Actually, not angels at all

A review of Steven Pinker’s Better Angels of Our Nature: Why Violence Has Declined

Steven Pinker’s Better Angels of our Nature (Viking, 2011) might be the most significant work of political science for a trade audience in decades. In the book, Pinker addresses the modern decline in violence, and explains the causes of that decline. How could the explanation not involve politics? Indeed, Pinker’s account is deeply political, but also willfully ignorant of political science.

Meanwhile, the fact that Better Angels is political science has so far eluded reviewers, readers, and booksellers almost entirely. Better Angels was at one point ranked by as #3 in their ‘Early Civilization’ category, #5 in “Social History”, and #10 in ‘Psychology and Counseling’.1 Powell’s also calls Better Angels a psychology book, and none of the categories Barnes & Noble associate with the book include politics or political science. Peter Singer, reviewing for the New York Times, writes that “Pinker draws on recent research in history, psychology, cognitive science, economics and sociology” — but never mentions politics. This is a problem.

Pinker’s central question is, why has violence declined as an element of human affairs?  Answering this question requires Pinker to accomplish two tasks: first, to convince the reader that violence has declined, and second to offer a convincing explanation for that decline. This is a massive undertaking, and the result is a massive book. Folks short on time can visit earlier versions of the argument presented in Pinker’s TED talk and an article in The New Republic, both from March of 2007, both of which argue concisely for the decline of violence as a feature of human affairs.

In Better Angels, Pinker arrays a much broader range of evidence in favor of his proposition, drawing from history, anthropological, pop-culture,2 but especially heavily from political science. At least two chapters are dominated by the work of political scientists — the reader loses count of how many times Pinker begins a sentence with “As the political scientist [name] writes”, or some variation thereof. How Peter Singer missed this is anyone’s guess, but those pressed into Pinker’s service include Bruce Russett, James Rosenau, Joshua Goldstein, James Fearon, David Laitin, Nils Petter Gleditsch, John Oneal, and Ted Gurr, among many others. What Pinker has done, at least from page 189 to page 377, is synthesize a significant portion of the evidence amassed by political science on war, civil war, genocide, and other state-related forms of violence to argue all indicators demonstrate a decline in violence.

Having established to his satisfaction that the puzzle exists, Pinker then turns to its solution. He argues that a number of related threads in human development have driven the decline in violence, among them the Enlightenment, literacy, and education. Yet, Pinker writes towards the end of the book: “A state that uses a monopoly on force to protect its citizens from one another may be the most consistent violence-reducer that we have encountered in this book”(p. 680) and we can infer he believes this true outside the book, as well. Pinker bases this argument on a wafer-thin reading of Hobbes, and refers again and again to the beneficial effects of ‘Leviathan’ in his book. Recall that Leviathan refers to an indomitable Biblical monster; Hobbes viewed Leviathan as an absolutist, unlimited state. Casual readers of Pinker might instead form the impression that Hobbes had in mind present-day Finland, but nothing about Hobbes’s understanding of the beast implies the eventual elimination of torture, capital punishment, or conscription — all of which Pinker points to as evidence of the overall decline in violence. Part of the problem is that Pinker has a reductionist view of the state; he writes that “Real-life Leviathans are essentially human beings, with all the greed and foolishness we should expect of a specimen of homo sapiens” (160). We had this debate in political science a while back, and decided states are more than just their leaders. Pinker missed that conversation, or ignored it.

Another problem for Pinker is that Hobbes’s vision of human nature is arguably false. The best recent academic book on violence is sociologist Randall Collins’s Violence: A micro-sociological Theory (Princeton, 2008), in which the author argues that “The Hobbesian image of humans, judging from the most common evidence, is empirically wrong” (11). Collins’s argument is simple: humans find violence difficult to perform. Much as we might wish an enemy injured or dead, we find the actual doing tough work. Humans’ aversion to violence is marked by tension and fear “that comes not from concern for bodily pain; this is something that people endure surprisingly easily” (8). Instead, the tension and fear is caused by our emotional ‘entrainment’ with one another, and desire to avoid confrontation, so that people will often refuse to act violently to resist violence  to themselves. Collins supports his case with careful and voluminous use of empirical evidence — including an extensive collection of photographs and incident reports — and demonstrates that violence is much rarer and more dependent on social context than is popularly believed. This has obvious parallels with Pinker’s work, and speaks to a specific premise of his argument, yet Pinker ignores it. That is, Pinker does cite Collins — more than once — but he does not discuss Collins’s refutation of Hobbes.

Apart from Hobbes, Pinker makes almost no use of political science theory to shape his arguments. He may think highly of our evidence — enough to borrow it, at least — but he holds the theoretical foundations of our discipline in low regard, especially those of International Relations. He points out that ‘many scholars’ in IR hold an “influential theory tendentiously called ‘realism’,” by which “the absence of a world government consigns nations to a permanent state of Hobbesian anarchy” (291). Here his understanding of the state gets him in trouble, because he argues that realism requires leaders-as-states to act “like psychopaths and consider only the national self-interest”; however, he writes, “humans are also moral animals” insofar as they are also guided “by moral intuitions supported by emotions, norms, and taboos”; thus, “it is neither sentimental nor unscientific to imagine that particular historical moments engage the moral and cognitive faculties of leaders and their coalitions in a combination that inclines them toward peaceful coexistence” (291). This is his demolition of IR’s ‘realism’ — which he later puts in quotes to avoid confusion with any more realistic realism (674) — yet even few non-realists in IR would attribute peace to the ‘moral intuitions’ of world leaders. Pinker has drawn some criticism for sneering at people he disagrees with in Better Angels, but the problem for political science is rather more silence. Pinker barely mentions our theoretical perspectives — except to snipe, very briefly.

Instead Pinker draws on the work of Norbert Elias, a sociologist/historian whose book The Civilizing Process was first published in 1939, and is — granted — every bit as relevant to political science as Max Weber’s work. Elias’s argument is subtle but profound: European elites, beginning around the 12th century, began developing elaborate codes of etiquette designed to structure relations between monarchs and nobles, so that the monarch could better control and define those relations. These rules  taught people to restrain their impulses and consider the interests of others, and they were subsequently absorbed by the nascent middle class to became the standards of normal behavior among ‘civilized’ persons. The ability to restrain yourself from farting in public turns out to be closely related to the ability to restrain yourself from punching a public-farting person in the face. The upshot of this process is that the words ‘impolite’ and ‘impolitic’ may have somewhat different connotations, but in fact mean roughly the same thing. What we consider polite or not is in fact a legacy of the rules which began to structure society in the late medieval era, and these rules established a very specific political system. Of course, Pinker does not describe it as such.

Pinker prefers to talk vaguely of ‘social inputs’ and ‘social organizations’. For example, in a modest concession for Pinker, who became famous for his arguments about the evolutionary origins of language, he avoids similar arguments about the decline of violence: “Since it is indisputable that cultural and social inputs can adjust the settings of our better angels (such as self-control and empathy), and thereby control our violent inclinations, we have the means to explain all the declines of violence without recent biological evolution” (621-622). Speaking of market economics as a domain of human activity, Pinker argues “It really should be lumped with other examples of formal social organizations that have been honed over the centuries as a good way for millions of people to manage their affairs in a technologically advanced society” (628). What are these mysterious social organizations, these social inputs reorienting human nature? We might name them, simply, ‘politics’ — in which case Pinker’s argument reduces to, ‘violence declined because humans got better at politics’.

The last third of the book is given over mostly to discussions of psychology of violence, and how that psychology is changing. Psychology being Pinker’s expertise, it is not surprising that he spends a lot of time talking about the changes and less about the changers, so let’s be clear: Pinker thinks “social inputs” and “social organizations” can change the way we think. Who is doing the educating in the Enlightened West? Governments, generally. Who is running the nutrition and vaccination programs to ensure healthy brain development? Governments, generally. Who is driving the civilizing process? Governments, generally.

To put Pinker’s views more succinctly: politics shapes human minds. It is not just that Leviathan makes the rational calculus of violence less rewarding, but that government can make violence literally unthinkable.

Pinker’s argument is a pretty staggering claim for the scope of political science. Minds are blown. Those of us calling ourselves political scientists should be quite glad for it: Pinker is opening vast grants of inquiry to our discipline, where previously we might have felt unwelcome. So why is this not obviously a political argument, much less a political science argument? Part of it is that Pinker dresses it up like a psychology book, and Pinker is of course a psychologist.

The bigger problem is that political science has never quite grasped violence as its subject of inquiry. Consider the history of the discipline: it emerged in the 20th century, well into the civilizing process Pinker describes. At that point, within-state violence was a minor concern, but inter-state violence was still a growing problem. Our lexicon reflects this: when political scientists talk about ‘war’, they usually mean inter-state violence. If they mean intra-state violence, they will modify to ‘civil war’. Political science is a late entry in the Enlightenment, and so our perspective on the problem of violence is limited. We see violence — ‘war’ — as the product of conflicting ‘social organizations’, and generally ignore the extent to which those organizations are themselves a reaction to violence. But politics is fundamentally a response — perhaps an instinctive response — to violence in human society. Consider that Collins defines violence as “a set of pathways around confrontational tension and fear” (198). Violence is not the only set of such pathways; so what do we call pathways around confrontational tension and fear which are not unlimited violence? Again, ‘politics’. Granted, politics can lead to violence; but when a rule says, ‘if x, then violence’, the likely purpose of the rule is to avoid violence. And Pinker makes clear that people have been developing rules for the management of violence since the dawn of human kind. Eventually those rules merged and meshed to form the state, a fairly sophisticated social organization for the management of violence.

This brings us to the second problem for political science: ‘policy relevance’. A lot of people in the discipline view this as the gold standard of political science, especially International Relations. What this means is that our work orients towards the interests of policymakers, i.e. the state itself, and not regular people. Even when we write for trade audiences, our books are policy-oriented. Clash of Civilizations, the one political science book our non-political scientist friends have all read, is very much policy-relevant (also very wrong-headed).3 Point being, political science is most interested in what interests policymakers, i.e. the state. This brings us back to the least developed part of Pinker’s argument, that the civilizing process affected the state because it changed state leader’s ‘moral intuitions’. This holds no water whatsoever; the state in most instances throughout the history of the form has resisted such limitations. For example, the U.S. government at this very moment is debating a bill that would allow the state to imprison with no limits its own citizens on the mere accusation of terrorist-related behavior. Is this an advance in our leaders’ moral intuition? Hardly.

A better explanation for the historical decrease in state violence is that the civilizing process, once effected upon the people, was reflected in them against the state. To quote the famous Admiral; “it’s a trap!” Once people became inculcated into the mindset that violence was wrong for them, they began to think it was wrong also for the state. This is especially true in democracies, where the people are nominally (more or much, much less) the state, but true to some degree for any society affected by the civilizing process. The state as an organization has no interest in its own limitation, but has instead become trapped by the civilization of its constituents.

What this means for political scientists is that ‘policy-relevance’ will often put us opposed to limitations on the state. Consider the literature that developed in political science over the last decade, which argues, more or less, the government can or should do pretty much anything it deems necessary to kill terrorists. This was an argument for more violence, not less — against limitations on the state. An argument which said ‘the state can’t do everything it wants’ would have been, by definition, policy-irrelevant, and few political scientists made that argument. More generally, political science finds it easy to ignore the problem of violence because such ignorance is in the interests of the state. A thorough understanding of violence might put political scientists in the uncomfortable position of having to criticize the state’s interest in an unlimited capacity for violence.

While an understanding of violence might not be policy-relevant, it is politics-relevant; violence is the fundamental problem of politics. And we — political scientists — don’t have that understanding, not yet. So we don’t talk about violence, don’t think about it, and don’t write about it (at least, most of us). We don’t have a literature, nor a vocabulary, much less a popularist version of the same, so no layperson confronted with the problem of human violence immediately thinks, “Aha – political science!” Which is why Steven Pinker was able to publish an 800 page treatise on political science, without anybody recognizing it as such.

The bad news is that it took a Steven Pinker to frame the puzzle; the good news is that he did not solve it. That is a job — maybe the job — for political science.



1 Amazon also has a ‘Social Sciences’ category, in which Better Angels is ranked #29 — just ahead of Glenn Beck’s Being George Washington at #30. In Amazon’s ‘Politics’ category, four different editions of Beck’s Being George Washington occupy positions in the top twenty, including #1 and #3. Better Angels is not counted as ‘Politics’, but presumably would be #1 if it were. There is no political science category. Granted this reflects more on the bookseller than the discipline, but it is a problem the discipline should wish to correct. All of this on 4 December 2011.

2 Pinker spends a couple of paragraphs talking about pop music’s reflection of changing social norms in the 1960s, eg the Rolling Stones’ “Street Fighting Man”. For Pinker, the ’60s were comparatively uncivilized, while the ’80s and ’90s saw a return to civilized norms. Given Pinker’s interest in pop culture, why then does he not write about the popularity of hardcore rap  in the 1990s — just when we were supposedly becoming more civilized? One of Pinker’s more disconcerting tics is his tendency to elide contrary evidence in this manner.

3 Criticism of Huntington’s work is readily googleable. For counter-example, Putnam’s Bowling Alone is also a political science book popular with trade audiences, and perhaps less chained to policy relevance.

Leave a Comment

Your email address will not be published. Required fields are marked *