15 May 2012

Facts and research
Child abuse which the child protection authorities do not want to know about - 2:
Violence against step-children compared to genetic children – Daly & Wilson's research
By Marianne Haslev Skånland, professor emeritus

An edition of this article in Norwegian has been published on various websites, first time in April of 2005.

Marianne Haslev Skånland has worked as a professor of linguistics at the University of Bergen and is now retired. Her major interests are analysis and criticism of science, both generally and in the areas of linguistics, psychology and child protection. She has been a member of the scientific board of the Foundation for Forensic Psychology in Stockholm, Sweden. She is engaged in social issues relating to human rights and health, and especially interested in the question of the scientific basis for attitudes to psychology and social life shown by social and judicial authorities.



1. About the research scholars Martin Daly and Margo Wilson
2. Central theory and hypothesis
3. The studies
4. Possible confounds and sources of error
5. Criticism, and the response to it
6. Conclusions
7. The explanation
8. Daly and Wilson versus representatives of social disciplines
9. Literature

1. About the research scholars Martin Daly and Margo Wilson


The Canadian psychologists Martin Daly and Margo Wilson are research scientists in evolutionary psychology (psychology based on evolutionary biology). They are in the forefront in modern research, perhaps especially in ethology and sociobiology, and in the history of biology and psychology. Parts of their research are of great importance in throwing light on why and how Norwegian child protection and psychology related to child protection go astray.

Daly and Wilson have been interested in researching to what extent evolutionary biology can predict types of behaviour, and whether these predictions are borne out by facts. Martin Daly is also a biologist. Originally they worked with animal studies but later on human behaviour has been an important research field for them. Their book
Sex, Evolution, and Behaviour (2nd edition 1983) is very informative. For survey articles giving a historical account of the research field and a clear picture of their views on many general and theoretical questions, see Daly & Wilson 1994/95 pp 1269-1273, and 1999 pp 509-516. A discussion and critique in the last-mentioned periodical is Smith, Mulder & Hill (2000).

In a series of publications from 1980 on, Daly and Wilson have taken up the occurrence of different types of violence in an evolutionary perspective. They have, for example, investigated homicides, locally and nationally in the United States and Canada, in relation to two variables: differences of income and average income, and have arrived at interesting results, cf Daly, Wilson og Vasdev (2001).

They have done several studies around violence within the family. Violence against partners is studied e.g in Daly and Wilson 1988, 1992, and 1994a. Here, the killing of partners by each of the two sexes is studied in relation to other factors possibly correlated with the sex-difference. The last of the articles gives a comprehensive treatment of theoretical questions concerning violence seen as behaviour. See also an editorial article in
Psychology Today 1993 (a very short, popular type summary of some main points).

Particularly valuable research has been carried out by Daly and Wilson concerning the occurrence of family violence towards children. In this area they have brought to light important facts largely overlooked or avoided by other researchers. These studies are the topic of my article here. For brevity I abbreviate Martin Daly and Margo Wilson's names as "D&W" below.

2. Central theory and hypothesis


On the side of theory, the starting point is this: A consequence of current hypotheses about how evolution functions is that it must be assumed to have led to the development of feelings and reactions causing individuals – human as well as animal – to prefer, and to contribute most help and care to, other individuals who will help the genes of the contributing individual survive and pass on: first and foremost one's own offspring, other close relatives who do not compete strongly with one's own offspring, and sexual partners with whom one has or may come to have children in common. If theory and hopotheses are right, these predictions about feelings and care should be manifest in practice, and if they are, it supports the hypotheses and the theory behind.

The same theoretical starting point regarding the consequences of evolution also leads us to expect that when conflicts arise, the individual will have a tendency to show greater degrees of indifference, enmity, aggression and violence towards individuals who do not contribute to proliferation of one's own genes, than towards individuals who do. This must be assumed to be valid regarding both partners and offspring, if the theory is right (1994/95 pp 1273-1274, p 1278; 1998 pp 11-17).

On the side of facts, such discrimination in favour of own offspring is well-known among several animal species. The clearest cases of all are those where adult individuals kill the offspring of others to further the opportunities to breed their own offspring (1998 pp 8-10; 1994/95 p 1279).

Coming to human behaviour, families with a step-parent stood out as a relevant group to investigate in order to find out whether similar conditons obtain to those found among animals. However, D&W found that certainly, a rich store of "step-parent stories" of an anecdotal type exists around the world, and there were also studies which could throw light on the question in some measure. But these studies had rarely made use of reliable statistics; no scholars seemed to have been sufficiently interested in the question to seek to clarify it through research on a general basis (1988 p 520; 1994/95 p 1278; 1991 p 420, p 421; 1996 pp 78, 80; 1998 pp 20-23; 1994b p 207: ‘The first reported evidence of this differential risk was Wilson, Daly, and Weghorst's (1980) demonstration that U.S. stepchildren were overrepresented as reported child abuse victims.’; pp 207-208: ‘The possibility of excess risk in stepfamilies had been overlooked by researchers lacking an evolutionary perspective’).

I shall return to a possible reason why the question has not been studied earlier (section 8 below).

Since nothing much had been done before, D&W have carried out such studies themselves, especially concentrated around ascertaining the extent of violence within families/households towards step-children compared to biological children - or, as they prefer to call them: "genetic children". Their studies were based on different surveys of reported and registered child abuse in the USA, Canada and Great Britain. The reliability of the surveys and possible sources of error were assessed.

3. The studies


When we want to find out how frequent, and therefore how likely, abuse of step-children is in relation to what it is like towards genetic children, we understand that it is not enough to count up how many children have been abused by a step-parent versus how many by a genetic parent (or, from a slightly different angle: count how many step-parents carry out abuse compared to how many genetic parents do the same). Absolute numbers must be seen in relation to how many step-parents there are and how many children live in households with a step-parent, compared to how many live in households with their two genetic parents (or in households with a single genetic parent). This is easy to understand, but still sometimes misunderstood or forgotten by many who do not deal in statistics on a daily basis. Especially in superficial newspaper articles we therefore see accounts claiming that by far most child abuse is carried out by the child's own parents, but without the journalist having investigated what the relative numbers are.

Such presentations of figures are misleading. It may be that children abused by their own genetic parents are a higher absolute number than children abused by step-parents. But if step-families are far fewer, the risk of a step-child being abused can still be higher than that of a child living with both of its own genetic parents.

D&W's study from Canada for the period 1974-1990 shows an example: 91% of children under the age of 5 were calculated to be living in households with their genetic father. In this period of time 178 children under 5 years were killed by their genetic fathers, amounting to 6.3 killed per million. In the same period 0.6% of children under 5 years of age were calculated to have lived in households with step-fathers. 67 such children were killed by their step-fathers, coming out as 392 killed per million (1994b p 210). There was, in other words, a 60 times higher risk for a child living in a step-family to be killed by its step-father than for a child living with its genetic father to be killed by the father.

Certain information about the frequency of different household types is therefore important if such studies are to give results. D&W have been painstaking in obtaining as reliable figures as possible and in studying and assessing the figures (1994b p 209, pp 212-213; 1985 p 197; 1996 p 78; 1998 p 27). When official statistics or other sufficiently reliable sources did not give the information on household composition which D&W wanted to be able to calculate such differences in risk – i.e information about what proportion of households in different social groups had a non-genetic care-giver (step-parent) – they have ascertained this in a separate investigation (1985 pp 198-199, pp 200-201; 1998 pp 29-30).


The various studies are described and referred to in several articles; in addition they are summarised in
The Truth About Cinderella. A Darwinian View of Parental Love from 1998.

•  The first study from USA (1998 pp 26-29; 1985 p 197) took its data from AHA (American Humane Association), which functions as official registrar of cases of child abuse in most of USA. The number of registered cases was somewhat above 87,000. The basis for registration of step-families in the population had been such that the number of step-families must be supposed to have been set too high; therefore the relative number of abuse cases in step-families would appear lower than it really was. Even so the result was: The probability for a child under 3 years of age, living with one genetic parent and one step-parent, to be registered as abused in an investigated and confirmed report in AHA's archives, was 7 times as high as the probability for a child living with its two genetic parents.

To locate possible sources of error due to reporting or to other factors than the step-relationship, the results were subjected to further analysis:

To guard against possible bias in reporting, the least serious cases were eliminated, since this is where errors in reporting is most likely. The result was that the over-representation of step-families increased. In other words, the over-representation of step-families involved in child abuse was not due to possible prejudice of people reporting or within the police or the health services. When all cases were cut out except the 279 in which the child had died from the abuse, the result was this: It was 100 times more likely that a step-child would be killed or subject to abuse leading to death, than it was that a child in a household with its two genetic parents would have the same fate (1998 p 28).

The elevated propability decreased with the age of the child. Thus, child abuse in step-families cannot be explained as conflicts between step-parents and defiant or provocative teenagers (1998 p 30).

The only potential confound the material made it possible to test for, was poverty. It was known that there is a higher frequency of child abuse among the poor. But going into the basic data, it was found that the over-representation of child abuse in step-families came in addition and was independent of poverty.

The results of the study were published in 1980 and 1981.

•  D&W's next study of child abuse was in Canada. Here they wanted to do a smaller, better controlled study in a local community and chose the Hamilton-Wentworth area in Ontario (1985 pp 198-205; 1998 pp 29-31). The basic figures were from cases serious enough to have been reported to the child abuse register in the province.

For school-children the risk of being abused turned out to be 40 times higher for children in step-families than in families with two genetic parents. The elevated probability decreased with higher age of the child; for step-children in their teens it was 10 times as high as for genetic children.

Sex abuse was one of the forms of abuse covered by the study. Such abuse was hardly present for the youngest children, but increased with increasing age of the child. In other words, abuse of step-children was not predominantly of a sexual kind (1985 pp 207).

•  They then studied child killings in all of Canada, based on state archives covering all killings known to the police and with a known perpetrator (1994b pp 209-212; 1996 p 78; 1998 p 32).

The over-representation of killings committed by step-parents was higher than the over-representation of non-lethal abuse committed by step-parents. The probability for a step-parent to kill a step-child younger than 2 years was 70 times higher than for a genetic parent; for step-children in their teens it was 15 times higher.

•  A study in Britain gave similar results (1994b pp 212-215). National registration of abuse cases came up with 32% of abused children coming from step-families, while calculation based on random sampling showed only 3% of children with the same age profile to be living in step-families. In Britain there was some over-representation of step-families in the poorer social groups, but when economic differences were controlled for, there was still en over-representation of abuse of step-children. (At this point D&W 1998 must probably contain a faulty number on pp 32-33. The elevated risk pertaining to the step-relationship is said to be 19 times as high as that for genetic children. But the quoted numbers, with 32% abused as against 3% step-children in the population, gives 15 times as high risk for step-children? If poverty is responsible for part of this increased risk, the part due to the step-relationship must be somewhat lower than the 15, not higher.)

•  Yet another study, done by forensic psychiatrist P.D. Scott, had figures which could be converted into probabilities: Regarding infants of average age 15 months killed by beatings given in anger by fathers or by step-fathers, it was 150 times more likely that the perpetrator was a step-father (1998 p 33).

A study done in New Zealand is relevant (1985 p 198), and D&W have also examined studies from other countries: New South Wales in Australia, Finland, Korea, Hong Kong, Nigeria, Japan and Trinidad. They have furthermore gone through relevant literature from anthropology and folklore studies. All have information or results that point in the same direction (1998 p 23, pp 32-36).


Some results are common to several of the studies:

Abuse/violence of all degrees against children is many times as frequent in step-families as it is in families where the children live with both of their genetic parents.

The more serious the abuse, the greater the over-representation of a step-parent as perpetrator. Seen from another angle: Statistically speaking the abuse is more serious when a step-parent is the abuser.

As regards particularly violent individuals, who might conceivably be over-represented in step-families, a study of the circumstances around the abuse shows that those step-parents who abuse/kill their step-children, at the same time tend to spare their own genetic children. D&W found no examples at all of the opposite: perpetrators who abused their own children but spared their step-children (1985 pp 205-206).

There are parents who kill their own genetic children as well. However, this is not only a far more rare phenomenon, but is usually carried out in other, less painful ways and with other motives. When genetic parents kill their children, the most ordinary methods are by shooting or suffocation (the supposedly least painful methods), the perpetrator often does it as part of a family tragedy in which the same person also kills the partner and himself/herself, the perpetrator is often in a state of mental imbalance or suffers from a psychiatric illness. Step-parents who kill more often do it by beating and kicking, the perpetrator does it in strong anger, and does not kill the partner or himself/herself (1998 pp 34-35; 1994b pp 208-212, pp 214-216; 1996 pp 78-79; 1985 p 198).

In cases of serious abuse and killing of the smallest children carried out by a step-parent, step-
fathers are by far the the most frequent perpetrators. However, few very young step-children live in households with step-mothers; they usually live with their genetic mother and a step-father. The studies therefore do not indicate that women are less dangerous than men when in the position of step-parent (1994b p 208; 1998 pp 60-62).

The incidence of violence in society generally varies between USA, Canada and Britain. Abuse of step-children seems to be independent of this (1994a p 524 footnote 18; 1994b p 212: ‘.. as is generally true where homicide rates are relatively low (Daly and Wilson 1988b), the proportion of cases perpetrated by parents and substitute parents is relatively high in Britain ..’).

4. Possible confounds and sources of error


Other factors than kin-relationship correlate with the occurrence of abuse. Such possible confounds exist:
    bad economy is an important risk factor,
    so is a young mother,
    and likewise a large family (many children).

Potentially relevant factors needing investigation are also:
    a possibly larger than average number of people with violent disposition among the remarried,
    possible bias in the likelihood of detection,
    possible bias in cases being reported,
    a possible underestimation of households with a step-parent.

D&W's statistical results were therefore tested, calculated or estimated against such possible confounds.

It turned out that none of these other factors could be responsible for the very high over-representation of step-children in abuse cases (1996 p 79; 1994b p 208; 1998 pp 27-28, 31-32; 1985 p 198, pp 204-206). There is indeed a higher indidence than average for certain groups defined by such criteria. But the over-representation of step-child abuse comes in addition to this and sends the total up even more. ‘All in all, although several additional risk factors were identified, step-parenthood held its place as the most important predictor, and its influence was scarcely diminished when the statistical impacts of all the other risk factors were controlled’ (1998 p 31).

Exactly the most serious cases, in which step-parents are most over-represented as abusers, are those for which there is the least risk that the abuse can be kept undetected, the responsibility be shifted onto some other perpetrator, be under-reported, or be explained or explained away as an accident, fighting, or friendly punishment having child-raising as its purpose (1996 p 78; 1994b p 208: ‘The present analyses are confined to killings of children under the age of five years. One rationale for this restriction is that these cases clearly cannot be construed as matters of mutual combat or self-defense.’)

The calculation of the proportions of different household types in the population (and therefore of how frequent step-families are) is discussed comprehensively for the Hamilton study (1985 pp 208-209).

If a grouping is made according to household type (step-family versus genetic family), the difference in abuse frequency is not quite so extreme as when a grouping is made on the basis of perpetrator, but is still very large (1991 pp 422-423).

Not all who are counted as step-fathers may actually be so; in some cases he may be the child's father who has established himself together with the mother only after the birth of the child. On the other hand, not all men are genetic fathers who are assumed to be, not even all who believe themselves to be the real father of the child. But when errors based on these divergences are estimated, it leads more in the direction of step-fathers being over-represented as abusers than the opposite (1994b p 209, pp 212-213).

The length of the marriage/partnership, and through that of the relationship child/step-parent, could not be calculated directly. D&W suggest that it should be investigated, and propose a hypothesis of how it might perhaps correlate with incidence of abuse, but they believe it is difficult to test reliably. They also hold that since the length of the relationship to a child of baby age is equally short for a parent and a step-parent, while the difference in incidence of abuse is at its maximum for the smallest children, it is unlikely that length of the relationship will turn out to be of significance (1985 p 206).

5. Criticism, and the response to it


In several of their articles D&W discuss objections and criticism, and account for the relationship to other aspects of behaviour and to theory and possible explanations.

Critics have attempted to throw doubt on just about everything and find alternative explanations. In my view, D&W have analysed such objections and alternative explanations satisfactorily and shown that they are not correct. Some of the objections (including discussion and refutation) are:

    that step-parents are not over-represented among child abusers (1998 pp 50-52; 1991 pp 422-423 about Malkin and Lamb 1989 and own studies; 1988 p 519);

    that D&W have not distinguished between real abuse and physical handling with other motives than enmity, i.e as disciplining or protection (1985 p 207; 1991 pp 119-20);

    that D&W have not been able to identify correctly the perpetrator in step-families and that their grouping based on household type is irrelevant (1998 pp 52-54; 1991 p 423; 1994b p 208, p 209);

    that D&W's figures for the proportion of step-families are too low (1998 p 49);

    that (single) genetic mothers pose a greater threat than step-fathers (1985 pp 207-208; 1988 pp 519-520, pp 521-522);

    that D&W's samples are not representative and that the critics' own studies are better (including one by Richard Gelles, in which people were rung up on the telephone by the interviewer and asked point-blank if they had abused children in the last year) (1998 p 51, pp 52-53; 1985 pp 200-02; 1994b p 209, p 212; 1988 pp 519-20; 1996 p 78);

    that the relatively less favourable position of step-children is due to unfounded stigmatisation and myths, not to less love from step-parents (1998 p 6, pp 56-58);

    that D&W believe child abuse to be an evolutionary adaptation (1998 pp 37-38, pp 51-58; 1991 pp 421-422; 1985 p 207, p 205 against David Finkelhor and others).


Smith, Borgerhoff Mulder & Hill (2000), mentioned above, is a clear and, so far as I am competent to judge, important critical analysis of evolutionary psychology and of D&W's theoretical standpoints in important respects. However, they criticise neither D&W's studies of child abuse, nor the results of the studies or C&W's explanations of the results. Their critique is therefore of little relevance in the present context.

6. Conclusions


The results, then, were unambiguous. The presence of a step-parent is the greatest risk factor for whether a child is likely to be abused (1998 p 31, p 35), and the step-parent is the person responsible for the greatly elevated frequency.

These statistical results do not mean that killing and serious abuse of children in their families is common. People do not live in their homes as sadists with a desire to torture their step-children (1988 pp 60-65). ‘Many, perhaps most, step-parents derive some pleasure from helping raise their partners' children, and many, perhaps most, stepchildren are better off than if their parents had remained single’ (1998 p 38).

The reality is rather that in situations under the strain of conflict, problems or demands felt to be unreasonable, people tend to react differently depending on the kind of person who disturbs, pesters or demands.

Abuse and killing are therefore not treated as isolated phenomena by D&W, but as a consequence of a more general difference in the relationship parent/child and parent/step-child:

There is a fundamental emotional difference between these two relationships, a difference great enough to yield statistically measurable results (1991 p 421). Step-parents have less love, less of an impulse to protect against harm from outsiders (1985 p 207), less thoughtful consideration for their step-children than parents have for their own children. The presence of step-children in a family is correlated with a higher divorce-rate, while the presence of children in common is correlated with a lower rate (1988 p 521).

    ‘Step-parents are primarily replacement mates, and only secondarily replacement parents.’ (1998 p 64);
    ‘But conflict in stepfamilies is not confined to these extremes’ [viz: serious forms of child abuse] (1996 p 79);
    ‘Step-parents do not, on average, feel the same child-specific love and commitment as genetic parents, and therefore do not reap the same emotional rewards from unreciprocated 'parental' investment.’ (1998 p 38);
    ‘There is a substantial literature on "reconstituted families," ranging from empirical surveys through anecdotes and autobiographical tips to exhortative pop psychology. The prevalent theme is
coping, and the literature abounds with acknowledgments that step-relationships are stressful’ (1985 p 209).

Step-parents invest less in the education of step-children and have lower goals for their education (even lower than that of single mothers with poor economy for their children), step-children leave their parental home earlier and feel to a greater extent that they are pushed out, step-children are over-represented among homeless teenagers (1998 p 22, pp 55-56). ‘... that we were looking at what might be called a 'reverse assay' of parental love. A paucity of heartfelt, individualized concern for the welfare of a child in one's care would seem likely to raise the incidence of any sort of misuse’ (1998 p 30).

7. The explanation


Why is behaviour like this?

D&W write quite a lot about feelings. Feelings are correlated with biological evolution (1994/95 pp 1272-1273).

Feelings make us react in certain ways in certain situations. The actions that spring from our feelings have a result. The result can affect whether we survive and breed. The evolutionary mechanism for parental love must be assumed to operate like this:

Which children will on average do best so that they survive until adult age when they can themselves breed children? – Those who grow up with adults who render care, i.e supply nourishment and protection for the children.

Which adults render the best care and protection for children? – Those who are not dependent on deliberation, reasoning, teaching or acquired morality in order to take good care of children but who are driven to do so by spontaneous feelings – by a strong urge to help and protect children. Such a drive is a personality trait lodged in the psyche of the individuals who possess it.

But if this quality is to be transmitted from an individual A possessing it, it must happen through inheritance, i.e because the quality has a basis in the genes of A and is transmitted when the carrier genes are inherited by the descendants B. Heritable drive to feel and act with love towards children can, then, only be transmitted from A to A's own offspring, who will therefore have the same drive.

However, the offspring B, having inherited their parents' tendencies to take good care of children, must try to survive till adult age before being in a position to exercise the instinct to care. If B's parent A practiced his eagerness to care not towards B but towards any children at all, A would be a lucrative target for the ambitions of totally different, egoistical grown-ups C, who could then get away with not taking care of their own children D but leaving them to altruistic A to rear. C would be free and fit to breed even more offspring. B would be neglected and pushed aside by D in the competition for A's care and would be less likely to survive than if he had had his parent A's undivided attention. If B died before he reached an age to breed, A's lineage would go extinct and with it the drive to care for children. The gainer would instead be D, who had far from inherited any such kindness from his selfish parent C.

Therefore, no such unselfish A – equally helpful to all – has won through in the course of evolution. The only lineages that have survived in species in which the offspring needs care and help from adults, are lineages with parents who feel a strong, protective and unselfish love towards
their own offspring. This discriminative love has two results: (1) It causes the possessor's own children to receive much better care and protection than the parent would have the strength to give to others. The offspring's possibility of doing well while growing to fertile age is thereby increased, so that the lineage can continue. (2) Offspring that in this way gets extra good care, has a greater chance of passing on to new generations the same specialised, discriminatory parental love because this love is genetically based.

Parental love is clearly no diffuse love for all and any children, it is strongly channelled towards the individual's own children. The quality of possessing a strong love of children can develop in some individuals through natural variation, but it can be transmitted only if it is directed toward the possessor's own children, in such a way that they are the ones who receive the better care (1994/95 pp 1269-70; 1998 pp 8-17).

The research results of D&W about the behaviour of parents and step-parents are seen to agree very well with hypotheses from evolutionary biology, which can also render an explanation of such differences in behaviour as would otherwise be difficult to find: The lineages that have survived and have been able to pass on its aptitudes and qualities to new generations through genetic inheritance, are those that have evolved a genetically based disposition to feel strong love for their own offspring, making them see to it that the offspring survives in the best way possible, both in struggle against dangers, in competition with others over resources, and in conflicts with the parents themselves.

8. Daly and Wilson versus representatives of social disciplines


D&W have not included adopted children as a separate category in all their studies. Parly this is due to some records not making it possible to distinguish adopteds from genetic children, partly it is because even where they can be identified in the records, the numbers are small.

But it is also because D&W assume equal motivation in both partners wanting to adopt. They furthermore assume that the possibility of returning the adopted child plus the screening of adoption-seekers by social agencies result in adopted children not being greatly at risk for abuse and thus in a different position from that of step-children. To the degree it has been possible to find numerical material, these assumptions are borne out (1985 p 206; 1988 p 524 footnote 31). They have also excluded foster children for seemingly similar reasons.

Even so, D&W's studies are very important in connection with the practice of fostering and adoption, too, because the studies show clearly that conditions are different when a child grows up with a care-giver who is not a close genetic relative from when it does not.

D&W seem not to be well informed about the activities within child protection in many countries. Considering the important information they have succeeded in bringing to light through their research, they might have a good basis for going into questions of social service agencies' actions regarding children (cf for instance an editorial article in Massachusetts News: "Social Workers Meet Counter Protest at State House"). On the whole, D&W are realistic and far from credulous, but strangely enough they seem a fraction naïve just when it comes to assuming that social workers' work with children is always real and helpful. A possible explanation might be that D&W have been assisted precisely by social agencies in obtaining records of child abuse and other information they needed for their studies; they may perhaps therefore have come to trust the representatives of these agencies generally.

Still, this acceptance is surprising, since D&W have in what they have written many observations and discussions of social science attitudes which go precisely to the core of the weaknesses of these attitudes – which in fact floodlight why it is that the ideologies and obsessions of social agencies lead to such catastrophic results as they do in practical child protection work. The proponents of these ideological trends are even the same as those who have opposed D&W and tried to explain away or deny the vulnerable position of step-children because it does not suit their ideology, an ideology which completely dominates in the professions in charge of foster and adoption administration. However, even though D&W could have good reason to be suspicious, it must be admitted that the professions dealing with children and child protection take very good care that real information about the state of affairs in child protection and allied psychology is not made easily accessible but is hidden away as much as possible.

People in the branches of social science and psychology dealing with questions regarding children, not least child abuse, are largely adherents of social-deterministic schools of thought, e.g of psychodynamic (Freudian) psychology, which must be classified as a very primitive type of determinism, and which has long ago been exposed as completely unscientific. Such trends or schools of thought figure heavily in some university circles and still more in organisations and agencies carrying out practical social work among children and the young (1998 p 58).

These trends have it almost programmed that only features of the material environment are interesting for behaviour, and that the same features deterministically govern behaviour and also feelings. Feelings are understood as purely a result of environmental influences, not as an independent motive force. Once again we have here a trend in culture viewing the mind of human beings as "tabula rasa" - a blank slate which can be imprinted equally well in any way whatsoever (cf Freeman 1983, ch I: "The Emergence of Cultural Determinism"). Although interested in the way in which different social environments influence children, the adherents of this simple determinism have not been alert to how important it is for children to grow up in their own biological environment – their kin group (1988 p 523: ‘Analyses of "family violence" have hitherto ignored crucial distinctions among relationships’ [viz: in type of relationship between individuals]). These professional groups are ordinarily badly informed and uninterested when it comes to evolutionary biology and biology generally and are intellectually distant from a framework which makes it natural to wonder if factors like genetic relationship between people living together could be of great, perhaps decisive, importance.

One of the results of social-determinstic thinking is a popular dogma in social work of our times: that the family is a so-called social construction – an arbitrary joining together of just any individuals, with no natural basis for the fact that parents and their own children are ordinarily the ones who stick together in a family. Those who believe in this dogma, will as a corollary have it that family life is activity in which the participants play social roles; one
is not mother, father or child, one acts in one of these roles (1996 p 80; 1998 pp 56-58). The ideology preaches that genetic relatedness is without importance. (A variant of the "role" theory also claims that belonging to one of the sexes is a matter of role-play which is induced socially and can easily be changed. This view seems to consider genetially and hormonally deviant individuals as proof, and extreme action taken on such a basis has not been without resulting tragedy, cf articles about Dr John Money (e.g Wikipedia 2012).) So-called "attachment theory" is an implementation of the same popular speculations of social-deterministic type.

Why D&W's results are unwelcome among people with such an ideological basis is easy to understand. Though many of them have been interested in child abuse and engaged in trying to predict and stop it, none of them had discovered the astonishingly large differences due to precisely biology, and it may well be difficult for them to understand such facts.

One should not get lost in speculations about the motives of researchers, but I think I sense, behind the unwillingness of social science professions to take in the implications of D&W's research, a professional upset and jealousy. D&W's studies have brought out facts which are unwanted in large sections of social professions and social sciences (1998 pp 15-16, p 42). Their work is not well known in the social professions. But to the degree it is, this is exactly where the protests come from. It is not unexpected, for instance, that an attempt at contradiction of D&W's results came from Richard Gelles, a well-known investigator of family violence, and from the American Humane Association and American Medical Association, taking Gelles as their point of reference (1991; 1998 pp 51-55). I wrote the following about Gelles' criticism in the article "To uheldige punkter i en høringsuttalelse fra Foreningen 2 Foreldre (F2F)" ['Two unfortunate points made by The Association 2 Parents in a submission to a departmental hearing']:
    ‘Reading Daly and Wilson's article "A Reply to Gelles", it can be seen that Gelles is a scholar with an agenda: he follows the trend of the times in closing his eyes to, in fact rejecting, the importance of biology for feelings and behaviour. Added to this he, considered a leading authority on child abuse, seems to dislike for others to contribute important results to the research in areas which he has failed to address. As can be observed from "A Reply to Gelles" (pp 422-23), Gelles believes he can prove that step-parents
are not over-represented against genetic parents as child abusers. Gelles bases this on a study he has conducted which he thinks is better. In Gelles' study of the occurrence of child abuse, people were called up on the telephone by unknown researchers asking them whether they had carried out different types of abuse or violence against their children in the last year. The answers obtained are thought by Gelles to be more reliable than the registrations in e.g police archives which Daly and Wilson have made use of. – Further comments about this should be unnecessary.’

D&W suggest a more respectable reason for the reaction of the social professions against exposing the situation of step-children, viz sympathy: ‘Perhaps the main reason is that the writers feel that stepfamily life is hard enough, without adding to the stigma’ (1998 p 56).

The result of the sympathy is no help, however: ‘Most of those professionally concerned with stepfamilies are practitioners first and scientists second. Fearing the insidious effects of 'self-fulfilling prophecies', they have understandably made it their business to offer cheer-leaderly encouragement. Unfortunately, in attempting to counteract stepfamily 'myths', they have created a counter-factual mythology of their own, in which social relationships can be reordered by fiat and the statistical facts about differential violence can be dismissed’ (1998 p 58).

I accept the sympathy-and-give-help explanation as valid for some participants in debate and/or active work, but I think I read the anti-biological trend of our times in parts of science and professional practice as more important, and as harmful derailing and going astray.

In contrast to social professions who ought to be wide awake to D&W's research results but who close their eyes, others do valuable research of the same kind as D&W, and come up with similar results. Flinn and England (2003) say (we bear in mind that cortisol is a stress hormone): ‘The family is of paramount importance in a child’s world. Throughout human evolutionary history, parents and close relatives provided calories, protection, and information necessary for survival, growth, health, social success, and eventual reproduction. The human mind is therefore likely to have evolved special sensitivity to interactions with family caretakers, particularly during infancy and early childhood’ (p 109); ‘Children living with non-relatives, stepfathers + half-siblings (stepfather has children by the stepchild’s mother), or single parents without kin support had higher average levels of cortisol than children living with both parents, single mothers with kin support, or grandparents... A further test of this hypothesis is provided by comparison of step and genetic children residing in the same households .... Stepchildren had higher average cortisol levels than their half-siblings residing in the same household who were genetic offspring of both parents’ (pp 119-20). Not unexpectedly, Flinn and England have a reference to Daly and Wilson's 1995 article.

In conclusion I can hardly do better than to leave the word to Martin Daly and Margo Wilson. They formulate the absolutely central insights that spring from their research work:

‘American social scientists have interpreted stepparenthood as a role, only partly coincident with that of genetic parenthood. The role concept has usefully directed attention to the importance of socialization and scripts, but it is at best a limited metaphor that has diverted attention away from motivational and emotional aspects of the social psyche. There is more to social action than mere familiarity with the relevant roles. Why are people motivated to embrace certain roles and to shun others? Parents are profoundly concerned for their children's well-being and future prospects, but human concerns have no part in role theorists' explanations of human action.’ (1996 p 80);

The profundity of parental love
    The proposition that stepchildren are not loved like genetic children strikes many social scientists as distasteful. We believe that it is this distaste that has motivated the curious attempts, which we discussed in Chapter 6, to sweep the data away. Ironically, this dismissive stance tends to go hand in hand with an assumption that interpersonal attachments are arbitrary 'social constructions', an assumption that implicitly denies the profundity of parental love and can lead to such inhumane policy prescriptions as have sometimes arisen in extremist Utopian communities, in which infants were removed from their parents to be reared by child-care professionals.
    Is the Darwinian world-view uglier in its implications? We think not. More generally, we reject the curiously prevalent notion that a scientific, materialistic, Darwinian world-view is uglier than its anti-scientific alternatives. Instead, we would suggest that more realistic world-views invite more humane attitudes and practices than fantastic ones, because they entail better models of human nature and hence greater sensitivity to human needs and desires’ (1998 pp 65-66).

9. Literature


Martin Daly and Margo Wilson (1983): Sex, Evolution, and Behavior. 2nd edition. Belmont, Calif.: Wadsworth. ISBN 0-87150-767-6

-  (1985): "Child Abuse and Other Risks of Not Living with Both Parents",
Ethology and Sociobiology 6 (pp 197-210). New York: Elsevier Science Publishing Co Inc

-  (1988): "Evolutionary Social Psychology and Family Homicide",
Science, 28. October 1988, vol 242 (pp 519-524). American Association for the Advancement of Science

-  (1991): "A Reply to Gelles: Stepchildren
are disproportionately abused, and diverse forms of violence can share causal factors", Human Nature Vol 2 no 4 (pp 419-426). New York: Walter de Gruyter Inc

-  (1992): "Who kills whom in spouse killings? On the exceptional sex ratio of spousal homicides in the United States",
Criminology 30 (pp 189-215).

-  (1994a): "Evolutionary psychology of male violence", John Archer (ed):
Male Violence (pp 253-288). London: Routledge. ISBN 0-415-08962-X

-  (1994b): "Some Differential Attributes of Lethal Assaults on Small Children by Stepfathers versus Genetic Fathers",
Ethology and Sociobiology 15 (pp 207-217). New York: Elsevier Science Publishing Co Inc

-  (1994/95): "Discriminative Parental Solicitude and the Relevance of Evolutionary Models to the Analysis of Motivational Systems", Michael S. Gazzaniga (ed):
The Cognitive Neurosciences (pp 1269-1286). Cambridge, Mass: MIT Press

-  (1996): "Violence Against Stepchildren",
Current Directions in Psychological Science 1996:5 (pp 77-81). American Psychological Society

-  (1998):
The Truth about Cinderella. A Darwinian view of Parental Love. London: Weidenfeld & Nicolson. ISBN 0-297-84161-0

-  (1999): "Human evolutionary psychology and animal behaviour",
Animal Behaviour, 1999, 57,3 (pp 509-519).

Martin Daly, Margo Wilson and Shawn Vasdev (2001): "Income inequality and homicide rates in Canada and the United States",
Canadian Journal of Criminology, April 2001 (pp 219-236).

Mark V. Flinn & Barry G. England (1995): "Childhood Stress and Family Environment",
Current Anthropology, Vol 36 No 5, December 1995

MV Flinn & BG England (2003): "Childhood stress: endocrine and immune responses to psychosocial events", in JM Wilce (Ed.):
Social & Cultural Lives of Immune Systems. London: Routledge press.

Derek Freeman (1983):
Margaret Mead and Samoa. The Making and Unmaking of an Anthropological Myth. Cambridge, Mass: Harvard University Press. ISBN 0-674-54830-2

(editorial) (2002): "Social Workers Meet Counter Protest at State House". Holliston, MA 01756: Massachusetts News.

(editorial) (1993): "Murder, She Said",
Psychology Today, Mar/Apr 93. Document ID: 1705. New York: Psychology Today Magazine.

(Wikipedia, updated 11 April 2012): "John Money"

Marianne Haslev Skånland (6 Jan 2005): "To uheldige punkter i en høringsuttalelse fra Foreningen 2 Foreldre (F2F)" ['Two unfortunate points made by The Association 2 Parents in a submission to a departmental hearing']. Oslo: Barnas Rett,

Eric Alden Smith, Monique Borgerhoff Mulder & Kim Hill (2000): "Evolutionary analyses of human behaviour: a commentary on Daly & Wilson",
Animal Behaviour, 2000, 60,4 (pp F21-26).