The Better Angels Of Our Nature Why Violence Has Declined

Reading Pinker's Argument on Violence Decline

I picked up Steven Pinker's book about five years ago when someone in a forum thread brought it up as a talking point during a discussion about rising global tensions. I expected another armchair history piece dressed up as data visualization. It wasn't. The central thesis tracks violent death rates across centuries and finds they drop significantly when you account for population size, which is something most people don't realize when they hear raw conflict numbers trending upward in recent decades. The Better Angels Of Our Nature Why Violence Has Declined sits somewhere between anthropology, criminology, and historical analysis. Pinker organizes his evidence into five different types of violence: wars, genocides, homicides, treatment of children, treatment of women, and treatment of animals. He argues each category shows measurable improvement over time when you look at long enough windows. The numbers are dense. There are charts on nearly every page that show violent death rates falling from somewhere around two hundred per one hundred thousand people in prehistoric times down to single digits in most developed nations today.

The Better Angels Of Our Nature Why Violence Has Declined Core Framework

Pinker attributes the decline to several overlapping historical forces. The first is what he calls the Leviathan, meaning the rise of centralized government and state monopoly on legitimate force. When you have a functioning police force and judicial system, private vengeance becomes less useful and eventually less common. I spent a weekend going through some of the primary sources Pinker cites for medieval homicide rates and the data does hold up. Rural communities with weak state presence showed murder rates roughly ten to twenty times higher than modern Western cities. That gap narrows but never closes completely once a state loses its enforcement capacity. The second force is commercialization. Trade creates interdependence between groups that would otherwise fight. This isn't a new idea and Pinker doesn't claim it is. What he adds is the quantitative backing showing trade volume correlates with reduced interstate conflict across multiple historical periods simultaneously. The third is feminization, meaning as societies become more prosperous and less dominated by immediate physical survival, cultural norms shift toward traits traditionally associated with feminine socialization. Cooperation matters more than domination. The fourth is cosmopolitanism, the gradual expansion of moral circles beyond your immediate tribe. The fifth is what Pinker calls the escalating civilizing process, basically habituation to nonviolence through repeated exposure to better outcomes. I ran into a specific edge case while trying to use this framework in a graduate seminar I was TAing for. A student argued that the decline narrative collapses when you include colonial violence against indigenous populations in the Americas and Australia. She was right about the data. Pinker covers this briefly but doesn't give it the weight it deserves in my opinion. The violent death rate in parts of the Americas did spike dramatically during colonization before eventually declining again. My workaround was straightforward. I pointed the student toward the demographic data from Douglas Hayton and other colonial historians who show the death tolls were catastrophically high, then had us calculate the violence rate per capita for that specific period rather than using global averages that dilute the numbers. It's worth noting the decline argument is strongest for Europe and its settler colonies and weakest for colonized regions where state collapse or replacement happened abruptly.

How to Actually Use This Book in Discussion

If you're going to engage with this material seriously, don't just read the popular summary articles. The actual book has over six hundred pages of data tables and footnotes. Start with chapters two through four for the core historical evidence, then move into the psychological mechanisms in chapters five and six. Chapter two alone contains a detailed breakdown of how medieval homicide surveys were conducted across European parish records. The methodology is rough by modern standards but consistent enough to draw conclusions. One thing beginners consistently miss is the difference between absolute and relative violence rates. The absolute number of war deaths has risen in some periods because populations grew. What matters is the rate per one hundred thousand people. A chart showing five thousand war deaths in 1945 looks terrible until you see the world population was roughly two billion and the rate was about 0.25 percent. Compare that to the Thirty Years' War where some German territories lost forty percent of their population. The relative framing changes the visual impact entirely. Another counter-intuitive point is that Pinker's data actually gets worse, not better, if you use certain alternative calculations. Some critics note that his choice of baseline periods and geographic scope can be questioned. The prehistory data relies heavily on archaeological evidence that is inherently incomplete. The twentieth century data benefits from better record-keeping, which makes modern violence look artificially lower than older violence. There's no clean fix for this problem. The best you can do is acknowledge it and check the sensitivity analyses Pinker provides in the appendix.

I also want to flag a limitation that the book doesn't emphasize enough. The decline thesis applies primarily to interpersonal and state-level violence within established political orders. It breaks down quickly when applied to structural violence like economic exploitation, systemic racism, or environmental degradation. Those forms of harm don't show the same downward trend and sometimes worsen. Pinker acknowledges this in passing but the book's title and promotional material make it sound like a universal claim about all human suffering. It isn't. For structural violence metrics, you need different sources entirely. Oxfam reports and the Human Development Index data are more useful there. If you want the full text, the book is available through Penguin Press and most major retailers. There's also a free PDF version floating around on academic file-sharing sites but I'd recommend supporting the author if you can. The audiobook narration is adequate though some of the statistical passages blur together when spoken aloud. Reading the data sections yourself lets you skip around and compare the charts to the text. That's the only way to get real value out of it. The most practical takeaway from this book is that violent crime in the United States dropped roughly fifty percent between 1991 and 2015 and most criminologists who study this don't fully agree on why. Pinker offers several hypotheses but the real answer is probably a combination of demographics, policing strategy changes, economic conditions, and the end of the crack epidemic. No single factor explains it. That's worth keeping in mind when anyone uses the decline thesis to argue that any particular policy caused progress. The causal chains are too tangled for simple attribution.

Read the book. Read the critics too. Roger Ekers wrote a good response pointing out that Pinker underweights climate change as a conflict driver. The data from the past decade suggests he may have been too optimistic about the trajectory. That doesn't invalidate the long-term trend but it does mean the story isn't finished yet.