Part 1 is here.
This may sound like a joke, but one of the serious problems with The Better Angels of Our Nature is that it's too short. A truly diligent scholar could write an entire book of this length on topics that Pinker rushes through in thirty-page chapters or sections of chapters. The desire to write a comprehensive book on violence has prevented him from writing a good one. The consequence of this compressed approach is that Pinker can seem to weigh opposing ideas diligently while actually examining them in a lopsided manner that I'd call polemical if I was more confident Pinker was aware of what he was doing.
If I describe his argument as reductive, defenders can point out scattered admissions of complexity, but it's the easiest thing in the world to list ambiguities in the abstract without applying them to your concrete conclusions. Lazy undergraduates (reader, I was one) know all about this, and professors who mean to be thorough aren't immune to it either.
A related problem is that, despite Pinker's occasionally slipping into colloquialism, this is fundamentally an academic book, dense with facts and arguments whose implications are described rather than explained. In a book aimed at a professorial audience, with the time, resources, and training to evaluate Pinker's claims, this approach would be fine. But The Better Angels of Our Nature is explicitly a book for ordinary readers, who are more likely to accept its overwhelming barrage of information without tracking the precise logic. A good book on popular science or other popular academics details its arguments clearly in accessible language. But very few such books exist, and this isn't one of them.
Chapter 2 (again)
While reading chapter 3, I came across a claim so unscientifically reductive that I felt compelled to check the sources Pinker cited for it. Unsurprisingly, I found some problems, but the important point right now is that one of them was to Lawrence Keeley's book 1996 War Before Civilization. That book, which made a splash in the popular non-fiction market on release, is Pinker's principal source for the claim in chapter 2 that nonstate civilizations were more violent than the states that evolved later, and for the data that back up that claim. I decided it would be worthwhile to dig around a little for responses to that book. I didn't find much, which according to preference you can interpret as a sign that the book was so brilliant it was hard to reply, a sign it was so ridiculous there was point in replying, or a sign that I'm not good at finding evidence of academic articles on Google. But I did come across the work of anthropologist R. Brian Ferguson, whose website includes PDFs of a number of articles on war in the pre-state world, some of which respond directly to Keeley (who had himself criticized Ferguson's approach). Even if you're not directly interested in this topic, I recommend taking a quick look at some of Ferguson's work, to get a sense of how complicated is the historical record that Pinker reduces to an incredibly violent past, and how distorted is his characterization of scholars who question that reduction as a "Peace and Harmony Mafia" pushing the image of the noble savage. Again, I'm not endorsing either side in the anthropological debate; I'd have to do a lot more reading to come to a decision on that. I'm only saying that Pinker fails to fulfill the responsibility of a popularizer to accurately reflect the state of the scholarly literature, even the parts that don't agree with him.
Because Pinker throws so much information and argument at the reader, even someone who's trying to read slowly and carefully can overlook odd or misleading assertions, including those with large consequences. Here's a quote from page 52: "During the 20th century the United States acquired a reputation as a warmonger, fighting in two world wars and in the Philippines, Korea, Vietnam, and Iraq. But the annual cost in American lives was even smaller than those of the other great powers of the century, about 3.7 per 100,000." This is equivalent to making the following statement about a father: "He has acquired a reputation as a child abuser, beating all of his children bloody several times a year. But he has never been injured himself." Two facts that are at best unrelated, and at worse in a mutual causal relationship, are presented as though they contradict each other.
Pinker's emphasis on the U.S. and Europe for statistics on contemporary wartime deaths (though to his credit he does include worldwide figures, which are also comparatively low) is part of a general habit of overemphasizing declines in violence and death in "major" and "developed" nations, which he attributes to, among other things, a civilizing process that will be the subject of chapter 3, which other nations haven't been privileged to experience. An alternate possibility, that rates of violent death remain higher in "developing" nations because Western technology has allowed the prosecution of wars that rain death on the citizens of those nations while leaving Western soldiers comparatively safe, has yet to come up.
Having argued in chapter 2 that violence declined after the rise of states, Pinker now elaborates that it declined further across the second millennium of the common era. Although he's once again blithely combining different types of data of varying reliability, they're more safely comparable than those from the previous chapter, and I'm willing to concede that violence did indeed decline over the given period, if not in the easily measurable way that Pinker's charts imply. Where Pinker falls flat is in his explanation for the decline, which, if I may be forgiven a little reductiveness of my own, is that people stopped murdering each other because they developed table manners.
He begins on a personal note, explaining how he never understood "the rule of table manners that says that you may not guide food onto your fork with your knife." But the scales fell from his eyes when he read "the most important thinker you have never heard of," one Norbert Elias. As with previous chapters, Pinker builds the rest of his argument around uncritical quotation of a particular thinker: Hobbes, Keeley, and now Elias. The gist of Elias' argument is that a civilizing process occurred, symbolized by the rise of medieval etiquette manuals. Pinker, never one to pass up the chance to describe the past as disgusting, quotes many colorful rules from these manuals, and links learning not to spit or piss in public with learning not to stab people to death, because both involve the inhibition of impulses. "The childishness of the medievals," which Pinker thinks has sometimes been exaggerated but with which he basically agrees, gave way to the maturity of moderns, which descended from the upper classes to the lower, but not always evenly.
This is one of those cases where Pinker will admit the weaknesses of his argument without really confronting them. Toward the end of the chapter he acknowledges that in a contemporary setting, informalization, the opposite of this civilizing process, is not consistently linked with changes in the level of violence. I've never even heard of this rule about moving food onto a fork with a knife, and if someone attempted to introduce me to it my inclination would be to tell them (softly and non-threatingly) to fuck off, but I don't think I'm any more likely than Pinker to behave violently. Pinker tries to close this gap by claiming it's only the entrenchment of nonviolent values that has allowed the dissociation of violence and politeness, but a more rigorous intellect would give serious consideration to the possibility that the correlation of rising politeness and declining violence in the middle ages was not causative, at least not in the precise way he interprets it.
Before getting into alternate interpretations of that decline in violence, I'd like to turn briefly to another couple instances of Pinker being cavalier with data. While reading reviews of The Better Angels of Our Nature, I came across an Internet comment questioning whether improvements in medical science over time might have had an impact on rates of violent death. That struck me as an interesting angle, and I was waiting for Pinker to get around to discussing it. Imagine my delight on discovering the following, his sole comment on medical care and violence prior to the 20th century: “doctors before the nineteenth century were quacks who killed as many patients as they saved." I don't think there's better evidence of Pinker's reductive tendencies when dealing with problems of his thesis. I hope it isn't necessary to explain that the undeniable badness of pre-19th century medicine when compared to the modern variety is not proof that pre-19th century medicine was equally bad in all places across centuries, or that its evolution had no effect at any point on rates of violent death. Comically enough, Pinker provides a footnote for this broad assertion, and though chasing down particular passages from Pinker's sources is not something I intend to do a lot of unless someone offers me access to an academic library and payment for my time, in this case I was interested enough to do some Googling.
The first source Pinker cites here is Lawrence Keeley's book, the relevant pages of which are long on anecdotal descriptions of bad medical practices and short on actual evidence that medieval medicine was consistently bad. But it's unfair to blame Keeley for this, since his argument in that passage is not about modern vs. medieval medicine, but about medieval vs. prehistoric. He's arguing that in some ways ancient tribes had better medicine than medieval states. Whether or not that's true (again, I'd have to do more reading), it goes against the incredible breadth of the Pinker statement it's cited as supporting.
Pinker's second source for that citation is an article that informs us that "most authors agree" on the insignificance of medicine for changing levels of violent death, and endorses another scholar's model for using information about time lapse between injury and death to estimate how many deaths might have been prevented by medical technology. The original article from which that model is derived presents it "with huge qualifications," bases it in part on personal judgments of "doctors familiar with violent assaults," and uses it only in the context of comparing the present day with nineteenth-century New York City. Used this way, such a model is defensible, even if it smacks of speculation in the pursuit of a definitive conclusion. But by a scholarly game of telephone, it becomes part of Pinker's reason to dismiss the evolution of medicine utterly.
Also discussed in the article with that model is an issue that one might expect Pinker to mention, since it counterbalances and might arguably negate the possible effect of modern medicine: the increased efficacy of modern weapons. But Pinker can't really bring that up, because one of the ways in which he has countered the general impression of contemporary violence is by emphasizing the goriness of ancient weapons. One of his anecdotes in chapter 1 is a prolonged description of the damage ancient Greek spears could do. Injuries that are unpleasant to read about and injuries that are more reliably fatal are, of course, not the same thing, but Pinker has so thoroughly blurred this distinction that he can't rescue it now.
Pinker's dismissal of non-modern medicine is of a piece with the general tone of his book, which emphasizes at every chance it gets how miserable various aspects of pre-modern life were. I'm broadly sympathetic with that: I'm gay, I'm an atheist, and I have a degenerative hip condition that has required and will require multiple surgical interventions to keep me out of a wheelchair. I have very little nostalgia for the past. But Pinker's views go beyond sensible jaundice toward a (here's that word again) reductive Whiggism that colors his interpretation of evidence.
One last point about presentation of data. On page 64 we have Figure 3-4, title "Homicide rates in Western Europe, 1300-2000, and in nonstate societies." The chart above places those nonstate societies on the left as a vertical series of data points, to the right of which is a roughly horizontal curve showing the homicide rates over time. The obvious implication of this presentation is that the nonstate data chronologically precedes the Western European numbers, but as anyone who has looked closely at these particular nonstate data will know, they're actually from roughly contemporary nonstate societies. The progression indicated by the chart is not objectively historical, but based on Pinker's subjective notions of social evolution. It's a consistent problem of his treatment of nonstate societies that, while making the theoretical claim that their historical "location" shouldn't matter, he doesn't state openly enough that these societies are contemporary, and have been affected by various aspects of the civilized modernity he presents as diminishing violence.
To turn from the concrete to the theoretical, my principal complaint with Pinker's notion of civilizing processes in the decline of violence is that he attributes too much causation to the social and cultural aspects of that process, ignoring their effect on more fundamental aspects of ordinary life. This has, I think, to do with his failure to consider in this context certain possible broader causes of violence. I don't have the patience or the psychological and historical grounding to do justice to the question of why people are violent, but one possible aspect of an explanation has been on my mind as I read Pinker's book. Might people be more likely to be violent, even if they regard violence as unfair, if they also regard the society in which they live as unfair? This may sound like the stereotypical liberal habit of blaming violence in poverty, and that's a question I'll get to in a minute, but the applications of this notion are much broader.
Hobbes talks about life in the war of all against all as being "nasty, brutish, and short," but even if we set aside violence the life of the distant past was, as Pinker so enjoys pointing out, unpleasant in many other ways. It wasn't just violence that ended medieval lives on a scale moderns regard as horrific. If apparently healthy people can drop dead of a mild wound, a toothache, or no visible cause at all, is it less obviously wrong to end their lives by violence? If life is a hard, cruel, geographically limited and socially proscribed experience, does destroying it seem a less significant act? I don't know, but I imagine it's worth debating, whether those limitations to fairness exist in medieval London or 21st-century New York City. Pinker is good at pointing out how life-denying ideologies, whether religious or honor-driven, can encourage violence, but what about less formalized intellectual processes? He's so caught up by the (comparative) benevolence of particular elite institutions that he can't recognize possible influences from the thoughts of "common" people.
When Pinker does occasionally discuss such motives, he's as brief and glib as he always is when treating topics he regards as distractions. On page 84, for example, we have this: "A third dubious belief about violence is that that lower-class people engage in it because they are financially needy (for example, stealing food to feed their children) or because they are expressing rage against society. The violence of a lower-class man may indeed express rage, but it is aimed not at society but at the asshole who scraped his car and dissed him in front of a crowd."
The first half of that is an absurd straw-man version of arguments about violence and poverty. I can't speak for everyone who's suggested such connections, but most of the time literal theft to feed starving children is a thought experiment in ethical complexity, not something anyone thinks happens very often. The second half confuses the aim of violence and its source. As anyone who has ever lashed out at the wrong person in a moment of frustration can attest, we don't always target the things that make us mad; we aren't even always fully aware what those things are. Unless he's cannily trying to generate sympathy, a lower-class man won't claim he was motivated by social rage, but that doesn't mean it wasn't a contributing factor in his behavior.
Apt to acknowledge injustice in the abstract but rarely granting it any importance (when addressing the nature of the 1960s counterculture, he gives such "hit[s]... to [elite] legitimacy" as pervasive racism and the "illiberal" war in Vietnam one paragraph, and loud music, poor hygiene, and rejection of self-control ten paragraphs), Pinker is the model of an elite liberal. In my commentary on the rest of chapter three, which deals more fully with American history, I'll explain further how that's so, but for now it's worth noting that, whatever Pinker's own political positions and intentions, his work has obvious implications for policies on and attitudes toward violence.
Books like Pinker's, presented as groundbreaking but actually the rejuvenation of old unscientific ideas with a new (but false) intellectual gloss, marketed to mass audiences but not written in a style accessible to them, exist not to be read and analyzed but to be pointed to in justification of particular policies and beliefs. In this sense The Better Angels of Our Nature is comparable to titles like The Bell Curve and Hitler's Willing Executioners. And that, unless you're thinking solely in terms of sales figures, is not distinguished company to be in.