A big book: the main text (in my paperback edition) is nearly 700 pages, unwide margins, and unsmall type. (There are occasional graphs, tables and illustrations, but still: it's a project.) But it's written by Steven Pinker, a guy I've enjoyed reading in the past. And it got rave reviews when it came out a few years ago. And cover blurbs most authors would kill for, for example: “One of the most important books I’ve read—not just this year, but ever.”—Bill Gates.
[Of course, Steven Pinker wouldn't kill for a blurb like that. As the book makes clear, he's a pretty peaceful guy.]
So yes, it's pretty good. If you're willing to invest the time on something a little denser than the latest Lee Child novel, I can recommend it.
The subtitle is: Why Violence Has Declined. Pinker first attempts to show that it has declined, and he is pretty close to irrefutable here. First, there is the trend of centuries and millennia: he debunks the hazy-hippie myth of ancient tribes living gently in sustainable communities. Instead, it was amazingly likely back then that you'd be violently killed by warriors of a neighboring tribe; if you were male, and lucky, you might take out a few of your antagonists first. Pinker argues, to the discomfiture of anarchists, that the development of the modern "leviathan" state moderated tribal violence to a fraction of the historical rate.
The modern state, of course, has its own nasty record of murderous violence, both against other nations and (in many cases) against its own citizenry. But (again) Pinker shows that this trend is also downward over the past centuries. There's the notable exception of what's called the "hemoclysm" ("blood flood") of the first half of the 20th century; Pinker argues that this really was an exception to the overall historical trend: states have gotten significantly better-behaved since then. (Yes, when you look at, say, Syria, things can seem bad; but they were much worse before, with much larger levels of violence going unremarked because they were so common.)
Pinker also considers interpersonal violence of all kinds: homicide, rape, assault, infanticide etc. To the extent that reliable statistics can be had, the trends are downward over the long term. (I was wondering if Pinker was going to look at abortion; yes he does. Although his discussion probably wouldn't satisfy the National Right to Life Committee, it's remarkably even-handed for a Harvard prof.)
After thoroughly documenting violence's decline over the years, Pinker gets to the topic promised by his subtitle: why has it happened? Pinker is a psychologist, and goes into great detail on brain physiology and function. ("The orbital cortex is strongly connected to the amygdala, hypothalamus, another parts of the brain involved with emotion.") How do violent thoughts get generated, and how do they get translated into action? There might be evolution at play, with selection over the past centuries operating in relative favor of brains that are better at controlling impulses, for example. (But, Pinker cautions, maybe not.)
Instead, Pinker argues that the decline is more likely due to actual old-fashioned progress: the spread and interconnectedness of rational thought, the easy availability of information, the victory of positive-sum free-market mechanisms over negative-sum command-and-control diktat. (And there's also the Flynn Effect: we really are just getting smarter as the years go on.)
Pinker is one of the best popularizers of science today. His style is, as always, accessible, occasionally funny, and very wide-ranging, with lots of pop culture references. One can imagine how the chapters developed from lectures provided to easily-bored Harvard students. Pinker is occasionally glib and simplistic, especially when wandering too close to current American politics. (He engages in some regrettable Dubya-bashing, which he supports by quoting dubious research.) But it's a big book, and you can hit the fast-forward during these parts and not miss much.