Pinker: I Don’t Get It
Ross Douthat identifies a serious flaw in the reasoning in an article that Steven Pinker published recently in the New York Times. Based on this and other posts about it, I read the original article. I’m hesitant to say this for fear that I’m the one not getting the joke (Pinker is a famous Harvard professor and all these smart people seem to take it very seriously), but the article struck me as poorly reasoned from front to back. I’ll focus only on one problem that I think is central.
Pinker cites the following hypothetical:
Julie is traveling in France on summer vacation from college with her brother Mark. One night they decide that it would be interesting and fun if they tried making love. Julie was already taking birth-control pills, but Mark uses a condom, too, just to be safe. They both enjoy the sex but decide not to do it again. They keep the night as a special secret, which makes them feel closer to each other. What do you think about that — was it O.K. for them to make love?
Referring to this and a couple of other hypotheticals, he says that:
Most people immediately declare that these acts are wrong and then grope to justify why they are wrong. It’s not so easy. In the case of Julie and Mark, people raise the possibility of children with birth defects, but they are reminded that the couple were diligent about contraception. They suggest that the siblings will be emotionally hurt, but the story makes it clear that they weren’t. They submit that the act would offend the community, but then recall that it was kept a secret. Eventually many people admit, “I don’t know, I can’t explain it, I just know it’s wrong.” People don’t generally engage in moral reasoning, Haidt argues, but moral rationalization: they begin with the conclusion, coughed up by an unconscious emotion, and then work backward to a plausible justification.
Here’s my hypothetical:
Julie is traveling in France on summer vacation from college with her brother Mark. One night she decides that it would be both interesting and fun to shoot Mark in the head. She takes careful precautions not to be caught, shoots him, hides the body and invents an alibi. She is never suspected of committing this action. She keeps it a special secret that makes her feel more in control of her own life. What do you think about that — was it O.K. for her to kill her brother?
The obvious difference is that coercion, more specifically violence causing physical harm, occurred in the second case. So what? Have you assumed the “conclusion, coughed up by an unconscious emotion” that it’s bad to kill people?
No, of course not, you’re a sophisticated moral thinker. As per Pinker, you recognize that we have evolved (or to be more precise, it is “plausible” that we may have evolved) so that we have biochemical processes that pre-dispose us to be averse to violent coercion (teenage boys, maybe not so much). After all, social groups can be more successful when people behave this way (or so it is plausibly theorized).
But why should Julie give a crap? After all, the premise of the hypothetical is that she finds it fun and interesting to kill her brother, suffers no material deprivation as a result, and feels no remorse afterwards. If you argue that nobody would ever feel this (an empirically false claim, as sociopaths do exist) and it is Julie’s anticipatable feelings of remorse that should prevent her from doing this, you are merely making the prudential argument that Julie is disadvantaged by this action – that some quirk of human programming makes us better off by not doing this. But then, the moral admonition against killing is merely a rule like “don’t touch a hot stove”, and your definition no longer allows for a special category of action called “moral” as opposed to “intelligently self-interested”.
But suppose, as per the hypothetical, that Julie is a sociopath. That is, it is in her material self-interest to kill h brother. Is it “wrong” for her to do it?
This is where Pinker completely falls down. He carefully avoids addressing this question. After using pages to mike-check various psychology experiments and theories, he gets to the point:
Perhaps we are born with a rudimentary moral sense, and as soon as we build on it with moral reasoning, the nature of moral reality forces us to some conclusions but not others.
Moral realism, as this idea is called, is too rich for many philosophers’ blood. Yet a diluted version of the idea — if not a list of cosmically inscribed Thou-Shalts, then at least a few If-Thens — is not crazy. Two features of reality point any rational, self-preserving social agent in a moral direction. And they could provide a benchmark for determining when the judgments of our moral sense are aligned with morality itself.
But he’s dodged the question. We are “born with a rudimentary moral sense” that conforms with the “nature of moral reality” – but is this merely a physically-contingent product of evolution, like the distribution of eye colors, or is it the implantation via a physical mechanism of moral laws that are independent of human opinion?
Consider his subsequent review of “two features of reality” that “point any rational, self-preserving social agent in a moral direction.” Ross’s criticism of these points strikes me as correct, but way too mild.
First, Pinker says that:
One is the prevalence of nonzero-sum games. In many arenas of life, two parties are objectively better off if they both act in a nonselfish way than if each of them acts selfishly. … Granted, I might be a bit better off if I acted selfishly at your expense and you played the sucker, but the same is true for you with me, so if each of us tried for these advantages, we’d both end up worse off.
This is simply a form of the argument that I am better off materially by cooperating and we’re back to the “hot stove” problem. That is, if I can get $10 by killing somebody in a manner guaranteed never to be detected, where is the argument that I shouldn’t do it? I’ll also note that his claim is empirically nonsense. I’m often way better off, at lest materially, playing you for a sucker if I’m more cunning and ruthless than you, often by cooperating with some third party to hose you. It’s like the guy never spent ten minutes in a high school.
Pinker then asserts, as the second “external support for morality”, that:
If I appeal to you to do anything that affects me — to get off my foot, or tell me the time or not run me over with your car — then I can’t do it in a way that privileges my interests over yours (say, retaining my right to run you over with my car) if I want you to take me seriously.
It’s kind of embarrassing even to criticize this. Isn’t obvious that in a such a situation I need to appeal to you in a way that doesn’t appear to privilege my interests over yours. Has Pinker never bought a car, negotiated a raise or listened to a politician give a speech?
Pinker was obviously wrestling with a huge issue in this essay, and has a lot of relevant knowledge, but it seems to me that he flinches from addressing the central question: as Ross put it in the title of his post, “Why Should We Be Moral”?
Since we’re conducting ‘innocent’ thought experiments, let’s ask aloud what exactly it is that we have to do voluntary — and indeed self-conscious — violence to to decide that it’s “fun and interesting” to carry out an incestuous sex act…and then to act on that judgment. Then let’s ponder whether the morality question perhaps has something to do with the responsibility one undertakes in choosing to disobey inherited moral interdicts (particularly the most venerable of moral interdicts), and then to indeed ensure that one does so. Far more interesting than Pinker’s asinine and lame hypothetical is the Blue Lagoon Hypothetical, the case of the brother and sister who didn’t know they were siblings and wound up in a romantic relationship. There’s a case in the news right now I think about twins who were separated at birth and fell in love and married each other. Moral philosophy here points us, in my judgment, in a RADICALLY different direction from the one we head toward given the brother-sister we’re-bored-let’s-screw schema. Why? The answer throws up some pretty worthwhile insights about the moral case for morality.
— James · Jan 16, 04:08 AM · #
I think that you folks may have misunderstood Pinker’s objective. His “two features of reality” . . .“point any rational, self-preserving social agent in a moral direction.” It seems to me that he is arguing only that they predispose humans to have moral systems.
Pinker claims that nonzero-sum games are prevalent. Citing exceptions doesn’t disprove his claim, nor do sociopathic behaviors. There is interesting research that seeks to identify stable social systems that find a need for some cheats (or defectors in game theory speak) to encourage the prevalence of cooperators.
A similar type of reasoning refutes your claims about the lack of need for reciprocation. Pinker’s claim is that this is required “if I want you to take me seriously.” The car salesman, often, does not. And, obviously, politicians do not expect to be taken seriously. They are judged for their ability to lie convincingly, but everyone knows that they are lying. When people speak of competent politicians this is what they mean.
There is game theory support for this too, and some attempts to define optimum strategies. One very successful strategy is grandma’s dictum: fool me once, shame on you; fool me twice, shame on me. This is also called tit-for-tat. A variant that is sometimes better is tit-for-two-tat. I’ll give you a second chance to be a stand up guy, but not a third.
You may disagree with Pinker, but you don’t seem to have actually engaged with his arguments so far.
— back40 · Jan 16, 04:54 AM · #
back40:
In a sense, you are right: I don’t address Pinker’s argument. The point of my post was that Pinker never actually addresses the question of morality, but merely considers prudence.
I guess my point could be further illustrated as follows.
Pinker cites arguments for why it can be a successful evolutionary strategy to be cooperative, that the heuristics that lead to such behavior may be hard-wired to a degree in humans, and that therefore there may be many situations in which these behaviors will occur even though they produce no obvious material gain. But consider the human behavior of ducking when something heavy starts to fall on us. This likely has some evolutionary survival value. It is almost certainly hard-wired to a degree, as evidenced by the fact that people duck even if they are in a situation such as a collapsing mine shaft in which ducking confers no material benefit.
How can Pinker’s argument distinguish the propensity to cooperate from the propensity to duck in a way that renders the former “moral”? If someone doesn’t duck when a large branch is headed for them, we might call them foolish or incompetent, but we would be unlikely to call them immoral. Why then, if someone proceeds with a tat-tat-tat-for-tit strategy would we consider them to be immoral, rather than merely foolish or imprudent?
— Jim Manzi · Jan 16, 05:19 AM · #
“The point of my post was that Pinker never actually addresses the question of morality, but merely considers prudence.” Um, but Pinker isn’t even trying to address the question of morality. As he says, “the new science of the moral sense does not make moral reasoning and conviction obsolete.” He is not at all trying to take on the question: if I didn’t already care about morality, why should I? Rather, he is trying to take on the (really more pressing) question: given that I do care about morality, but I also have all this scientific evidence that makes my moral judgments look like they might often be illusory, how do I figure out which judgments to trust and which to discount? And the places where morality and reflective prudence point us in the same direction, he is suggesting, are good places to look for some firm points to stand upon.
— philosopher · Jan 16, 05:31 AM · #
philosopher:
given that I do care about morality, but I also have all this scientific evidence that makes my moral judgments look like they might often be illusory, how do I figure out which judgments to trust and which to discount? And the places where morality and reflective prudence point us in the same direction, he is suggesting, are good places to look for some firm points to stand upon.
Why? It seems to me that often the opposite would be the case.
— Jim Manzi · Jan 16, 05:51 AM · #
Because these can be taken to be actual “features of reality”, in his terms. Note that he is explicitly endorsing the analogy with mathematical knowledge from the previous paragraph — moral realism is a starting point for the paragraphs that follow, not a conclusion from them.
— philosopher · Jan 16, 06:11 AM · #
<i>Isn’t obvious that in a such a situation I need to appeal to you in a way that doesn’t appear to privilege my interests over yours. Has Pinker never bought a car, negotiated a raise or listened to a politician give a speech?</i>
To be generous to the professor, Pinker answers this question in his authors@google youtube video: there are different starting positions in moral judgments, roughly dividing into three categories.
To be ungenerous, Pinker and Hauser do undermine the notion of an extrinsic moral reality, but they cannot face it or do not want to admit it, because both authors fill their books with assurances to the contrary.
The problem, though, is the research is sound. It’s their assurances, and their philosophical commentary, which is flawed, not their science.
The evidence strongly suggests an uncomfortable truth: independent of human cognition, Moral Truth is a meaningless construct.
— JA · Jan 16, 06:16 AM · #
Here are three different questions that one might be asking in this neck of the woods:
(i) Are there moral facts?
(ii) If the answer to (i) is “yes”, then do I have reason to let those facts guide my actions?
(iii) If the answer to (i) and (ii) are both “yes”, then how can I know what these facts are, so that I may let them guide my actions?
Jim has taken Pinker to be giving an answer to (ii), which I have tried to suggest is a mistake. JA takes Pinker (and Hauser, etc.) to give us evidence against (i), but I just don’t see it. I think that Pinker is concerned with the ways in which the scientific results seem to give a skeptical answer to (iii). It is an epistemological question, not a question of practical reason nor of metaphysics.
— philosopher · Jan 16, 06:27 AM · #
In a way, all of this is sound and fury: Morality is just a word to describe a vague category of sensations and intuitions. Perhaps it’s better to say that we are all born with a slew of prepared emotional responses to the environment, but, thanks to neo-cortical self-modeling, we are able to step in, terminate the dominion of instinct by the franchise of reason, and exercise some measure of control over our moral mindspace. Does that clear it up a bit?
— JA · Jan 16, 06:28 AM · #
philosopher,
There are certainly “moral facts”, but none which are independent of human cognition. In other words, there are facts about moral judgments that emerge out of human interaction — facts that scale and facts that don’t — but there the evidence is against any extrinsic moral reality that we either do or don’t tap into when we make moral judgments.
— JA · Jan 16, 06:32 AM · #
To say that there are facts about the moral psychology of humans, but no facts beyond that, is simply to deny that there are any real moral facts. (Compare: we would never say that there are “witch facts”, which are the psychology of people when they think about witches. There just aren’t witch facts. Or, if one would rather, there is just one, completely negative, witch fact.)
Now, that particular brand of moral non-realism may be true, but nothing in the work by Hauser, Haidt, etc. are any good as premises in arguments for such a conclusion.
— philosopher · Jan 16, 06:35 AM · #
I cite their work as examples of rolling back the mystery of moral judgments.
Reality scales from quarks to us without any fundamental break in its laws. The seat of morality, whether it is “moral reasoning” or “moral intuition”, is in the human brain. Outside that realm, or, alternately, outside its daughter realms (society, e.g.), there is no such thing as external, eternal moral truth.
— JA · Jan 16, 06:42 AM · #
That doesn’t really make any sense. No one thinks that a scientific investigation of moral judgement was going to find some weird Cartesian extra-natural link to a splendidly separate moral realm. It’s a starting point of such investigations that our brains are natural organs, made from material substance, governed by physical law. So, if such investigations could show what you say they show, the actual results of such investigations wouldn’t be necessary to show them. Or, to put it differently: there wasn’t any “mystery” to begin with, so there’s no work to be done in “rolling back” this nonexistent mystery.
But they don’t show that at all, or anything close — again, because moral realism just doesn’t require that moral facts have some weird causal status in the world order. Moral facts just don’t need to be that kind of fact. Note that one could, by parity of reasoning, show that mathematics is all a mental myth as well. Most folks would take that to be a reductio of the argument, but if you’re willing to swallow mathematical nihilism, well, you’re welcome to try.
— philosopher · Jan 16, 06:56 AM · #
Very gratifying to see that folks still like to stay up really late philosophizing. Unless y’all are in California, in which case never mind.
— Noah Millman · Jan 16, 09:46 AM · #
Pinker claims that nonzero-sum games are prevalent. Citing exceptions doesn’t disprove his claim, nor do sociopathic behaviors.
As I have said elsewhere, discussions of morality, particularly those that contain any question “what is moral” or “what are the criteria by which a moral ethic can be constructed” cannot use language of pathology. To do so begs the question of what is moral, and suggest a reading of morality that is dependent on norms.
To be ungenerous, Pinker and Hauser do undermine the notion of an extrinsic moral reality, but they cannot face it or do not want to admit it, because both authors fill their books with assurances to the contrary.
And this is my frustration. Philosopher says “Um, but Pinker isn’t even trying to address the question of morality. As he says, “the new science of the moral sense does not make moral reasoning and conviction obsolete.” He is not at all trying to take on the question: if I didn’t already care about morality, why should I? Rather, he is trying to take on the (really more pressing) question: given that I do care about morality, but I also have all this scientific evidence that makes my moral judgments look like they might often be illusory, how do I figure out which judgments to trust and which to discount?
And yet I find that not entirely in the character of Pinker’s piece, and certainly not in the character of the popular understanding of evolutionary moral psychology. The Times, in a piece published earlier this year entitled Scientist Finds the Beginnings of Morality in Primate Behavior", said “Biologists argue that these and other social behaviors are the precursors of human morality. They further believe that if morality grew out of behavioral rules shaped by evolution, it is for biologists, not philosophers or theologians, to say what these rules are.“ These kinds of claims are common, and I confess I find Pinker strongly hinting at the same in his article. What is so frustrating about talking about this subject, not just in reference to Pinker’s article but in general, is that evolutionary psychologists seem constantly to suggest that they have the tools to create a scientifically sound and rigorous approach to morality— that they can use science and the sense of certainty it brings to craft a “true” moral ethic. And yet when the leaps of logic that are required to make that true are pointed out, there seems to be a backtracking (as is going on here) about precisely how strong of a claim is being made by evolutionary psychology. It makes for a discouraging conversation when the goal posts seemed to be moved so often.
Outside that realm, or, alternately, outside its daughter realms (society, e.g.), there is no such thing as external, eternal moral truth.
I find more and more people amenable to this claim, particularly among people in the sciences. And yet I’m discouraged by the refusal to follow this kind of reasoning to its logical ends, and recognize that there is also no external, eternal scientific truth. But that’s a discussion for another time.
— Freddie · Jan 16, 04:06 PM · #
And yet I’m discouraged by the refusal to follow this kind of reasoning to its logical ends, and recognize that there is also no external, eternal scientific truth.
I agree with this. Scientific truth, insofar as it exists, is refracted — or, as Husserl would say, bracketed. Remove the sentient brain, and poof, “scientific” truths disappear.
However, without a doubt there is an external reality independent of humanity. The goal in science — and philosophy — is to get our bracketed truths as close as possible to reality’s unbracketed truths.
An example: think about renormalization. In quantum electrodynamics, the unbracketed truth of reality throws out infinite integrals for even the simplest of problems. Now, obviously we cannot compute infinite integrals, so we were forced to develop conceptual tools to allow us to approximate what reality really is. But what tools! — the results acquired by using the conceptual tools of quantum electrodynamics matches experimental data to an unbelievable degree: nine decimal places and counting, by far the most precise and successful conceptual tool in science.
So, yes, renormalization cannot be considered an external truth, but it is so isomorphic to reality that it doesn’t matter. And thus, our standard. Is it more isomorphic to reality to say that morality exists independently of human cognition, or to say that morality is a complex combination of intuition, emotion, imagination, and self-reflection, completely dependent on human cognition? I think it’s pretty clear it’s the latter.
Think, if one of the parameters of the universe were changed ever so slightly — say the spin of the electron — and as a result galaxies, stars, planets and us never existed, would it still make sense to talk about morality?
— JA · Jan 16, 06:09 PM · #
I’ve read the Pinker article, and your and Routhat’s responses to same. I think Pinker’s intentions have been misread, perhaps even by Pinker himself.
He’s primarily describing moral thought, as opposed to proscribing moral thought. There are strong parallels in grammar (which is more in Pinker’s line). Is a language’s grammar a formula for how one should speak it? Or is it a recipe for how it is spoken? Almost by definition a descriptive explanation cannot serve as a guide for “individual action.” (I can devise a functional grammar for Ebonics but it won’t provide any argument for why I should or should not speak it.) Pinker reaches for a proscriptive grammar by the close but ultimately fails — not because of his bias that morality should be rational, but because rational theories of morality primarily describe only what people do, not what people should do.
If you — or, for that matter, Pinker— are looking to neuroscience to figure out why some things feel wrong, you’re looking in the wrong places. That’s like looking to neuroscience to help with the declension of the verb “to be.”
— Paul S · Jan 16, 08:59 PM · #
Paul S:
rational theories of morality primarily describe only what people do, not what people should do.
But isn’t “what people should do” pretty much the common-sense definition of morality? (I mean this as a legitimate question, not a snide comment.)
— Jim Manzi · Jan 16, 09:38 PM · #
Freddie: There is a natural confusion that people can fall prey to here, because the word “morality” is ambiguous between something like “the practices and psychology of moral judgment” on the one hand, and something like “the actual moral facts (or the lack thereof)” on the other. But I don’t see any evidence at all that Pinker is confused about this — rather the opposite, in fact. And in the NYT article you mentioned, really only Franz de Waals seems to be making anything like this confusion; both the journalist and many of the people cited in the article are very clear about respecting those two different things.
Paul S.: I think you mean “prescriptive”— only a wanton amoralist would prescribe a proscription on morality. ;-) But I think you’re halfway right about Pinker; only halfway, though, because in the article that has sparked all this attention, he’s clearly concerned with the question of, given what the kind of descriptive account that is developing, how can we still claim to be able to tell moral right from wrong? And this is a normatively loaded question, albeit (I’ve been arguing) an epistemological one.
— philosopher · Jan 17, 08:43 AM · #