Steven Pinker’s “The Better Angels of our Nature, Why Violence has Declined”

Beware — this is a big book. There are nearly 700 pages, exclusive of the copious end notes, bibliography and index. The pages are large with small type. Pinker does not repeat himself and there is very little fluff. Occasionally I will slow down and pick apart the text and meaning to see how it fits into the thrust of the book. So far it has always fitted.

I wonder how Pinker got the will power to write this book. I lack the writing ability to write many of the books I read, I lack the knowledge to write most, but for a book like this it feels to me like it would been a lifetime task and I would lack the stamina.

I am morbidly fascinated by the number of ways that Pinker finds to slice the data. Yet he supports a number of interesting hypotheses thereby. This is indeed a data driven book.

Very good so far. I quibble about some game theory terminology: Games have a sum, which is the sum of the values to the two or more players. A positive sum game may leave one player in the hole, but if the game is voluntary, the transaction may be adjusted with a side payment to provide an incentive for all players. Plunder, as referred to on page 77, is not a voluntary game and it is not zero-sum; it is almost always a negative sum game. (Pinker makes the correct distinction in the rest of the book.) Pinker’s comment about justice pushing against such plunder is good. Robinhood lore suggests the possibility of positive-sum plunder, but that is unusual. This abuse of game theory terminology is widespread.

Pinker recounts the arc of levels of organizations, from genes to societies. Less evolved systems have smaller elements that fight among themselves. (the ‘selfish gene’) There are two general sorts of organization between people, into larger groups—voluntary and involuntary. Today’s (2012 Aug 25) Wall Street Journal had a column by Matt Ridley describing recent findings about continuing raging battles among the genes of insects.

The section on the history of American violence is fascinating. (Pages 98-128) Southern states today have higher violence than the rest of the U.S. and Europe. Gun laws are symptomatic of the problem but by no means the whole problem. Guns are used for robbery in the South no more than in the North, but they are used much more in revenge situations in the South. ‘Honor’ seems to play a rôle in the South where government justice solves many similar problems in the North.

The section beginning on page 212 (“Just as the uptick in civil warfare …”) is fascinating. There he organizes the history of violence from the perspective of ideologies, identifying a number of ideologies created by academia as well as from the classic ideologues. He mentions anti colonialism, democracy (as an ultimate cause of commonweal), and self-contained economies cut off from global trade. Confirmed anarchists will find little to cheer in Pinker’s diagnosis.

I was skeptical of ‘Peace Keeping Forces’ but Pinker quotes numbers and logic that changed my mind. Similar logic has worked to some degree as model constitutions are published, especially by Africans, that shame other Africans to conform. I don’t recall who first made this point to me, perhaps Collier.

I am surprised that Pinker makes a hard and fast distinction between bandits and governments. I am impressed with Mancur Olsen’s spectrum with war-lords at one end and democracies at the other and the natural trend towards the benign in particular cases. Today’s world populates that spectrum with few gaps. The UN must draw a line when it considers membership. On page 336, (“figured in Rummel’s definition of democide”) Pinker again makes a categorical distinction; without claiming that all governments are good, he categorically denies that warlords are governments. This categorical difference between Pinker and Olsen does not really blunt Pinker’s points however; it is merely an irritation. I think that Pinker would presume that even an evil government would reign in a local warlord for its own convenience and stability.

Pinker examines the expanding sphere of rights. He notes the problem of how large a radius to apply to animals—how many taxa. We feel we should be kind to animals because we are animals. We feel that it is natural for predators to kill their meal, but if we are animals is it not then OK to kill our prey? About the only entity he does not discuss is the AI and its rights.

Chapter 8 (Inner Demons) begins well but I have one quibble. In the third paragraph Pinker lists mental faculties that bear respectively on violence and forbearance. He lists reason in the latter category. Alas reason has a dominant rôle in modern military matters as well, and often leads to war.

On page 491 Pinker makes an interesting observation:

I think that an organism that stashes beliefs (propositions) can reject a painful one. That is almost the scenario when we are convinced by a reductio-ad-absurdum proof. In the case of cognitive dissonance it is an unconscious act to reject the hypothesis — close to confabulation. In reductio-ad-absurdum it is a conscious rational act, abetted by the unconscious. Pinker then describes a complicated experiment by Valdesolo and DeSteno seeming to show that the contradictory information was still in the head, but conditionally accessible.

On page 509 Pinker includes ideologies as sources of violence. I claim that ideologies are an example of “committed belief” as illustrated here in a harmless context. The ideologue may be aware that there are some mistaken people who do not share his ideology. I presume that there are self images or investments that would make it painful to disengage from the ideology.

Page 515, “Dominance”: Pinker mentions several phrases that suggest a dominance play. Among them is to “Draw a line in the sand.”. I agree that that phrase is usually a dominance play but it may as well be a suggestion that “here is where I think that my property ends and yours starts. Do you agree?”. The phrase suggests an outcome that is not necessarily a capitulation by either side. It is compatible with a negotiated settlement; the other phrases are not.

I have a libertarian leaning. This book is the first document that I have seen in very many years that has something good and principled to say about government. I think that libertarian rhetoric must address some of the issues raised here. It goes beyond the somewhat familiar observation that even corrupt governments usually prevent a great deal of violence. See note below about page 678.

Page 581: Quote:

I am slightly startled here that Pinker seems to advocate expansion beyond perhaps the concomitant changes that make this desirable. (Pinker lists some downsides to empathy on page 591.) I and most readers welcome the violence decline that Pinker portrays. Clannish behavior evolved for a reason however. Pinker noted that there is little benefit to humans of expanding the circle to include most animals, aside from the feel-good phenomenon. There is not much cost either.

I am tempted to butt in here and invoke Jane Jacobs and claim that this expansion has succeeded as well as it has because of markets which greatly expand the possibilities of win-win situations. Darwin’s evolution, as applied to DNA gave us hierarchical organizations (Jacob’s ‘guardians’) that were not smart enough to invent win-win interactions between larger groups. Jacob’s ‘merchants’, often working outside the hierarchies, showed the way to win-win. The merchants were due to evolved memes—evolved by somewhat Darwinian means. They came about via the brain which evolved Darwin’s way. Marx would not applaud.

I am surprised that Pinker has made little attempt to say why ‘kill the neighbors’ was good then and not now; why was it once adaptive? Generally works with an evolutionary perspective do that.


Recent Axelrod stuff (More Recent Tournaments: McCullough, 2008; Novak, May, & Sigmund, 1995; Ridley, 1997; Sigmund, 1997)
Pinker refers several times to the pattern of hyperbolic discounts, once by name. I suggest there is an easier explanation for this behavior. The person making the offer, $5 today vs. $10 next week, is a bird-in-hand. If the choice is to delay there is a probability that the bird will have flown, or the recipient may forget to collect or the recipient may be 1000 miles away when the appointed time arrives. All of these latter considerations are unconscious or indeed evolved instincts that are indeed adaptive, even today.

Pinker’s consideration of the place of reason (page 642) is good. I have an easier comeback to those who say that reason is useless. Darwin’s mechanism, the ultimate judge, has tested it and pronounced it fit. Sure there is confabulation where we find flakey ‘reasons’ to do what we wanted to do for emotional reasons; but that is just a reflection of the general value of logical reasons. We would not have evolved to value them if they were not usually germane.

On page 644 Pinker lists several ways that reason can lead to reduction of violence. I think he missed what may be perhaps the greatest. It generally takes reason, even combinatorial logic, to find win-win solutions to conflicts. Such logic assumes a theory of mind as grist; you must predict what situations your ‘opponent’ will want.

At the end of page 647 Pinker seems to confuse logic with the supposition that we should not harm others—somehow subsuming the Golden Rule. Alas that is an extra normative ethical axiom. Logic deals well with hierarchies and asymmetric situations. Given theories of mind and ‘meta logic’ which allows one to reason about what another might reason, win-win situations may be found and peace follow. Logic is necessary for such advances.

Page 656: I think that Pinker gives a sufficient explanation of the Flynn effect.

Pages 678-670: Here Pinker gives a simple quantitative model of the advantage of the state’s monopoly on power. I am very glad to see this as it is on the level of the prisoner’s dilemma including the new rôle of the state. It provides a quantitative model for optimum police power, including the nature of corruption of police power.


Erasmus, Bacon, Hobbes, Spinoza, Locke, Hume, Kant, Bentham, Jefferson, Madison, Mill