even remotely concerned with "oughts". Economics determines (or tries to!) the relationships between various entities. So, for example, when a business person wants to use economics to determine their future strategy it is a pointless exercise however, if they wish to have a particular goal then economics is the ideal tool to help determine the best course of action.

      This is an almost perfect analogy to what Harris is talking about and why he's wrong. While Harris may say it's all about well being, the non-economist would say it's all about money, but is it? Would a business owner not also value security, market share, long term profits over short term? Are these not entirely valid goals, and, more importantly, would there not exist different business people who would value them in a different order and so require different strategies? This is what Harris ignores, he seems to think that any two strategies can be measured on a single scale, completely ignoring the multitude of factors that go into making up his magical well-being and also the fact that different people have different weights on each metric – and different metrics! Should we be maximising total, average or median well-being? Should we eliminate suffering before worrying about increasing well-being? Is there a trade off to be made between suffering and well-being, even between people? These are entirely valid moral questions and Harris' system ignores them to such a degree that it is painful to read his book and try to take it seriously.

      Harris blithely states, without evidence, that the Taliban's treatment of women is obviously less good than ours. But surely we need to measure well-being before doing that. If they Taliban believe that heaven and hell are real then they may well be acting entirely rationally and believe they are increasing well-being of the women by saving their souls. Their peak of well-being would be much higher than ours, we simply cannot measure the good parts of their world-view. Which is not to promote naive moral relativism, it is simply to point out that under Harris' own system that is a real possibility.

  • Rosemary Lyndall Wemm says:

    You make excellent points.

    I have not yet read Harris's book so I have no way to assess whether you have got the correct gist of what he is saying. If you have, then Harris has indeed attributed too much to science.

    As a fellow neuro-scientist, I am happy to use information I gain from the study of my material pertinent to my profession and any other information that impinges on how human beings work and how things affect their health and well-being. I am happy to use that information to improve the fairness of my moral judgements. I am happy to concede that the increasingly sophisticated socialization of human beings is in large part due to genetic influences resulting from evolutionary processes. I am not willing to concede that there are not other influences at work or that morality is entirely, universally and uniformly objective, even given total knowledge of all the scientific discoveries made so far. While I may be genetically, socially, developmentally and environmentally programmed to start with the first premise you ascribe to Harris's schema I still consider it to be a value judgement. If it were not, then every society would be identical in what it prescribed as "moral".

  • Malcolm Forbes says:

    Are you just pissed off because Sam Harris thinks philosophers are a bunch of wankers?

  • William Theseus says:

    Man has an animal nature and a human nature. The highly evolved human nature in man is sublect to will. The default action is always the animal nature. Morality is the struggle between these two natures in Man.

    See Will in The Humanist Bible.

Authors

Affiliations