Killers and victims
Martin Daly and Margo Wilson
Martin Daly and Margo Wilson worked for much of their careers with
monkeys and mice, but nowadays they are mainly known for their research on
homicide. Among their many publications is Homicide (1988). The following
interview took place at the 1996 conference of the 'Human Behavior and
Evolution Society' in Evanston, Illinois, USA.
Why did you investigate murder?
Margo Wilson:
We had different ideas and hypotheses about interpersonal conflicts,
for instance marital conflict, sexual conflict, parent-offspring conflict, and
conflicts in a competitive setting too. We were trained in animal-behavior, and
we have maybe an innate distrust of what people say.
About sexual conflict, a long time ago I thought, well, maybe we can go
to shelters of battered women and talk to the women, but then we thought: No,
who is using these facilities? It is a very biased sample. If you know before
what way the bias is going, then you can deal with it, but we did not know.
One day we were moving from California to McMasters in Canada and came
through Detroit, and we thought, my God, the murder-capital of America, why
don't we use homicide? All cases get reported and investigated, so there is no
bias in the sample.
We contacted the chief of police in Detroit; he had a Ph.D. in sociology
and was very interested in research, and he said "Yes, come!" So we
went there, the first time in our life we were in a police-station, looked at
the information he had, and realized we could do a ton of things.
In your book you give statistical data on victim-killer relationships, for instance how often men kill other men, how often parents kill offspring and the other way around, how often males kill females, females males, etc. You write in your book that you were amazed to discover that no one had ever compared an observed distribution in victim-killer relationships to what was expected in the light of any sort of theory of interpersonal conflict. Are you saying that, except for your own theory, there actually is no theory of homicide?
Martin Daly:
Homicide research has been dominated by sociologists who are interested
in what they call 'macro-social determinants of variation in gross homicide
rates'. Another group of homicide researchers has been psychiatrists interested
in 'abnormal perpetrators'. But there has been very little attention to the
demography and epidemiology of homicide, that is, to questions like:
"Within the married population, what are predictors of who is likely to
kill whom?" I think it fair to say there has been essentially nothing,
except for statements like: "People in bad circumstances might be more
likely to kill their children".
You show with a long list of data from all kinds of places and times, that males kill other males at least fifty times more often than females kill other females. I once heard the explanation that as one male can inseminate many females, it is good for the survival of the species that males do all the dangerous, risk-taking jobs. What is your opinion about this explanation?
Martin Daly:
Survival-of-the-species, or good-for-the-species kinds of arguments are
unpopular among evolutionists, for the reason that it doesn't make sense that
natural selection should have favoured attributes which serve the interests of
species. Natural selection is overwhelmingly a matter of differential
reproductive success of individuals within a species. Doing the right thing for
your species is likely to lose out in competition with doing the right thing
for the proliferation of related individuals likely to carry the heritable
substrates of the same traits, whatever they are. Nepotistic attributes are
likely to spread in a population to be evolutionary stable, whereas those
things that are good for the species nevertheless cannot evolve.
So why are males more violent than females?
Margo Wilson:
Females don't have the same potential for reproductive success as males
do. A woman can only have a child once every few years until menopause, so, in
contrast to males, there is a finite possibility. Compared to a man there is no
payoff for being dangerously competitive.
It's a zero-sum game among males, and competition among males for access
to females, a form of sexual selection, has designed the male mind to be both
more confrontational and dangerously risk taking. Some men will have a lot of
access to women, others will have very little, and this may have to do with
attributes like how many resources they have, how powerful they are, and how good
they are in keeping other men away from their females.
Martin Daly:
The general notion that seems to apply across the animal kingdom is that high variance in outcomes and high rewards for being a winner, and high probability for being a total loser, in combination select for higher risk competition tactics. Perhaps it is worth saying in trying to explain this argument, that the woman who stays out of trouble and keeps her nose clean, is likely to have successful pregnancies; a male who merely keeps out of trouble is likely to die celibate, and dying celibate is no better than dying young trying harder from a natural selection perspective.
You show that co-offenders are six times more likely to be blood-relatives than victim-killers. Why is this something you would expect?
Martin Daly:
The prevailing criminological model in sofar as there was an explicit
one for who is likely to kill whom, is an opportunity model. Obviously you are
not going to get into conflict with people you don't interact with. The more
often and intensely you interact with them, the more opportunity for conflict
to arise. We wanted to say, yes, obviously there is a lot of truth to that, but
over and above that, given a certain level of opportunity or interaction, there
is differential likelihood of conflictual versus cooperative motives to arise
in relation to relatedness. Your kin are, to use the jargon of evolutionary
theory, the vehicles of your fitness; -your offspring the most obviously, but
collateral kin as well. So we thought it would be interesting to investigate
the victim-killer versus co-killer relationship. We don't know how much people
interact with relatives and non-relatives. But whatever that distribution is,
if the opportunity to form cooperative alliances in dangerous endeavours is
distributed according to interaction frequency, the opportunity to come into
conflict is similarly distributed. By a pure opportunity-model these two
things ought to be similar. But it doesn't work like that. People tend to collaborate
with their relatives and they tend to do damage to nonrelatives.
Step-children seem to be a special case. You show that step-children are a hundred times more often fatally abused than genetic children.
Margo Wilson:
Parental investment is a valuable resource that you could allocate to
different activities. It takes a lot of time and effort and self-sacrifice, so
you would expect that selection would have shaped both mothers and fathers in a
biparental species to allocate that investment to own offspring, because
otherwise you would be contributing to rivals. Children that are not yours, or
not your kin, you would expect the psychology to be such that you would discriminate
against them, and you would be more reluctant to invest in them. The emotional
experience would be that perhaps you love them less.
A year or so ago we published a paper on qualitative characteristics
that are different for when genetic parents kill their children versus
stepparents. It looks like the emotional context where the genetic parent kills
is in sort of sorrow, not in anger, while in stepparents it is in anger. The
stepparent is more likely to assault to death, beat it to death, while in the
genetic parent it is smothering or maybe it is shooting.
Also, the genetic parent is more likely to kill self in the same
episode, and the stepparent doesn't. So there are cues from the context of the
homicide that sort of betray us that there is a very different background that
is causal to this outcome.
But killing is a rare outcome, the tip of the iceberg.
Martin Daly:
You may have heard of the sexually selected infanticide phenomenon, the
phenomenon of a new male who takes over a troop of langurs or pride of lions
and more or less routinely kills off his predecessor's offspring. This
terminates, from his perspective, the female's wasted investment in his
predecessor's young and gets her on to breeding with him sooner. At a very
distal level there is an analogy, but we do like to stress in this context that
it seems pretty clear that there is not anything like a sexually selected
infanticide adaptation in the human animal, certainly not like in a lion or a
langur. Humans don't routinely do that anywhere. And for everyone who kills or
dramatically mistreats a stepchild, there are many who make some degree of
parental-like investment to the stepchild, and actually do it a favour.
Couldn't the behaviour of the stepchild be responsible for the aggression of the stepparent?
Margo Wilson:
That is something you would expect with older children, who are talking,
walking and are annoying to the parent. But I know of nothing that would give
you the expectation that a small infant treats a stepparent differently than a
genetic parent. And we found that the biggest risk is for the youngest, under
three years.
In Holland when someone treats you badly, they sometimes call this a 'stiefmoederlijke behandeling'. That is, you are treated as badly as a stepmother supposedly treats you. Why a stepmother and not a stepfather?
Martin Daly:
We looked at a sort of cross-cultural compendium of folklore, and a
'stiefmoederlijke behandeling' is apparently worldwide. These stories are
everywhere, and stepmother-stories are apparently much more prevalent than
stepfather ones. When you think about stories, folktales and those sort of
things, you have to ask: Why do these things exist? They have to fulfil the
social purposes of the people who tell them, and they have to be interesting
to the people who hear them. My take on cruel stepmother stories is that the
people who tell them are genetic mothers; they tell children how awful
stepmothers are, and the secondary message is: The worst thing that could
possibly happen to you is for me to disappear and your father replace me.
But why is the father not telling the same thing about stepfathers?
Martin Daly:
He probably doesn't tell anything to the kids, but if he does, he is
more likely to tell cruel stepfather stories. But whether stepmothers are more
risky to kids, we don't have actual information about that. Our best estimate
is that the risk is similar.
Nowadays very young children almost never live with stepmothers. The
'Snowwhite,' 'Cinderella' and cruel-stepmother stories probably originate
from the times when stepmotherhood was not all that rare, because of high
mortality of young mothers.
Why do husbands kill their wives, and why do wives kill their husbands?
Margo Wilson:
One situation that is associated with men using violence against their
wives is that of sexual jealousy, that is, the man thinks or fears his woman is
having an affair with another man. Another situation that may turn dangerous
for the woman is when he thinks she intends to terminate the relationship. In
both situations the man is at risk of losing control of his wife's reproductive
capacity and is hence losing ground in the reproductive competition between
men.
Men use violence and threats against their wives in an attempt to regain
this control, and an extreme outcome is the man who kills his wife.
Martin Daly:
The motivations for women who kill their men appear to be completely
different. The dominant theme here is a struggle to resist coercion. Most
commonly, women kill in self-defence against husbands who are abusive against
them, their children, or both. Regardless of which spouse ends up dead, the
husband is usually the instigator of violence.
When people think of biologists, they often imagine someone putting rings on the legs of birds, or looking through a microscope. Your proposal that the social sciences should consider themselves branches of biology, probably sounds surprising to these people.
Margo Wilson:
We mean by the word biology what the dictionary says, which is the study
of life. And since people in the social sciences say they are also studying
'life', then, of course, we are all in the same business. We have different
specialisations which have to do with what aspects of life, or the kinds of
perspectives we bring to bear, or the kinds of explanations we are interested
in. People, including biologists, often misuse the word biology. When they
should be talking about physiology or endocrinology or genetics, they should
say that. Biologists sometimes say about a cultural phenomenon: "This is
not my subject, because I am a biologist." But cultural phenomena are by
definition biological phenomena.