Vol 4, no 8 August, 91


Conspecific comparison - a reply to Paul Gilbert


Paul is right. I should simply have presented the theory (that depressive  states evolved as part of the yielding component of agonistic behaviour),  rather than confuse everyone with trying to rebut an objection (that  depression is essentially yielding behaviour but depressives sometimes seem  adept at getting their own way). Also, the title should have been  "Metaphors of Yielding", a term which encompasses both voluntary submission  and depressive yielding.

  In his distinction between social comparison by intimidation and social  comparison by attraction I think Paul has made a very fundamental point -  one that has certainly clarified things for me. This thinking about social  comparison in an evolutionary context starts with Darwin's idea of sexual  selection, and then there is Ginsberg's view of social organisation as  providing an arena for social comparison, and then Wynne-Edwards idea of  conventional competition for conventional goals;  but I cannot recall  anyone getting anywhere near Paul's analysis. We can now see ritual  agonistic behaviour as one form of social comparison which is one form of  the social competition which subserves intrasexual selection. Since this  is all very new, I think it might be helpful if I just reflect back to Paul  my current thinking about the evolution of social competition as it has  been affected by his contribution.

   The evolution of human social life can be seen as the evolution of ever  more sophisticated and effective methods of sexual selection, and it might  be helpful to enumerate some of the possible stages, concentrating mainly  on the intrasexual component of sexual selection:


1. Unritualised social competition


Many insects kill members of the same sex, some worms plug each other's  sexual orifices, some beetles spray each other with anti-aphrodisiac gas; this category includes any action to reduce the other's viability or  fertility over which the victim has no role in "consenting". Possibly the  suppression of sexual development by pheromones in some rodents and new  world monkeys comes into the category, otherwise it does not occur in  vertebrates.


2. Ritual agonistic behaviour


In ritual agonistic behaviour the loser, being unharmed, must consent to  lose. He has the option (at an unconscious level) of not consenting, and  can be said to choose between consenting and non-consenting strategies.

   In evolutionary terms, ritual agonistic behaviour seems to be performing  two rather separate functions. To the extent that it takes the form of  intergenerational conflict, it serves to delay reproduction until later in  the life span. To the extent that it is intragenerational conflict, it  serves to create lifelong variation in fertility within a socially  interacting cohort of cconspecifics. It is this second function which  subserves intrasexual selection, and which concerns us here.

   According to the simplest view, each individual has to choose between  two strategies, a dominant strategy in which he reproduces more and a  subordinate strategy in which he reproduces less. The dominant strategy is  designed to maximise his own reproduction, the subordinate strategy is  designed to maximise the reproduction of his close kin (and depends for its  selection on the "kin selection" component of inclusive fitness). For  individuals in this system, the overall strategy is similar to Maynard  Smith's "assessor" strategy;  on at least one occasion in their ontogeny  they have to assess their chances and choose between the dominant or "hawk"  substrategy on the one hand, and the subordinate or "dove" substrategy on  the other.

   How does they make this choice?  There are a number of possibilities  which have not in fact evolved. They could leave it entirely to the  opposite sex and adopt the strategy "If chosen as a mate, adopt dominant  strategy;  if not chosen, adopt subordinate strategy", thus relying  entirely on the intersexual component of sexual selection and eliminating  the intrasexual component. Or, they could do it by counting heads, such 

as, "If the home/nest contains more than x individuals when you reach age  y, adopt subordinate strategy, otherwise adopt dominant strategy."  What  has in fact evolved is a form of social comparison, which bears a certain  resemblance to co-consultation. We can imagine a primitive vertebrate,  scratching its head and wondering whether to adopt a dominant or  subordinate strategy, so it chooses a consultant, and says, please help me  make up my mind. The consultant says, use me as a yardstick, if you find  yourself superior to me, your chances are good and you should adopt the  dominant strategy;  otherwise you should play safe and adopt the subordinate strategy. The consultation takes the form of a fight in which  our indecisive individual uses the strength of the consultant as a  yardstick to estimate his own strength, and after the consultation he  either says to himself "I am a strong person" and adopts a dominant  strategy, or "I am a weak person" and adopts a subordinate strategy. Of  course the interaction is symmetrical and the "consultant" is making a  similar decision (it is a co-consultation). This could be called dyadic  comparison because, in each comparative episode, each individual compares  himself with one other. It is the main form of vertebrate social  comparison and is called ritual agonistic behaviour.

   As Paul points out, group living provides the opportunity for more  sophisticated social comparison, and probably the selection mediated by  ritual agonistic behaviour in groups is more effective than occurs in  territorial species. The opportunities for effective comparison in a  single fight are limited, but if animals live in groups they have extended  time in which to evaluate each other's strengths. Fights can follow a long  period of mutual assessment, can be protracted, and be divided into bouts.

   As a result, the rank order in a group should reflect small differences  in strength, skill, intelligence and courage (the components of RHP). Ritual agonistic behaviour amplifies these small differences into gross  social disparity.


3. External mediation of intrasexual selection 


In human evolution there has been a major change in ranking behaviour. Instead of two rivals A and B fighting it out between themselves, the  choice between A and B is made by C, D, E etc. This is the change which  Paul has pointed out as so important. In order to achieve social succcess,  A has to make himself attractive to C,D and E rather than make himself  intimidating to B. Selection is now by external judges rather than by 

interaction between the rivals themselves. The scope for greater  efficiency of selection, and for cultural variation in the criteria of  selection, opens up an entirely new "ball game" of the sexual selection  process. In fact, this development must have been about as important as  the development of sexual selection itself. To distinguish it from the  dyadic comparison which occurs in ritual agonistic behaviour, the  evaluation of A and B by C,D and E etc could be called polyadic comparison. Of course it is seldom as simple as that, and in most cases everyone is evaluating everyone else.

   Does this kind of sexual selection occur in animals as well as man?  In  macaques, baboons and chimpanzees the outcome of ritual agonistic behaviour  is affected by alliances with same-sexed conspecifics, so that the capacity  for alliance formation is being selected for as well as fighting ability. The choice between two potential allies offers a primordium of polyadic  comparison, in that the criterion of choice is not so much "Does he  intimidate me?" as "Is he likely to intimidate the other fellow (and, if  so, is he likely to favour me)?";  this is still some way from, "Which of  the two is more attractive?", but it as a major advance from the evaluation  of others entirely in terms of dyadic comparison. In chimpanzees, in  addition, the influence of female group members affects the rank order in  males, and this is a further step towards intrasexual polyadic comparison;   in fact, it is similar to the situation in at least one tribe of American  Indians in which only males are allowed to run for office and only females  are allowed to vote.

   In human society polyadic comparison has been enormously increased in  importance, particularly due to language and the opportunity this gives for  the comparers to discuss those being compared, and for the careful  allocation of prestige;  it also gives the group members the opportunity to  discuss the criteria for the allocation of prestige. But it has not  replaced the other forms of social competition, and so we see them  operating side by side.


Some consequences of polyadic comparison


1. Proscription of agonistic behaviour by society. Groups practising  polyadic comparison would have an enormous advantage over groups still  limited to dyadic comparison (agonistic behaviour). Culturally they would  be at an advantage because their leaders would have those characteristics which are the criteria for the allocation of prestige, and in most human  groups these appear to be a combination of competence and dedication to the  interests of the group. Groups with such leaders should outperform groups  whose leaders were selected for power to intimidate. Genetically, the  polyadic groups would tend to have more members with qualities of  competence and unselfishness because there is a correlation in most human  groups between prestige and reproduction;  therefore we have probably  experienced a gene/culture co-evolution for competence and group loyalty.


   Among those groups practising polyadic comparison, there would be an  advantage to those groups in whom selection was entirely by polyadic  comparison, and therefore there would be an advantage in preventing  agonistic behaviour as much as possible. Therefore we can expect ritual  agonistic behaviour to be proscribed by groups, both in their childrearing  practices and in their code of behaviour for adults. In childhood there is  an enormous parental influence towards non-intimidatory behaviour, see for  instance the life histories described by Vaillant in his Adaptation to  Life, in which a cohort of American college men report severe sanctions on  aggressive behaviour during their childhoods. The proximate reason for  parents stopping their children from quarelling may well be that they find  the noise irksome, or that they consider it bad manners, or that they think  the children should spend the time improving themselves in some way;  but  the ultimate, evolutionary reason may be that they want to decide the  children's rank order themselves by the giving and withholding of praise  and criticism, and so they do not want the rank order decided by the  chidren themselves in the course of quarelling (ritual agonistic  behaviour). Also they want to develop in their children the mentality that  looks for SAHP in the form of praise rather than RHP in the form of the  submission of others, so that when they leave home they will still be  oriented towards polyadic comparison. The widespread existence of bullying  in school playgrounds (1) might seem to gainsay this thesis, but it is  probably due to the fact that there were no schools in our 'environment of  evolutionary adaptedness'. Glanz and Pearce (2) have pointed out that in  hunter/gatherer society children seldom interact with each other in the  absence of adults.

   In adult life, fighting between same-sexed adults is also proscribed.  Duelling was forbidden by monarchs, not because of the fear of loss of life  (which was slight), but because the king wanted prestige to go to people he  approved of rather than to those who were skilled with the sword or pistol. What dyadic competition is allowed between adults is governed by society's  rules rather than by nature's. Fine differences in ability can be assessed  by pitting individuals against each other in sport and in intellectual  tests. But these are polyadically controlled dyadic comparisons. Prestige  is allocated not only for performance but also for sportsmanship, and bad  marks are allocated to those who are seen to cheat or who do not accept the  decision of the referee.

   Because of this proscription, ritual agonistic behaviour is only seen in situations over which society has little control:  in prisons, on street  corners, in the school playground, in the family and in situations in which  master and servant are alone together. Also, society does not proscribe  ritual agonistic behaviour in marriage;  in fact, sayings abound to the  effect of "Never interfere between husband and wife". This may well be  because the rank order within marriage does not affect the rank order in  the group as a whole, and therefore it affects neither the choice of  leaders nor the correlation between prestige and reproduction.

   One or two contributors to this debate in ASCAP have questioned the idea  that ritual agonistic behaviour does not occur in everyday human social  interaction. It may indeed occur in subtle forms (such as damning with  faint praise in committee meetings) but I very much doubt whether the mild  forms that may occur contribute to rank order.


2. Development of latency period. Students of baboon social life have  pointed out that the brief period of immaturity before the adolescents join  the adult dominance hierarchy is a time in which they evaluate each other,  and each group of peers has worked out its rank order by the time the  canine teeth have developed. The human latency period allows a much  extended time of mutual evaluation by the peer group. It also allows the  previous generation to play an important part in the evaluation, and of  course in human life we see a whole professional class of evaluators  ranking our adolescents according to adult standards. Therefore, whereas  the accepted function of the latency period is to allow more learning, we  can add the additional function of allowing ranking according to ability to  learn and according to other attributes which are manifested at this stage  of development.


3. Religion and war as projective tests. Society wants individuals who are assertive and capable and yet have the capacity for submission of their  individual goals to those of the group. The induction of children into  religious practices allows an evaluation of this capacity for submission,  and also provides a test of memorising capacity by requiring the child to  learn scripture and ritual.

   The wars of primitive man are ritualised and the death rate is low. There is much observation of individual fighting attributes. In this way  society can allocate prestige to those who will risk their lives for the  sake of the group. This is a possible explanation for the universality of religion and war in human groups:  those groups that lacked these aids to  polyadic comparison did not survive.


4. Why people are nice. On the whole society allocates prestige to people  who are nice. Nice means that they are decent, honest, reasonable,  cooperative people who put the good of the group before their own selfish  interests;  they are also likeable and interested in their fellow human  beings. Thanks to polyadic comparison human goups have been selecting for  niceness for millions of years, and we have become very good at it. Therefore we have to some extent overcome the legacy of dyadic comparison  which is to select for intimidating, selfish bullies. The genes may be  selfish, but the people are unselfish, and it is the people we have to  interact with, not the genes. I think in this sense the message of  evolutionary biology is an encouraging one. We are nice because, for a  very long time, we have selected each other to be nice.


Hedonic dyadic comparison 


Of course in the hedonic mode there is a lot of comparison of social  attention-holding power (SAHP) on a dyadic basis. This takes two main  forms. There is furtive comparison, in which, for instance, a woman will  look round a room she enters to make sure she is the most attractive woman  present;  this involves a comparison of herself with each other woman  separately. Then there is the episode of mutual appraisal when two people  meet. This is similar to the assessment stage of ritual agonistic  behaviour;  the differences are that it is SAHP rather than RHP that is  being compared, and that evaluations of favourable relative SAHP are not  signalled in the form of catathetic signals;  in fact, politeness often  directs the dyad into an exchange of anathetic signals (compliments)  whatever the result of the evaluation. This helps to prevent a switch to  the agonic mode, because inappropriate signals of favourable relative RHP  lead to loss of face (loss of SAHP). It seems likely that dyadic  evaluations of unfavourable relative SAHP, either furtive or mutual, may  cause social anxiety or possibly even depression (fall of SAHP) but this is  a matter for future empirical study.


Use of same mechanisms


It is likely that the SAHP system of hedonic polyadic competition developed  out of the RHP ritual agonistic behaviour system, rather than starting from  scratch. Thus in the RHP system we have catathetic signals in the form of  threat and attack causing a fall in RHP which triggers a further fall in  RHP which is the core element of depression. And in the SAHP system we  have catathetic signals in the form of disapprobation causing a fall in  SAHP which triggers a further fall in SAHP which is experienced as  depression. Although the nature of the catathetic signals is different, it  seems likely that the mechanism for receiving the catathetic signals is the  same, also the mechanism that converts receipt of catathetic signals into  fall of RHP, and also the hardware that calculates whether a fall in RHP is  sufficient to trigger the depressive "devaluation" of RHP that takes the  form of depression. And in the SAHP system, as in the RHP system, the  withdrawal of an anathetic signal has the same effect as a catathetic  signal. The main difference in the SAHP system is that catathetic signals  are no longer signals of favourable relative RHP; in fact they have no  comparative component;  instead of signalling "I am better than you" they  signal "You are no good."  Likewise, anathetic signals are no longer  signals of unfavourable relative RHP, signalling, "You are better than me";   instead, they signal "You are good", without any implication as to the SAHP  of the speaker. This may be one reason why there has been such an enormous  developmemt of anathetic signalling in the polyadic system.

   SAHP and RHP are components of human self-esteem, and the evolutionary  sequence RHP --- SAHP --- self-esteem goes a long way to explaining why  there is such a wide variation in human self-esteem, why there is "global"  self-esteem rather than separate self-esteem for each characteristic, and 

why depression is associated with a global fall in self-esteem rather than  just the component which is relevant to the social situation.

   In summary, I think Paul's development of the SAHP/polyadic comparison  system has made it possible to relate the yielding hypothesis to actual  human behaviour, to integrate it with all the current psychological work on  social comparison (particularly concerning social anxiety, shame and guilt)  and to relate both the above to evolutionary biology. Keep going, Paul.

1. Tattum, D.P. (1989) Violence and aggression in schools. In Bullying in  Schools, ed D.P.Tattum & D.A.Lane. Stoke-on-Trent: Trentham Books. Pp 7-19.

2. Glantz, K & Pearce, J.K. (1989) Exiles from Eden: Psychotherapy from an  Evolutionary Perspective. London:  W.W.Norton.