Sunday, September 8, 2013

Are moral philosophers scientific pessimists?

Scientism is back in the public discussion, and it's back in full swing. Although the debate seems never really to have settled down in recent years, a couple of noteworthy events recently have thrust it into the spotlight. First, Steven Pinker published an essay in The New Republic urging, among other things, that "scientism" is largely a canard. Second, being the scapegoat of much of the preceding and ensuing debate, Sam Harris issued a challenge to his critics, inviting them to topple the central thesis of his 2010 book The Moral Landscape: How Science Can Determine Human Values. As Harris's doppleganger, whom I call "Straw Harris," is routinely thrashed over the claims of Sam's book, it is my opinion that his challenge is entirely warranted, and yet a cadre of ruminants continues to arrive to nibble at Straw while Sam sits watching with what must be ever-decreasing patience.

One such interested writer is Op-Ed columnist Ross Douthat, of the New York Times. Douthat addressed Pinker's essay, and now he's gone on to address Harris's challenge--as a warm-up act for his shot at a winning essay? I wonder and have my doubts. "Sam Harris and Scientism," Douthat wrote, straight to his point, in his Opinions page space on September 5, a piece I would have all but ignored if it weren't for noted philosopher of science, etc., Massimo Pigliucci calling it a "good argument" against Harris.

I asked Pigliucci about this good-argument claim on Twitter and was assured that it is one, "if I look at it" (I had twice by that point). Pigliucci, of course, accuses Harris of being one of scientism's worst offenders and has taken his own shots at Pinker's essay. Pigliucci asserts, primarily, that Douthat exposes the "disanalogy between well-being and health"--an odd point to make since unseating this analogy, which appeals to gaining intuition into the matter, if refuted would not unseat Harris's central premise.

To avoid making this over-long, I want to focus on Douthat's piece, even if he technically isn't a moral philosopher. Hopefully, I'll be able to provide those looking to topple Harris with some ammunition on how to do it correctly, if it can be done.

Douthat almost immediately gets to quoting Harris's summary of the rebuttals to his claim that science can determine human values, and much of the rest of Douthat's column is devoted to arguing why those criticisms hold up against Harris's responses. That summary Harris provides is:
… there are three, distinct challenges put forward thus far:

1) There is no scientific basis to say that we should value well-being, our own or anyone else’s. (The Value Problem)

2) Hence, if someone does not care about well-being, or cares only about his own and not about the well-being of others, there is no way to argue that he is wrong from the point of view of science. (The Persuasion Problem)

3) Even if we did agree to grant “well-being” primacy in any discussion of morality, it is difficult or impossible to define it with rigor. It is, therefore, impossible to measure well-being scientifically. Thus, there can be no science of morality. (The Measurement Problem)
Harris goes on to claim that by swapping the term "medicine" for "morality" and "well-being" out for "health," these criticisms are obviated. Douthat summarizes Harris's position (quoting him at times) in this way.
The concept of well-being is to morality as the concept of health is to medicine, and while people who deny that the pursuit of conscious creatures’ well-being is the proper end of any moral system may be entitled to their opinion, they are in the same position with respect to moral dialogue as “Christian Scientists, homeopaths, voodoo priests, and the legions of the confused” are with respect to medical conversations. They have “perverse and even self-destructive ideas about how to live,” and their position is rightly excluded from rational debate. [Inside quotation marks in this blockquote are Harris's words, the rest being Douthat's.]
Douthat's first agenda from here is, rather than to counter Harris directly, to attempt to undermine this analogy so that he can make the case that Harris has not adequately responded to these rebuttals that "seem like compelling points," in Douthat's own words. If Douthat is right, Harris may have a little more work to do in obviating these arguments, but even if Douthat is right here, he hasn't defeated Harris's arguments but rather a particular defense of them. And yet Douthat does not appear to be correct here.

Immediately, for instance, Douthat bogs us down by seizing upon Harris's “Christian Scientists, homeopaths, voodoo priests, and the legions of the confused,” noting that these various people are concerned with health but are equipped with methods that are identifiably poor at achieving success with it, stating that "no matter how blurry the proper definition of 'health' may be, the cemeteries are filled with clear examples what it doesn’t look like" [emphasis his]. He names these people as some "who do not actually exemplify problems (1), (2), and (3) in Harris's medicine analogy."

There are two prongs of response here: (A) the best Douthat can hope to do by this point is to damage the analogy Harris used to respond to the challenges, and (B) we aren't that much worse off at identifying miserable people (and animals) than we are at identifying dead people, and over time we will continue to get better at making this identification, and so also we have an empirically grounded sense of what well-being doesn't look like.

We'll return to (B) in a minute. Here's why, to (A): Harris's use of “Christian Scientists, homeopaths, voodoo priests, and the legions of the confused” provides examples of people with broken approaches to medicine. Douthat notes, by contrast, that they aren't people that set out directly to be injurious to the goal of health, and that's true--on a presumption of their ignorance. For the voodoo priests or witch doctors, there is a good claim to be made here, but for Christian Scientists, such is not so clear, at least not in countries like the United States. Those people claim to value health but, knowing what they must know, or at least see, about modern medicine, value it less than they value their beliefs about health.

That aside, it still isn't the case Douthat needs to make unless damaging this analogy is all he wants to do. If Harris's analogy is flawed, so be it, but that doesn't unseat his case. Douthat dismisses Harris's intuition-pumping analogy but doesn't drive to the core of the problem: he needs to show that people who actually do exemplify (1), (2), or (3), that is people who actually do not value well-being, think of their well-being in only the most narrow terms, and people who assert that well-being cannot be measured, even in principle, have good justification for their moral attitudes that is beyond empirical methods to correct.

There are people who do not value well-being for themselves and/or others (masochists and sadists), and we do not include their attitudes in rational moral discussions. Indeed, we consider their attitudes a pathology. There are people who put themselves and their concerns ahead of those of others (sociopaths), and again, we hardly hesitate to consider their attitudes in (most) rational moral discussions. In both of these cases, empirical data unequivocally could illustrate that a worse outcome, in terms of well-being and suffering, is obtained by making space for their attitudes than by not doing so.

There are also people who seem to think that well-being cannot be measured and thus, even if it is a bedrock moral value, it cannot make for a scientific basis for morality (a few scientism-calling moral philosophers, among them). For the moment, we do include their insights in our rational discussions. Perhaps that will always be the case, though we might hope that, if it proved warranted, they would release the grip on so much primacy in this academic domain, traditionally theirs or not.

The point (B) above, though, cuts across this third group cleanly. Harris's case isn't that we can have such a scientific morality today, or even that we can ever develop one in total. It is that in principle such measurements of suffering (or even flourishing, happiness, etc., as components opposed to suffering--ones recognized from the days of Greek philosophy as fundamental components of a "good life") could provide profoundly clear answers to our moral questions. Particularly, we should be able to get much better insight into what we should value by considering the realities of how we suffer and how we flourish, using scientific methods and discoveries as a finely honed tool for doing so.

Our understanding of psychology, sociology, and (though it's a boogeyman to the scientismists due to some hyper-excited overreach in short-term claims about the field) neuroscience are already shedding keen insights into the nature of well-being and suffering, and it's hard for me to imagine that it's anything but profound pessimism with regard to our capabilities to advance these sciences that motivates claims that we'll never be able to make meaningful measurements of well-being. It's worth noting that Harris doesn't even claim that there is a unique ethic or set of moral attitudes that produces optimal results--just that we can use scientific methods to ground our various moral systems in something more salient than pure thought. (It's always worth noting for me at this point that the old philosophers held reason so far above the nitty-gritties of reality that they were able to conclude incorrectly that men and women have different numbers of teeth and, far more staggeringly, that opening mouths and counting them would be a deficient way to evaluate that claim.)

The Measurement Problem, then, appears to be little more than a call to complexity, if it isn't a ditch-effort punt, in order to maintain the primacy of non-empirical methods of working in this field. The Persuasion Problem is hardly a problem on even very muted consequentialism as we can easily see that people who have a very narrow band of concern possess more potential to injure well-being outside of their tribe and thus, statistically speaking, will do so on average. It isn't a stretch of reason, and would fall afoul of data gathered on well-being and suffering, to realize that this problem is extremely likely to produce a worse result than otherwise.

One additional point about the Measurement Problem--to address another common, but weak, objection that seeks to defend it--is that even if the set of possible moral states is only partially ordered, particularly that there are literally situations that are incomparable (e.g. how many prevented rapes is a murder worth? or some such), that does not preclude the ability to measure, nor does it preclude the ability to identify those higher or lower in value when that is appropriate. Harris, of course, makes this point via his landscape metaphor, as noted previously and puts no requirement for total comparability of all potential moral positions.

Douthat may sense the issues with the Persuasion and Measurement Problems, though, and so he turns his bigger guns upon the Value Problem. That's fair enough, since no matter what we want to do regarding morality, we do at some point have to smuggle in at least one value. Harris's essential claim is that well-being is a bedrock moral value, though, one that holds a deserved special status among values and even potential for empirical evaluation over other choices of values. Douthat notes, attempting to undermine this thought:
If I say to Harris, the good is the beautiful and the beautiful the good, and therefore the best and most morally admirable society is the one that produces the most beautiful artifacts, the loveliest lyric poetry, and the most scintillating prose, and this remains the case no matter how low the inhabitants of that society score relative to our own on subjective measures of personal well-being, we are in disagreement on a much more fundamental point than the Christian Scientist and the surgeon. The latter two are aiming at the same goal with different methods, whereas I deny that Harris’s goal is necessarily the one we should be pursuing in the first place. 
But Douthat misses, in an important way, that the Value Problem is very likely to be caught up, at least in part, in the Measurement Problem, which, again, I don't think has good legs to stand on, particularly in light of Harris's claim to being able in principle to resolve it. If we actually can get deeply into an understanding of what leads to suffering and well-being, then we can evaluate Douthat's hypothetical claim that "the good is the beautiful and the beautiful is the good" by analyzing how "the beautiful" impacts his thinking and conscious experience in a way that leads to him wanting to define it as "the good." Incidentally, in a hypothetical society that maximizes well-being--which is not the same as a society without suffering--an individual Douthat could hold this particular view, or many other views, even if he would suffer to some degree by having to live in it.

Of further note, here, Douthat gives away the ghost a little by bringing up this classic "good if and only if beautiful" idea. Particularly, he didn't boldly venture into attempting to claim that "the good is the hideous, the ghastly, the ugly; and the hideous, the ghastly, the ugly is the good." Using the Orcs of Mordor as a hypothetical culture, he could imagine a tremendous "good" coming of cacophony and chaos and corruption, of pain and suffering and discord, and he could defend it upon the notion of placing value on those ideals. Of course, he didn't! Why? Because it would fly in the face of all of our rational moral intuitions, which are based upon the very notion that "good" means something that isn't repugnant, something fundamentally connected to the idea of avoiding misery and promoting well-being. We may smuggle in that value, at bottom, but we do so because on the broadest consensus we might imagine, that's simply what "good" really means.

Douthat continues that paragraph with a rather bald assertion resting on the Measurement Problem.
And whereas the voodoo priest’s worldview is ultimately refuted scientifically by the corpses of his patient, there is no comparable scientific refutation that Harris can offer to someone who sides with Orson Welles’s Harry Lime (Link to relevant video clip, provided by Douthat).
Indeed, given the rate of improvement we are seeing in our insights into the nature of human (and animal) experience, and given that even without keen scientific measurements to tell us what misery looks like, this particular claim of Douthat's lands flat and should be gathering dust somewhere in the Appeal to Complexity forest that protects so many bad ideas borne in fuzzy thinking.

Of course, Douthat doesn't agree with Lime:
Not that the Lime worldview cannot be refuted! But the methods of the microscope and laboratory do not suffice to do so — whereas, again, they do generally suffice, given even trial and error and a large enough sample size, to refute the arguments of witch doctors and pseudoscience-spouting gurus. In the latter case, there are sick people and dead people to back up one side’s assertion about what counts as medicine; in the former case, there is only the bare presupposition, hanging unsupported by anything dispositive.
See again, though, the call to the Measurement Problem to try to back the substance of the Value Problem, while ignoring the plain fact that we have some pretty keen insights into the makings of misery. (Indeed, the existence of torture proves that human beings have almost a perverse affinity for understanding suffering.) So here, in the former case, instead of a bare presupposition, there are miserable people and people suboptimally well facing unnecessary and ultimately purely deleterious and, hypothetically, avoidable suffering. Pardon them, I suppose, for not being so obvious in their misery as is being dead. Under certain circumstances, these people could sue for damages to their qualities of life (e.g. negligence, which could be construed as valuing the interests of the self more highly than potential injuries to others), but I only mention this since Douthat decided to use the legal term "dispositive" to try to make this case.

Douthat comes around at the end by digging into a claim that morality need not actually consider well-being or suffering. He writes
But there is also no necessary reason why one must presuppose [the value of avoiding the worst possible misery for everyone] in order to pursue moral inquiry. ... [T]he aim of “avoiding the worst possible misery for everyone” is not definitionally integral to morality; it is one possible definition of the good that morality pursues. And, once again, the scientific method cannot settle the question of whether that definition is correct.
Morality, to be clear, is "the differentiation of intentions, decisions, and actions between those that are 'good' (or right) and those that are 'bad' (or wrong)," (definition by Wikipedia). So, then, Douthat's case rests upon the idea that there are meanings for the word "good" and "bad" that cannot be resolved to have something integral to do with well-being and suffering. Alternatively, we could say that Harris's case rests upon the notion that a bedrock definition of "bad," and thus "good" by negation, is the actualization of or tendency toward "the worst possible misery for everyone."

Perhaps this shows my bias, but if there is a more salient definition for "good" than "that which moves us away from the state of the worst possible misery for everyone," I don't know what it is and can't think of it after a great deal of trying. Douthat's "beauty is the good" boils down to an assessment, however subjective, of beauty, and it is imminently reasonable that one may presume that the reason it is taken to be either beautiful or good is because it provides some subjective sense of reduction of misery or increase in wellness, if only fleeting, if only psychological, if entirely individual.

Of course, we must note that certain kinds of transient or willful suffering, as with running a marathon for most people or (though now becoming obsolete) intentionally contracting chicken pox as a child to confer immunity as an adult, may produce a net benefit in well-being, so a simple understanding of well-being and suffering that takes into account only a moment-by-moment assessment of subjective quality of life is not sufficient.

Of note--and not a problem for Harris's case--it is likely to be a fact that our evaluations of our own well-being have a great deal to do with the so-called Persuasion Problem. I may very well value my own well-being, and that of my close family and friends, more highly than that of distant others, and my well-being may depend upon this. That's perfectly fine, though. If the goal is to maximize well-being while minimizing suffering, and if being perfectly egalitarian in valuation diminishes well-being for individuals, then it is likely to diminish it globally as well unless it necessarily leads to some unforeseen positive sum that, in light of our evolution into relatively small tribal groups, seems rather far-fetched. (Except in the most general terms, I cannot conceive of the immediate or long-term needs, goals, or passions of seven billion people--or maybe even more than a hundred--for instance, and I don't suspect I'm deficient in humanity because of this.) If that's how it is, then that's how it is, and empirically understanding the realities of human social psychology can only help clear the mud from this still-murky water.

Consequentialism and utilitarianism, of course, are part of what is under scrutiny here, each position possessing its own philosophical problems, but when it comes to moral evaluations, even on apparently incomparable situations, an evaluation of consequences and utility ultimately play a significant, if not major, role in guiding our moral intuitions. If I imagine receiving information that requires me to choose between loyalty to a friend and honesty to another, it is a careful weighing of the consequences, involving the interpersonal relationships, feelings of my friends, and my own state of being able to live comfortably with my choice that informs much of my decision process.

I might wonder what it is that a moral philosopher is using to determine moral attitudes if not some evaluation of the imagined, potential consequences of the hypothetical (and real) dilemmas in each case. Ultimately also, I might wonder upon what those evaluations are based if not experience, personal or shared, which are themselves a form of loosely empirical information that could conceivably be sharpened by clearer scientific insights into the nature of well-being and suffering. I suppose a virtue ethicist might suggest adherence to particular norms or ethics, for instance full-disclosure, always being truthful, keeping the peace, etc., but even the ancients were clear that these choices have to be judged situationally--and how, if not with at least some consideration of consequences and/or utility? All remaining answers to this question strike me as lacking real salience.

Harris's case may rest upon giving special status to the value of avoiding the worst possible misery for everyone, but if anything can be taken as a moral axiom, it is that this particular value constitutes substantive moral bedrock. On this thinking, then, the scientific method, equipped with a (currently mostly hypothetical) tool by which we can assess well being or misery with some degree of grounded meaning, absolutely can help us settle the question of whether or not Harris's definition of "bad," and thus "good," is correct. Different definitions of "good" within moral, cultural, or individual ethical frameworks, can be evaluated in terms of their impact upon this bedrock moral value, and as Douthat notes, "if you know what moral ends you’re driving at, then clearly science can be of assistance in your quest."

Thus, anyone, be it Douthat himself, Professor Pigliucci, or another, that seeks to win Harris's challenge has it ahead of him or her to demonstrate plainly how well-being, as Harris defined it, fails to constitute a bedrock moral value, with "bedrock" meaning backed by the empirically recognizable states of misery and happiness in sentient minds. That I cannot fathom a way to do this is specifically why I do not endeavor to enter Harris's contest, and so I'd be fascinated to see a serious attempt at this. Scientific pessimism and vague appeals to the complexity of the problem, though, just aren't going to cut it.

Author's note, 9/9/2013: Minor changes were made to this essay from its original published form to correct some typographical errors and to clarify meaning in a few areas.


  1. I wonder whether Harris's theory runs into trouble if it's seen in an evolutionary context. Should we think of an individual's well-being in evolutionary terms? For example, does sexual jealousy contribute to a person's well-being? Sexual jealousy may well serve an evolutionary purpose. It makes people very alert to the possibility of infidelity in a partner and motivates them to try to prevent it. This makes evolutionary sense. However, the experience of sexual jealousy is a very unpleasant one. Would people be better off if they didn't have the experience, and weren't troubled by their partner's unfaithfulness? Presumably, such people would be at an evolutionary disavantage, and would be less likely to pass on their genes to future generations.

    Competition for status provides another example. People of low status tend to be unhappy. From an evolutionary perspective that's a good thing because, again, it provides people with the motivation to change their circumstances.

    Reducing inequality is usually seen as an obviously good thing. If inequality means that some people have plenty to eat and others are starving then no one would qestion the need to tackle it. But what if the poorest people in society have a good standard of living and yet they are suffering simply because they are aware that other people are more successful than them? And what if this suffering is due to their evolutionary programming?

    Our emotions are tools that evolution has invented to make us behave in such a way that we are more likely to pass on our genes. We experience optimal well-being when we live the life that achieves that goal. We are happy when we are more successful than others and unhappy when we are less successful. The search for happiness is part of the game that evolution plays with us.

    1. Of course, I expect you realize this, but this isn't what Harris's contention is about. His contention is that science can inform us on what our values should be based upon empirical insights into well-being and suffering, if I've understood him correctly.

      As to your question--I don't think so, not necessarily. I don't think Harris's contention leads to any necessary challenge of what we might call the human condition. If we evolved to be narrow to a degree, and our overall sense of well-being and satisfaction in life depend upon that, then I would suggest that his claim is that science will help us uncover that and tell us a great deal more about how to do it well (at the very least better than pre-scientific guesses, even when those are good philosophy but especially when codified into religions).

      To your examples, if sexual jealousy is a fundamental part of being human (for some, most, or all people), then we don't have any reason to abandon that fact in a quest for moral values that optimize well-being and minimize (not eliminate) suffering, particularly gratuitous suffering. Again, Harris's case seems to me only to suggest that science is a preferred tool for delving into the question, not something that dictates what the necessary values must be (or that they be universal).

      The same applies, mutatis mutandis, to your other examples. It's kind of a dystopia effect, which I think was alluded to nicely in the film The Matrix. In the film, the computer character Agent Smith indicates that the original Matrix was a utopia, and that human nervous systems rejected it. If that's the case, then it's the case. Again, Harris's argument is that science is the tool that can inform us on that better than others.

      Psychologists (Haidt, at the least, e.g.) already recognize that we derive meaning and value to our lives--which are integral to happiness, which is integral to well-being--via struggle. Indeed, in cases where there is insufficient struggle, there are psychological problems. Here's an example of science determining that the value of eliminating struggle would be erroneous. Struggle, of course, carries with it some kind of suffering, and that's good for us apparently--but again, by studying it empirically we find it out. Yet again also, we see that eliminating apparent suffering is not necessarily the optimal solution, although I think an excellent case can be made that a strong moral value attached to Harris's case is the elimination of gratuitous suffering.

      (To refer to Harris's health analogy, we might consider the nervous system possibly needing certain amounts of interaction with various pathogens to become more effective later in life and possibly to preclude certain autoimmune disorders, if those hypotheses are correct.)

    2. A science of morality couldn't work on our emotions, but rather on what we should do with them. To talk about jealousy, is not different from talking about hunger or being horny, which are subjective experiences of conscious creatures like us. Those things are part of our nature, therefore not subject to moral examination. Nevertheless a science of morality could indeed inform us on what to do when you feel those things. For example, with jealousy it would be more conducent to well-being a dialogue than a murder. When it comes to hunger it would be more conducent to well-being to eat a fruit than a human being, just for sake of setting some silly example.

    3. Sexual jealousy had an evolutionary advantage but what science will tell us is that for maximizing well-being, we are much better off not paying attention to these negative hard-wired emotions.
      Passing on genes is no more the goal of life as it used to be. Unconditionally loving people is something that makes no evolutionary sense but on average makes for mentally healthier individuals and society. Science can come up with such finding(buddhism already teaches this) in principle.

      Science in principle can tell us to be satisfied with our circumstances and not compare. It can probably give techniques for kids so that they learn knowledge for the sake of learning, not for competing in exams with other kids. Extrapolate this to adult form and people wont be dissatisfied even if they are less successful because atleast they are doing what they enjoy and not comparing with others constantly. Those who continue to suffer by comparing oneself with others might get motivated to do better but that can again be another peak in multiple peaks of this moral landscape.
      Noticing such human behaviors(which counter evolutionarily hard-wired behaviors) and finding their usefulness (by collecting data and scanning brains) can come from science only.

    4. Thanks for the interesting replies. There is no doubt that science can illuminate questions of morality. I find it interesting that evolution has, as it were, "invented" the capacity for well-being. Well-being, like any evolved capacity, is a means to an end. Happiness is the reward for doing the things that evolution wants us to do. The idea that we should respect the happiness of others is also likely to be an invention of evolution.

      The interesting thing is that once evolution has invented these things they seem to take on an independent life of their own. Promoting well-being becomes a goal that thwarts the original purpose for which the capacity evolved.

      Once we have the capacity for well-being we can decide to have more of it than evolution intended.

  2. The central argument of his book in my opinion is not how it would look if we would to try to determine moral values just that there is an objectiv morality. Every action will either move the total wellbeing (how ever we decide to define it) up or down (or the same).

    It is very easy to grasp if you just simplify the universe to one small island with 2 people on it (maybe some fruit) that exist for one week. It is very easy to realise that there are better or worse actions these people can take to make their stay there as pleasent as possible. Our universe is the same just a bit bigger.

  3. This is an excellent analysis Dr. Lindsay. Great comments too everyone. I'll be taking on the Moral Landscape Challenge and I think what I read here resembles my own conclusions.