The interdependence of agonistic and affiliative behaviour


by John Price and Leon Sloman


ASCAP November 93. 6, 10-14


What follows is a response to John Birtchnell's accusation (September, 1993, ASCAP) that some of us are "vertical thinkers" and talk about social hierarchies to the exclusion of affiliation.  It represents the results of discussions at odd times during the Human Behavior and Evolution Society (HBES) meeting at Binghampton, August, 1993.  We tried to list as many areas as possible in which agonistic behaviour (the primitive work-horse of the vertical dimension) and affiliative behaviour interact.  We found they could be classified into two main categories:




"The fiercest anger of all, the most incurable, is that which rages in the place of dearest love" said Euripides, and still over two thousand years later we find that the fiercest battles over the vertical dimension occur in loving marital and family relationships.  In a loving relationship, the cause of agonistic behaviour may be a conflict over affiliation, or the weapon used may be borrowed from the affiliative repertory.


Use of withdrawal of love (or threat of same) as a weapon


When someone feels angry, they often want to hurt the other as much as possible.  They will use any weapon that comes to hand.  If they are in a loving relationship with the person they want to hurt, they may know that the most hurtful thing they can do is to deny their love.  Therefore they say "I hate you" or "I no longer love you" or, even worse, "I never really loved you", or, playing it dirty, "The child you think I had by you is not really yours."  These statements are not true, usually, and therefore have nothing to do with affiliative behaviour, or with the termination of affiliation (unless they are taken literally, and the weapon misfires).  Such statements are used because they are the most effective way of inflicting pain, which is one of the main goals of the angry person.  The result may be to cause suffering, depression, and even a clinical depressive illness, as the following case illustrates:


     A male university student was referred with severe endogenous depression.  He was indeed very depressed, and said, "I can't concentrate on my studies, I can't laugh with my friends, it all seems pointless."  There was impairment of sleep and appetite.  He told the therapist that a month previously his fiancee had asked him to take off a cheap metal ring which had been given him by a former girl-friend.  He had refused, saying, "It's only a bit of metal".  She had withdrawn her love, and shortly after this he became depressed. 

        The therapist pointed out to him that a ring is a symbol of commitment, and that if he wore a symbol of commitment to another woman, he could not expect his fiancee to commit her love to him.  The patient accepted that he was in the wrong and had acted badly, and as an act of atonement he gave the ring to his fiancee.  She was then able to forgive him, and his mood started to improve.  The active submission of taking off the ring had replaced the involuntary yielding of depression.  As a result of this episode he learned that he could not have his own way all the time, but must accommodate to the needs of his partner.


   This case illustrates one complex interaction of agonistic behaviour and attachment behaviour.  In the agonic mode the mental state is one of wanting to hurt the other person, and any methods that come to hand may be used.  This patient's fiancee knew that the most hurtful thing she could do was to withdraw her love, and this was far more effective than punching him on the nose.  The withdrawal of love was not a part of attachment behaviour;  it was borrowed, as it were, from attachment behaviour for use in agonistic behaviour, and it was not real.  But the recipient can easily be confused and mistake it for real, in fact its effectiveness in causing pain depends on just such a confusion.  Therefore it is a dangerous weapon which can easily get out of hand and go too far.

   This case also illustrates the different attitude to value adopted by men and women.  The patient thought the ring to be of no importance because it was made of some base metal, and had no financial worth.  His fiancee was not interested in its resale value, but only in its symbolic value as a sign of commitment to another woman.  Its transfer to her own possession was a symbol of his renewed commitment to her, and possibly also a symbol of his ability to learn to accommodate his base male attitudes to her own more refined value system.  We expect she will keep it till it rusts away.


Conflict over affiliative behaviour


Agonistic behaviour is the primitive method of settling conflict, and the vertical dimension reflects the result of previous agonistic interaction (and predicts the outcome of future agonistic interaction).  Conflict can be about anything, including affiliation;  in fact, in a loving relationship conflict is very likely to be about affiliation.  A person either gets too much of it from the other, so that his need for "distance" is unsatisfied; or, more often, he gets too little of it, so that his need for "closeness" is unsatisfied. In either case, he suffers "frustrative non-reward" which can give rise to either aggression or depression (depending on the result of a complex appraisal system).  We know that one of the common causes of the "battered wife syndrome" is an attempt by a wife to leave her husband.  The latter at first coerces her to stay, and finally takes the view that, "If I can't have her, no-one is going to have her."


Inhibition of aggression due to affiliative factors


A man may want to hit his wife, but does not do so because she has told him she will leave him if he does so, and he does not want to lose her.  Or he may refrain from hitting because his mother told him that men do not hit ladies;  or because a similar rule has been absorbed from other cultural sources.

   This inhibition is of particular interest in the causation of depression.  We have suggested that a catathetic (putting-down) signal lowers RHP if it not reciprocated in full measure;  and lowering of RHP is probably the final common path in the complex network of causal chains leading to depression.  It does not seem to matter what reason there is for non-reciprocation;  it could be the usual one that the person concerned is losing an agonistic encounter, but it could also be an affiliative reason that has nothing to do with losing.  And yet the effect seems to be the same.  RHP is reduced and may trigger the "involuntary subordinate strategy".  An exception to this is when the husband is so secure that he is unaffected by his wife's aggression, and, for instance, discounts it as her "time of the month" or uses the arousal associated with it to switch the context of interaction from agonistic to sexual.


Strengthening the affiliative bond through agonistic interaction


Lorenz claimed that an affiliative bond which had been tested by a sequence of agonistic interactions followed by reconciliation might well be stronger than a bond not so tempered in the fire of agonism.  Some support for this comes from De Waal's work on reconciliation in chimpanzees.  This is a large subject which we will not do more than mention here.


Attractiveness of power and resources


It is well known that human females are attracted to high ranking men, and that "money is the best aphrodisiac".  This is natural because such men are better able to provide resources for their children.

   For the same reason, possibly, men may find rich and high-ranking women attractive, although many men are put off by dominating behaviour in a woman (and perhaps nowadays women are increasingly turned off by dominating behaviour in men).  This is a field being explored by David Buss, who spoke at the meeting.


The demonstration of attractiveness by valour in battle or successful generalship


Agonistic performance has been monitored during history probably as closely as batting and bowling averages are monitored in England today.  Battle against neighbouring groups is an excellent projective test for females to assess the desirability of men as husbands and lovers, and for parents to discriminate between competing sons-in-law.  Bravery, reliability, altruism, co-operativeness and endurance could be observed in the old days and now can even be seen on television.

   Anthropologists enquiring about the sources of prestige in various societies are often told that the greatest reputation goes to the man who leads the group successfully in battle.


Exchange of affiliation for agonistic support


Those who study reciprocal exchange in monkeys have found that the reward for support in agonistic interactions may be not only support in future interactions but some affiliative behaviour such as grooming.




Dependent rank


Ranking theory cannot ignore affiliation because in most primates, and most of all in human beings, rank depends on alliances with other group members.  Concerning non-human primates, Bernstein & Gordon (1) wrote:


      ...animals living in a society depend more on alliances and coalitions rather than on individual fighting skills to maintain their social position.  Thus a scrawny old female supported by many generations of offspring and long associations with other females and adult males may maintain a position of unquestioned superiority over young males of much greater fighting ability.  So, too, may an old male retain his high rank, eventually losing to a challenger not because of his failing fighting abilities but because of the successful recruitment of support from group members by the challenger.  After such a defeat, a new order is established incorporating all animals into a society that recognises the new relationships.


This dependence of rank on alliances has been amply confirmed by more recent primate studies.  Bernard Chapais (2) sums up his review of ranking behaviour in female Cercopithecine monkeys as follows:


     Perhaps the most general principle emerging from the present review is perhaps that any female seeks to outrank any other female against whom she is given sufficient alliance power.


In their introduction to their book on alliances, Harcourt and De Waal (3) sum up their view of alliance formation and agonistic behaviour in man and other primates:


     If we view the dominance hierarchy as the vertical component of social organisation, the network of affiliative and kinship ties can be viewed as the horizontal component.  In many species these two components exist side by side without much interplay.  The remarkable social complexity of human and non-human primates is brought about by their capacity to (1) alter competitive outcomes and dominance positions through collaboration, and especially (2) establish social bonds for this very reason.  Alliance formation links the vertical and horizontal components of social organisation by making an individual's dominance position dependent on its place in the affiliative network.  Consequently, this network becomes an arena of dominance-related strategy.


These phenomena can be described under the headings of parental influence, female dependence on her male consort, and the forging of alliances:


Parental influence.  We know that in humans social rank is frequently hereditary.  If we doubt this we only have to go to Buckingham Palace and make an application to be crowned king.  What is now being realised is that the same applies to a lot of monkeys, and it seems likely that kin-dependent rank was a characteristic of the common simian-human ancestor some 40 million years ago.  So there have been a lot of generations for natural selection to act on the links between rank and kin affiliation.  We recommend the volume by Harcourt and De Waal, especially the chapters by Chapais and Datta.  In these monkeys, interventions by parents and other relatives in the quarrels of youngsters result in the rank ordering of whole matrilines, within which the daughters are ranked according to the "youngest ascendancy principle", so that each youngest daughter ranks immediately below her own mother, and is dominant not only to her older sisters, but to all females who rank below her mother.  One  can visualise the hectic social life of the group as the youngest daughter of the dominant female achieves a rank reversal with every other female in the group, except her mother, as she grows up.  In these Cercopithecine monkeys, rank and kinship are the same, at least for the females.


Female dependence on male consort.  In a number of animal species, the rank of a female depends on the rank of the male she is associated with.  This results in some very flexible rank situations, especially in harem situations such as the hamadryas baboon, in which the male consorts with whichever of the harem females is in oestrus.  As a result, each female rises to the top of the female hierarchy once a month, and just as surely falls to the bottom once a month;  and it seems as if, when they enjoy their brief reign at the top, they are more concerned to avenge past slights when they were down in the past than to build up credit for when they will be down in the future.

   Real, if less dramatic, changes occur with human beings.  When an English lady is widowed, she traditionally hands over the manor house to the wife of her eldest son and retires to the dower house, where her power is much reduced and she yields formal precedence on social occasions (such as order of going into dinner) to her daughter-in-law.  In other cultures, the situation of widows is even worse;  for instance, it may be thought that "widows are unlucky and should be shunned" (4).  Shweder continues:


     In the Hindu moral world the death of a husband has more than moral significance, and its metaphysical meanings run deep.  Traditional widows in India spend the balance of their lives absolving themselves of sin (fasting, praying, withdrawing from the world, reading holy texts).  In their world of retributive causation, widowhood is a punishment for past transgressions.  The fact that your husband died first is a sign telling you that you must now undertake the task of unburdening yourself of guilt, for the sake of your next reincarnation on earth.  In such a world the flame of your husband's funeral pyre is appreciated (understood and experienced) as though it was the romantic analogue of the last plane to Lisbon in the movie Casablanca.  If you are not on the plane it is likely to be a very long time until you see him again, if ever. A shared cremation absolves sins and guarantees eternal union between husband and wife, linked to each other as god and goddess though the cycle of future rebirths.


The forging of alliances.  Two can fight better than one, so the result of any conflict depends more on who is fighting with whom than on the fighting ability of any particular individual.  Alliances may be formed with kin and non-kin.  A famous mythical example of the latter is the alliance between Theseus and Pirithous, who formed an unbeatable team and must have caused a lot of grief around the Aegean Sea in those early days.




We think it must be clear from the above that agonistic and affiliative behaviour are inextricably intertwined in human interaction.  The reasons we talk more about agonistic behaviour are twofold.  First of all, it is in relation to agonistic behaviour that the biological advantage of depression lies.  If there were no agonistic behaviour there would be no depression.  You cannot say the same about affiliative behaviour.  Depression plays no adaptive role in relation to affiliative behaviour - its manifest connection with affiliative behaviour is secondary, due to the enormous importance for agonistic behaviour of alliances and dependent rank.  We have hypothesised that depression is an intensified and prolonged involuntary subordinate strategy and this should not be needed for affiliative behaviour which should be egalitarian (at least between adults).  Depression can maintain the stability of close relationships that are threatened by conflict, but at the expense of symmetry;  the depressed person switches into a subordinate role and the relationship becomes complementary (in terms of power);  if there were no vertical dimension this form of " relationship homeostasis" would not be necessary.

   Secondly, the relation between depression and loss of loved ones has been well documented and discussed by others, whereas the relation of depression to agonistic behaviour has, we feel, been largely overlooked;  and the reason for this is the lack of an evolutionary perspective in previous thinking.  In particular, the failure to distinguish the relative rank of the object of hostile expressions in depressed patients has concealed the very considerable degree of agonistic inhibition in depression, which we believe to be its main adaptive function.

   The treatment of depression should include attention to affiliative networks.  It is not sensible to ask the depressed person to make new friends, because all social initiative is inhibited, and the therapist would just be setting the patient up for further failure.  However, the mental state of depression is conducive to mending feuds in existing relationships, especially when the feud is due to pride on the part of the patient.  As Alexander Shand said many years ago, "Sorrow abates pride" (5).  When the patient has recovered, there is the possibility of extending the social network, and of resuscitating friendships which have lapsed;  we believe that a wider and closer network of friends and relations would be one factor in preventing relapse.


1.  Bernstein, I.S. & Gordon, T.P. (1974) The function of aggression in primate societies.  American Scientist, 62, 305-311.


2.  Chapais, B. (1992) The role of alliances in social inheritance of rank among female primates.  In: Coalitions and Alliances in Humans and Other Animals ed A.H.Harcourt & Waal.  Oxford: Oxford University Press.  Pp. 29-59.


3.  De Waal, F.B.M. & Harcourt, A.H. (1992) Coalitions and alliances: a history of ethological research.  In: Coalitions and Alliances in Humans and Other Animals ed A.H.Harcourt & F.B.M. De Waal.  Oxford: Oxford University Press.  Pp. 1-19.


4.  Shweder, Richard A. (1991) Thinking Through Cultures:  Expeditions in Cultural Psychology.  Cambridge: Harvard University Press.


5.  Shand, A.F. (1913) The Foundations of Character.  London: Macmillan.


If somebody attacks the habits...which characterise me...they are negating me.  If I care deeply about that other person, the negation of me will be still more painful."  Bateson Steps (pp. 212-213) quoted by Sebastian Kraemer (1993) in Books reconsidered: Steps to an Ecology of Mind by Gregory Bateson.  British Journal of Psychiatry, 163, 265-268.