Pinker is wrong

A bit late, I know, but I just read Steven Pinker's "Science is Not Your Enemy," and I have to say I think he is badly wrong. Not that science is your enemy, but much, perhaps most, of what he says in defence of the thesis in his title is false.

He begins with the claim that the "great thinkers of the Age of Reason and the Enlightenment were scientists." In the next paragraph he names them, and they are all philosophers: Descartes, Spinoza, Hobbes, Locke, Hume, Rousseau, Leibniz, Kant, Smith. Now Descartes certainly was a scientist as well as a philosopher, Leibniz was a mathematician, and Hume a historian who wanted, in his philosophical work, to be a second Newton of the moral sciences (i.e., roughly, a psychologist), but they are all philosophers, and their philosophical work is not the same kind of thing as their work on optics, calculus, etc. Pinker seems to be blurring this distinction from the start. He claims that methods and insights from the sciences are shedding light on ancient problems and that "writers in the humanities" should be delighted about this. So now philosophy is all the humanities, and the fact that some philosophers also did science means that distinctions between science and the humanities are artificial and bad. He doesn't quite say that, but it seems to be the idea. When you do say it explicitly it is clearly a bad idea.

From here he moves on to defend 'scientism,' reclaiming the word from those who use it in an allegedly vague and certainly pejorative sense. As he redefines it, 'scientism' is distinguished by commitment to the beliefs that the world is intelligible and that the acquisition of knowledge is hard. The intelligibility claim is connected to the belief that phenomena can be explained by principles that are more general than the phenomena themselves. 

One immediate problem with this is that understanding comes in different forms. It is one thing to understand a thing in terms of some general principle, another to understand something particular in its particularity, i.e. not as an instance of a general rule. I need to give examples or this idea will be very obscure. Say my mild-mannered neighbour turns out to be a serial killer and I struggle to understand how this could be, how he could have killed all those people. Perhaps all I need is statistical information about the correlation between mildness of manner and violent behavior. But perhaps that will not help me at all. Perhaps what I don't understand is how this man's particular mildness could coexist with the brutality of these crimes. I need some way to see the gentle neighbour as the same person as the serial murderer. One thing that might help here would be if he is like some character in a novel and you point this out to me. Then I can read the novel and, if you are right, I might gain the insight, the understanding, that I need. But novelists are not scientists. They don't offer general principles. Another example would be if I want to understand how Hamlet or a painting by Jackson Pollock works. I might want to know things about the brain and the effects on it of certain patterns of sound or colour, but what I want is probably something quite different. I don't want general laws. I want to know how this work is put together, how its various parts rhyme or contrast. To understand Hamlet, for instance, I need to know the plot, the characters, and the language of the play. I need to understand how these things work together in the play, in this play. What I want to understand is essentially particular. General principles can help at most only a little.

Maybe that's still too obscure. One way to understand something is to see it as an instance of a general rule, and science does a good job of providing these rules. But sometimes what we want is to understand a particular person or thing not as an instance of a general rule but as the particular thing it is. Why do the particular features of this work go so well together? It can't be because of some general rule because only this work has these features in just this relation, and telling me that any work that has these features in this relation is good is telling me nothing helpful. "Hamlet is good because of its Hamlet-iness" is not illuminating. Nor is, "When you watch Hamlet blood flows to these parts of your brain." If I want to understand a person, similarly, it might help if I know that all teenagers are like that or that such-and-such behaviour is common in old men, but equally this might just make me puzzled about all old men, or simply be no help at all. If I want to make sense of x it might be that I need to find a way to relate various things about x to various other things. Putting things in context is a way of understanding that does not involve applying general rules to particular cases.  

A second problem with Pinker's benign scientism is that according to it "the findings of science entail that the belief systems of all the world's traditional religions ... are factually mistaken." At most this applies to literalist versions of religious belief, and even there it is hard to believe that there isn't something else going on, that the people who insist we could get to heaven in a rocket or find hell by digging in the right place would at some point in the experiment reveal that they had not meant all of it, including the 'literally' part, literally. Maybe I'm over-optimistic there, but the idea that science could prove that God did not create the world, or become a man, or send his messenger to us, shows an outstandingly crass conception of religious belief (at least for someone who wants to be taken seriously as a critic of religion). This is a point that has been made repeatedly in response to people like Dawkins, but Pinker seems to be ignoring it. 

He then moves on to ethics:

in combination with a few unexceptionable convictions— that all of us value our own welfare and that we are social beings who impinge on each other and can negotiate codes of conduct—the scientific facts militate toward a defensible morality, namely adhering to principles that maximize the flourishing of humans and other sentient beings.   

Science, or facts combined with "unexceptionable convictions", supports utilitarianism! Incredible that seemingly intelligent people would still bother to think about ethics any other way! Who needs philosophy? Incredible also that Pinker would come so close to endorsing Mill's version of utilitarianism without, apparently, having read what Mill has to say on the subject. Mill offers a reasonable defence, of course, but he is far from claiming that the issue needs no argument. Indeed he brings out just how much room for debate there is on, for instance, the nature of flourishing.

When Pinker moves on to politics it is hard not to see shades of Jonathan Haidt:

The new sciences of the mind are reexamining the connections between politics and human nature, which were avidly discussed in Madison’s time but submerged during a long interlude in which humans were assumed to be blank slates or rational actors. Humans, we are increasingly appreciating, are moralistic actors, guided by norms and taboos about authority, tribe, and purity, and driven by conflicting inclinations toward revenge and reconciliation.

There is more wrong with Haidt's thinking than I can go into now, but for an introduction see here. Pinker has a lot of the same flaws, especially ignorance of the philosophical work on the topic he is discussing and ignorance of the fact that what he is getting into is philosophy.

I largely agree with all the criticisms of Pinker's essay made here, here, and here, but let me try to add a brief point of my own. Ignorance of philosophy is certainly a big part of Pinker's problem, but so is the difference between matters of fact and matters of value. If I want to know what makes a poem so good science won't help me. If I want to know in what circumstances if any abortion is OK science won't help me. If I want to know how best to relate one thing to another science will often be of limited use to me. This is not to deny that science has great value, or that open debate and hard work are valuable. It is, though, to push back against the idea that science is the way to solve all problems, i.e. to push back against scientism.

Let me end with some points made by others I agree with to save you from having to click on all the links above. First, Massimo Pigliucci:

Yes, quantitative methods can (and should) be used by historians, though this will always likely be complementary to, rather than substitutive of, classical historical methods. And yes, the fruitful collaboration between philosophers of mind and cognitive scientists is a shining example of how to bridge the divide between the two cultures. But no, quantitative analyses of Jane Austin novels interpreted in evolutionary psychological key are frankly ridiculous (I've seen it done), and while clearly the study of the physiology of visual or auditory perception are fascinating fields in their own right, they are far less useful to my enjoyment of a Picasso or a Beethoven sonata than knowledge of the history of art or of music.

Next, P. Z.Myers:

Science is a fantastic tool (our only tool, actually) for probing material realities. Respect it for what it is. But please, also recognize that there’s more to the human experience than measurement and the acquisition of knowledge about physical processes, and that science is a relatively recent and revolutionary way of thinking, but not the only one — and that humans lived and thrived and progressed for thousands of years (and many still do, even within our technological culture!) without even the concept of science.

Scientism is the idea that only science is the proper mode of human thought, and in particular, a blinkered, narrow notion that every human advance is the product of scientific, rational, empirical thinking. Much as I love science, and am personally a committed practitioner who also has a hard time shaking myself out of this path (I find scientific thinking very natural), I’ve got enough breadth in my education and current experience to recognize that there are other ways of progressing.

And finally, Ross Douthat:

Like Sam Harris, who wrote an entire book claiming that “science” somehow vindicates his preferred form of philosophical utilitarianism (when what he really meant was that if you assume utilitarian goals, science can help you pursue them), Pinker seems to have trouble imagining any reasoning person disagreeing about either the moral necessity of “maximizing human flourishing” or the content of what “flourishing” actually means — even though recent history furnishes plenty of examples and a decent imagination can furnish many more. Like his whiggish antecedents, he mistakes a real-but-complicated historical relationship between science and humanism for a necessary intellectual line in which the latter vindicates the former, or at least militates strongly in its favor. And his invocation of “the scientific facts” to justify what is, at bottom, a philosophical preference for Mill over Nietzsche is pretty much the essence of what critics mean by scientism: Empirically overconfident, intellectually unsubtle, and deeply incurious about the ways in which human beings can rationally disagree. 

P.S. Then there's Ray Monk on Wittgenstein and scientism:

Scientism takes many forms. In the humanities, it takes the form of pretending that philosophy, literature, history, music and art can be studied as if they were sciences, with “researchers” compelled to spell out their “methodologies”—a pretence which has led to huge quantities of bad academic writing, characterised by bogus theorising, spurious specialisation and the development of pseudo-technical vocabularies. Wittgenstein would have looked upon these developments and wept.

There are many questions to which we do not have scientific answers, not because they are deep, impenetrable mysteries, but simply because they are not scientific questions. These include questions about love, art, history, culture, music-all questions, in fact, that relate to the attempt to understand ourselves better. There is a widespread feeling today that the great scandal of our times is that we lack a scientific theory of consciousness. And so there is a great interdisciplinary effort, involving physicists, computer scientists, cognitive psychologists and philosophers, to come up with tenable scientific answers to the questions: what is consciousness? What is the self? One of the leading competitors in this crowded field is the theory advanced by the mathematician Roger Penrose, that a stream of consciousness is an orchestrated sequence of quantum physical events taking place in the brain. Penrose’s theory is that a moment of consciousness is produced by a sub-protein in the brain called a tubulin. The theory is, on Penrose’s own admission, speculative, and it strikes many as being bizarrely implausible. But suppose we discovered that Penrose’s theory was correct, would we, as a result, understand ourselves any better? Is a scientific theory the only kind of understanding?