Wednesday, January 30, 2008

Natural-Born Killers?

Lately I’ve been thinking about some of the darker facets of human nature, particularly the human capacity for killing each other. There’s enough going on around the world to justify sinking in to a reflective funk about this persistent and troubling behaviour, from rising gun crime on UK inner-city streets to Darfur Iraq and, more recently, Kenya. But I’ve had another reason for dwelling on the nature of the murderous mind: I have a feature article out this week in Nature on current trends in thinking about evolution, the brain, violence and murder.

As a species we’re built to compete for resources (money, sexual partners, status, power and so on), and from time to time the friction caused by everyone rubbing up against each other ignites an emotional explosion leading to murder. And we very likely have an evolved coalitional psychology that binds ‘us’ against ‘them’ in conflicts with outgroups (be they defined along national, religious or ethnic lines); the sparks created by abrasive groups pushing against each can all too easily set off an all-encompassing conflagration that threatens to burn down whole societies (you can fill in your own historical or contemporary examples here).

It’s not all doom and gloom though. Human history, as revealed by palaeontologists, archaeologists, anthropologists and criminologists, has always been plagued by violence and murderous mayhem, both at the inter-personal, one-on-one level, and also in terms of the death tolls exacted by tribal warfare. Yet on timescales from millennia to decades, things seem to be getting better. A much smaller percentage of the populations of modern democracies meet their end through murder of any kind than has been the case for most of the past 5,000 years (see my Nature article, and Steven Pinker’s essay in The New Republic, for more on this).

On the shorter timescale of decades things also seem to be on the up. For instance, the number of armed conflicts around the world, and the number of people dying in genocidal purges, is also on the decline. The Human Security Brief 2006, published by the Human Security Centre at the University of British Columbia, has documented a number of encouraging trends (though there is still, obviously, much progress to be made):
“Notwithstanding the escalating violence in Iraq and the widening war in Darfur, the new data indicate that from the beginning of 2002 to the end of 2005, the number of armed conflicts being waged around the world shrank 15% from 66 to 56. By far the greatest decline was in sub-Saharan Africa….The steep post-Cold War decline in genocides and other mass slaughters of civilians has continued. In 2005 there was just one ongoing genocide—in Darfur. In 1989 there were 10…The number of military coups and attempted coups fell from 10 in 2004 to just 3 in 2005, continuing an uneven decline from the 1963 high point of 25.” (These positive developments are tempered somewhat, the report notes, by increased international terrorism, and greater targeting of civilians in campaigns of political violence.)
In my Nature article, I try to draw out both the causes of violence and murder, and the reasons why they might on the decline over the long term. These are clearly enormous topics, and one could write a big book – or an entire bookshelf – trying to answer these questions, and still leave something important unsaid. In a 4-page feature, space constraints and the need for a coherent arc through the piece mean something has to give. One topic that I wasn’t able to go into in as much detail as I would’ve liked is the possibility that evolution has sculpted the mind to kill. So I’ll explore that idea in a bit more detail here.

Evolved Killers
Most evolutionary see murder as a by-product of our evolved minds, not as a behaviour that natural selection has sculpted humans to engage in. The business of getting through, and getting ahead in, life invariably brings people into conflict with one another. A colleague’s promotion may come at the expense of our own advancement; a competitor in the sexual market may monopolise the attention and affections of those we desire to have as our own; or a rival in the race for power, status and wealth may stand in the way of our goals. The risk that competition over material resources and reputation (which often serves to enhance the attainment of desirable resources) will escalate into murderous violence is particularly acute among men, who in common with many animal species have both more to gain and lose in the in the evolutionary game of successfully reproducing — and therefore greater incentives to place bigger, and riskier, bets at life’s table (essentially, the variance in reproductive output is greater among men than women, so that some men do really badly and others really well, whereas most women cluster around a similar average success) .

On the by-product account, the majority of murders happen when the normal brakes on aggression (fear of retaliation, empathy, and behavioural inhibition) are weak or temporarily overwhelmed by the momentum of an aggressive impulse: after a couple of beers, two hot-heads in a bar start trading insults over pool game, start fighting, take it to the parking lot, and one gets hit, falls, and smashes his head on a curb stone and dies.

And this failure to apply brakes on our aggression can also pose a threat to our supposedly nearest and dearest. Evolutionary psychologists Martin Daly and Margo Wilson of McMaster University, Ontario, Canada, have argued that men frequently use aggression and violence to control and coerce women, which perhaps surprisingly puts them on the same page as many feminists thoroughly antagonistic to evolutionary explanations of relations between the sexes. Occasionally, Daly and Wilson suggest, physical attempts by men to threaten, intimidate and punish their spouses (perhaps to make them think twice before leaving them for another man) occasionally result in “tragic slips” that leave the women dead.

Against this mainstream evolutionary account of homicide as a by-product of psychological mechanisms designed to facilitate completion and control of other people, David Buss, of the University of Texas at Austin, and Joshua Duntley, of the Richard Stockton College of New Jersey in Pomona, have developed a much more ‘adaptationist’ argument [1-3]. According to their controversial homicide adaptation theory, natural selection has fashioned psychological mechanisms to produce homicidal behaviour in a wide range of contexts, and for a correspondingly diverse array of reasons.

Buss and Duntley are critical of approaches to violence that treat it as a singular phenomenon explicable by a single cause — a tendency toward competiveness and risk-taking, or a lack of emotional control and behavioural inhibition, say. The by-product account, contend Buss and Duntley, is too vague, too broad-brush, to be up to the task of explaining the myriad manifestations of murder. Instead, they argue that homicide – the killing of one person by another — covers a range of diverse behaviours, including infanticide, step-child killing, intrasexual rivalry homicides, mate killing, and warfare killing; and their homicide adaptation theory suggests that there is a cold, evolutionary logic to these diverse forms of murder.

Men, for instance, can eliminate a potential sexual rival, acquire his resources, gain access to his mate, ascend a status hierarchy, and send out a message about their readiness to use violence in defence their family and property through murder. Women can use killing to protect themselves against a violently overbearing spouse, or a threatening stalker, as well as eliminating potential mate-poaching females. Natural selection could even favour the killing of biological children in situations when, put in the cold calculus of evolutionary logic, the benefits to investing in the new child are outweighed by the costs, or unlikely to be realised (for instance, if times are tight and other kids are already on the scene, then investing in the newborn might be detrimental to the survival prospects of existing offspring; and if the baby is born deformed or otherwise unlikely to thrive, then further investment— and it is painful to write this — might be the biological equivalent of throwing good money after bad). As Duntley says, “Homicide can be such a beneficial solution to adaptive problems in certain, specific contexts that it would be surprising if selection had not fashioned mechanisms to produce lethal aggression.”

Given the impact and enormity of murder, it is little wonder it looms so large in literature, art and films. It also plays on our minds a worrying amount too, psychologists have found. Building on research by psychologists Doug Kenrick and Virgil Sheets into homicidal fantasies, Duntley and Buss have found that upwards of 90% of men, and more than 80% of women, have had a vivid fantasy about killing someone.

Of course, most people do not act on theses fantasies. And the fact that people fantasise about something does not prove that the act in the fantasy is part of our evolved psychology; many people (mostly teenagers and young men) have fantasised about playing video games, but nobody’s suggesting that gaming is part of human nature. (However, a fascination with pornography is also not something directly selected for over evolutionary time, and there is no evolutionary benefit today for a man to spend time alone with a magazine, or at his computer, rather than getting out and meeting real women; nevertheless, pornography is able to exploit the desire to see naked, sexualised human forms, and the mind has no defence against being tricked and aroused by 2-D images and not just 3-D people in the real world. So it is possible that the predilection so many men have for computer games is explicable by some similar ‘misfiring’ of evolved psychology in a modern context).

At the same time, homicidal fantasies, or ideation, can provide a window onto the murderous mind, just as sexual fantasies illuminate the sexual mind. It is not the existence of these fantasies per se, however, that impresses Buss and Duntley. Rather, it is the pattern of these fantasies. Buss, one of the pioneers of evolutionary psychological studies of human sexuality, draws an analogy with sexual fantasies. Romantic and sexual liaisons are enjoyable, so it is little surprise that fantasies about them preoccupy the minds of both men and women. What’s more interesting is the differences in the types of sexual fantasies that men and women engage in. According to Buss, these map onto the different sexual psychologies that men and women have evolved through eons of sexual selection.

The same holds for homicidal ideation. Not only do the sexes differ in the types of homicidal fantasies they typically entertain; these map onto the different situations in which killing would have been beneficial to men and women over evolutionary time, according to Buss and Duntley. So men frequently fantasise about killing other men who have dissed them or otherwise challenged their social standing (as well as men trying to steal their girl, or men who personally threaten them); women are often moved to homicidal thoughts by abusive boyfriends (although they too report thoughts of killing mate-poachers, a threat to both men and women).

While most people do not act out their homicidal flights of fancy, sometimes they do. Murder does not always result from inflamed passions ignited in the heat of the moment. For example, many men who kill their wives (for cheating on them, say) plan and think through the act before committing it. For Buss and Duntley, this doesn’t make sense according to the standard line in evolutionary psychology that spousal murders typically result from tragic ‘slips’ in men’s attempts to use violence to coerce and control women. On top of this, argue Buss and Duntley, the patterns of homicidal ideation and fantasy correspond to the actual patterns of killings recorded in crime statistics.

For Buss and Duntley, this adds up to some compelling evidence in favour of homicide adaptation theory over by-product accounts. As I mention in the text of the Nature piece, other evolutionary psychologists remain to be convinced. Curiously, evolutionary psychologists are often portrayed as eager to propose evolved functions for all aspects of human behaviour, seeing adaptation and function everywhere they look. Yet the response to homicide adaptation theory shows this is anything but the case. Many think that there is no need to propose separate adaptations for homicide on top of adaptations for aggression as part of the system of competition – murder occurs as a by-product, not as a result of evolutionary design (D&B would counter that this account is woefully under-specified as an explanatory theory of the actual data on homicides).

Even those who remain sceptical about the claims of homicide adaptation theory don’t reject the possibilities of adaptations for murder in principle. As I quote Daly as saying, “I wouldn’t want hitch my wagon to the by-product argument, but I don’t think anyone, including Duntley and Buss, has figured out a good way to identify the hallmarks of homicidal adaptation”. There is also a debate about what counts as an adaptation for killing. Take the case of infanticide. In many species, including langur monkeys and lions, males frequently kill the infants in groups they have recently joined, which eliminates competitor’s offspring and brings females into oestrus so the invading male can have offspring of his own. Anthropological studies have shown that human mothers sometimes kill their own offspring in predictable situations — such as when the baby is deformed or unlikely to thrive, or when current circumstances are poor for raising a child.

It might be objected that in the case of human infanticide, the death of the child usually results from the mother simply walking away, rather than a lethal assault. Unlike in the case of lion infanticide (by males, in this case), the behaviour in human is not routine, and nor does it involve a specific infanticidal act, such as a deadly bite. And while there are likely to be psychological adaptations for assessing the current situation or the prospects for the kid, Martin Daly suggests that this doesn’t require anything that “deserves to be called an infanticidal adaptation – it is a de facto infanticidal act if you just walk away.” Duntley and Buss, for their part, think it’s a mistake to focus on whether death results from neglect or direct action. “We argue is that if there is evolved psychological design that reliably produces the outcome of a dead body, then that is design for homicide,” says Buss.

There are clearly theoretical, empirical and conceptual threads to the debate over whether humans have an evolved psychology to kill in certain contexts and situations. Some people accept the case of adaptations for infanticide; others for the coalitional psychology of war (see Nature article). Not many would go as far as Buss and Duntley in proposing such a wide range of homicidal adaptations. But that’s how science progresses: by people putting forward bold hypotheses, which the scientific community then discusses, evaluates and tests. If homicide adaptation theory can come up with better, more specific predictions, about who should be expected to murder and when — and it claims it does — then we will have to confront the unsettling possibility that we are, in part, evolved killers. On the flip side, this recognition may lead to a better understanding of what drives people to kill, and improved strategies for preventing the frequently senseless loss of life at murderous hands. And that can’t be a bad prospect.

1. Buss, D. M. The Murderer Next Door: Why The Mind Is Designed To Kill (Penguin, New York, 2005).

2. Duntley, J. D. Adaptations to dangers from humans. In The Handbook of Evolutionary Psychology (ed Buss, D. M.) p224–254 (Wiley, Hoboken, New Jersey, 2005).

3. Duntley, J. D. & Buss, D. M. The plausibility of adaptations for homicide. In The Innate Mind Volume 1: Structure and Contents (eds Carruthers, P., Laurence, S. & Stich, S.) p291–304 (Oxford Univ. Press, Oxford, 2005).