ASCAP July, 1990


The yielding hypothesis of depression


Readers of ASCAP will be familiar with the hypothesis that depressive  states evolved as part of the yielding component of ritual agonistic  behaviour. The hypothesis has been outlined and examined in other  publications, here summarised:


1. One of the basic plans of vertebrate social organisation is asymmetry  between conspecifics of the same sex. This social asymmetry tends to be  manifested as hierarchical rank or territorial ownership/non-ownership, or  both. Some form of social asymmetry was predicted by Darwin's theory of  intrasexual selection;  that is, that a social process rather than natural  factors decides who in any given generation is going to reproduce.


2. The social asymmetry is usually produced by ritual agonistic behaviour,  which consists of threat, fighting, escape and submission.


3. The losing or yielding behaviours of escape and submission can usefully  be thought of as consisting of two components:  an action component and a  message component. The action component consists of not fighting back, and  of desisting from whatever the original fight was about;  the message  component consists of informing the opponent that one does not intend to  fight back. In the case of escape, the message is conveyed by the act of  fleeing;  in the case of submission, the message is conveyed partly by the  cessation or absence of fighting, and partly by the emission of specific  submissive signals.


4. In social primates yielding consists of submission more than escape,  and the very existence of the group depends on effective submission by  subordinates, who, in spite of being in full health and strength, and  sometimes in spite of having enjoyed social dominance in the past, may have  to remain for long periods in a state of behavioural inhibition,  particulalry in relation to social and sexual objectives. Therefore, 

underlying the repertory of "voluntary" submissive behaviours, there is an  involuntary form of yielding which I will call depressive yielding, and  which may be switched on whenever voluntary yielding is ineffective, thus  acting as a safety net or fail-safe mechanism to ensure that some form of  yielding occurs. The relation between depressive yielding and voluntary  yielding is similar to that between shivering or vasoconstriction on the  one hand, and putting on warm clothes or switching on the central heating  on the other hand.


5. Depressive yielding may be acute or chronic. In the chronic form,  which characterises those individuals who never achieve territory or high  social rank, and cannot accept this fate "voluntarily", it consists of  depressive neuroses or personality disorders characterised by low self- esteem, fatigue, anxiety and indeciciveness. In the acute form, which  occurs in individuals being pushed down in rank, or having attempts to rise  in rank thwarted, it takes the form which we recognise as depressive  illness.


6. The depressive yielding reaction evolved because it permitted  individuals to remain group members during periods when the social  situation was unfavourable to them. Those who did not have the capacity  for depressive yielding were killed or driven from the group.



7. Ritual agonistic behaviour is so widespread in the vertebrate sub- phylum that its underlying mechanisms may well be homologous, in which case  the depressive state in humans may share the same neurochemistry as  yielding reactions in animals such as the defeat state described by Henry  et al. (1986).


8. The implication for therapy is that we should look for relationships in  which the patient is losing but is not yielding voluntarily. He or she  should be encouraged to settle the cause of the conflict by peaceful means;  or, if that is not possible, to bring the matter to a conclusion, which he  might do in a number of ways, such as by conquering, by submitting or by  escaping from the relationship. This approach is not included among  current psychotherapeutic models of depression. Of course, we must bear in  mind that our patients are men and not monkeys, and that the conflict may  be in a symbolic rather than a personal relationship. C.S.Lewis in The  Problem of Pain proposed that the function of depression is to enable man  to submit to God, and to take the difficult step of abandoning the  attitude: "My will, not thine, be done."  Many avenues of human "salvation"  such as cult membership are based on voluntary submission, often in extreme  form (Gallanter, 1989).


9. Like other evolutionary hypotheses, and like the theory of evolution by  natural selection itself, the yielding theory of depression is not  refutable and is therefore not strictly scientific in the Popperian sense. However, the hypothesis has heuristic value in two other senses:  first, it  draws attention to relationships which might otherwise be overlooked;  and,  secondly, it gives rise to directly testable hypotheses.

   Both these features of the hypothesis can be illustrated by the case of  research on the expression of hostility in depression. The hypothesis draws attention to the relative rank of the patient and the object of the  hostility as a potentially important variable in this field, a variable  which has not been considered in any of the many published studies on the  expression of hostility in depression (Price, 1968). If depression is seen  in the context of social ranking, we become aware of the fact that for any  group-living primate, there is all the diference in the world between an  act of hostility expressed up the hierarchy and a similar act expressed  down the hierarchy;  therefore a patient who expresses hostility to his  boss cannot be considered in the same category as a patient who is hostile  to a subordinate or to his child. It was also possible to derive a  prediction about the expression of hostility in depression:  that, if  depression occurred in a partner to a marriage which had developed  complementarity along the dominance/submission dimension, the expression of  hostility to the spouse would be increased if the depression occurred in  the dominant spouse but reduced if the depression occurred in the  subordinate spouse (Price, 1968).


Problems with the hypothesis


Although the hypothesis has face value in that it accounts well for the  "giving in and giving up" mental state of depressed patients, there are  several problems which have prevented the hypothesis from influencing our  thinking about depressive states and so from assisting in obtaining funds  for research into them. These problems, it could be argued, are due to the  very different social environment in which we now live, compared to the  environment in which depressive yielding evolved and became integrated with 

other hominid behaviours. Rank order between two human beings of the same  sex is seldom determined by ritual agonistic behaviour (except in places  where society has little control such as the street corner and the school  playground) but rather it is determined by other group members external to  the dyad. This means that it is unusual for depressive reactions to be  directly due to losing in ritual agonistic encounters (except within the  nuclear family, such as between husband and wife). This may account for the  objection that whereas depression is more common in women, ritual agonistic  behaviour has always been regarded as an essentially male phenomenon, and  was actually classified by Moyer as "inter-male aggression".

   Another objection is that depression tends to be triggered by "exit  events" whereas the yielding hypothesis predicts that it should be followed  by "entry events" (someone new to yield to), but since rank order in man is  so dependent on patronage and alliances, the loss of a loved one is more  likely than the arrival of a potential competitor to result in loss of  rank.

   One other serious objection, to which I want to devote this essay, is  that depressed patients sometimes appear not to be submissive, and even to  use their depression to get their own way.


The depressed patient as yielder


Most writers agree that the basic attitude of the depressed patient is one  of giving up and giving in. The patient feels inferior and is inclined to  self-effacement. He is not likely to proclaim his objectives, still less  to achieve them. Beck (1976) puts it well:


   "The term "loser" captures the flavor of the depressive's appraisal of  himself and his experience. He agonises over the notion that he has  experienced significant losses, such as his friends, his health, his prized  possessions. He also regards himself as a "loser" in the colloquial sense: 

He is a misfit - an inferior and an inadequate being who is unable to meet  his responsibilities and attain his goals. If he undertakes a project or  seeks some gratification, he expects to be defeated or disappointed. He  finds no respite during sleep. He has repetitive dreams in which he is a  misfit, a failure." (p106).


Not only does he not get his own way in the present, he has no anticipation  of getting his own way in the future:


   "The predictions of depressed patients tend to be overgeneralised and  extreme. Since the patients regard the future as an extension of the  present, they expect a deprivation or defeat to continue permanently. If a  patient feels miserable now, it means he will always feel miserable. The  absolute, global pessimism is expressed in such statements as "things won't  ever work out for me"; "life is meaningless....It's never going to be any  different." The depressed patient judges that, since he cannot achieve a  major goal now, he never will. He cannot see the possibility of  substituting other rewarding goals. Moreover, if a problem appears  insoluble now, he assumes he will never be able to find a way of working it  out or somehow bypassing it." (p117).


Thus, whether or not his objectives conflict with those of others, the  depressed patient is not in a mood to realise them, and we could say that his mental state was ideally suited to a strategy of not getting his own  way.


The depressed patient as non-yielder


However, various views have been expressed which conflict with the idea  that depressed patients accomodate themselves passively to the wishes of  others. Freud (1917), in his classic paper Mourning and Melancholia, wrote  of his depressed patients:


"They are far from evincing towards those around them the attitude of  humility and submission that alone would befit such worthless persons;  on  the contrary, they give a great deal of trouble, perpetually taking offence  and behaving as if they had been treated with great injustice." 


Fenichel (1983) wrote:


"The depressed patient, who seemingly is so extremely submissive, is  actually often successful in dominating his entire environment." (p116).


Bibring (1953) wrote:


"It is hardly necessary to discuss the conscious and unconscious secondary  gains which many patients derive from a depression. This may proceed on  the external as well as internal level. By demonstrating their sufferings  they try to obtain the "narcissistic supplies" which they need, or they may  exploit the depression for the justification of the various aggressive  impulses towards external objects, thus closing the vicious circle." (p46).


   In his reconsideration of Szasz's Myth of Mental Illness  Birtchnell  (1989) raised the issue of whether or not psychiatric patients get their  own way. Szasz used words like "domination" and "coercion" to describe  their behaviour. Hooper et al. (1959), commenting on their intensive study  of marital interaction, wrote that "it is possible to see the whole  depressive stance as a massive attempt to exercise control over the marital  relationship".

   In the clinic and the psychiatric ward it is not uncommon to encounter  instances in which depressed patients do not yield as easily as one might expect. Sometimes, in clinging to depressive but grandiose delusions, such  as that they are responsible for all the suffering in the world, they  refuse to be persuaded by the opinions of others. Some "stubbornly"  decline to be treated with ECT or even with drugs. Nurses often regard the  depressed appearance presented at ward rounds as "put on" for the sake of  the doctors in order to avoid discharge or some other undesired outcome.

   Relatives may also regard depression as a form of manipulation. Recently I treated a couple who used to spend alternate weekends with their  respective parents. The wife's depression tended to be worse when the time  came to visit the husband's parents;  she was too depressed to prepare  herself and too anxious to make the journey so that often they cancelled  the visit. She enjoyed visiting her own parents and visits there were  seldom cancelled. She was getting her own way in visiting her own parents  and avoiding visiting her husband's parents. In another case the husband  wanted to move house to be nearer his work, but the wife wanted to stay  where she was;  in the end they stayed, because the husband felt that to  move away from her few friends would make his wife even more depressed.



Do depressed patients wield paradoxical power?


Recently Michael McGuire made some relevant comments. Introducing a  session entitled "Evolution of mood and anxiety" at the 1988 Ann Arbor  conference on psychiatry and psychology from the point of view of  evolutionary biology, he raised some important issues which he exhorted the  participants to bear in mind, of which:


"The second issue deals with affect as a form of manipulation. While this  may be minimally apparent among those persons considered to be "normal",  the use of affect as a manipulation "tool" often seems apparent among  persons suffering from psychiatric disorders. To the degree that we  encounter persons who are "effective affect manipulators" -- that is, they  get us to change our behaviour -- we also encounter a potential paradox. Effective affective manipulation presupposes that the manipulator is  capable of accurately "reading" how others will respond to particular  affects. This possibility is not easily reconciled with what we assume  when we encounter persons with psychiatric disorders".


McGuire is making the point that to use one's depression to manipulate  others requires social skill, whereas lack of social skill is part of the  impairment of function which characterises the deregulation of  psychophysiological systems in depression (McGuire 1988).

   Ethological studies of depressed patients have concentrated on speech  and non-verbal communication and they have not addressed themselves to the  outcome of conflict or the reaching of contested decisions. The simple  question "Do depressed patients get their own way by virtue of being  depressed?" cannot be answered from available data, and therefore it seems  excusable to make a foray into theory and see if the problem can be  reformulated in such a way that the paradox described above is resolved.


The agonic and hedonic modes


It is useful to distinguish between situations in which a group or dyad is  oriented towards agonistic behaviour on the one hand, and those situations  on the other hand in which it is oriented towards sexual, nurturant or  affiliative behaviour or to the performance of some task. In a number of  publications, recently summarised, Chance (1988) introduced the concept of  the agonic mode for a group which was oriented towards fighting (even if no  fighting was taking place) and he pointed out that this mode affects a  number of behaviours of members of the group, such as cognition (especially  the type of interpersonal evaluation which is undertaken), attention,  physiology and muscular tension;  if this mode is prolonged the members are  liable to stress disorders and depression. In the hedonic mode, on the  other hand, there is no orientation towards fighting;  the members are  relaxed and their attention is free to undertake learning, tasks and  constructive thought. Chance gave the rhesus monkey as an example of a  species which operates in the agonic mode, and the chimpanzee as one which  favours the hedonic mode (because fights are followed rapidly by  reconciliation). It is clear that man combines both agonic and hedonic  tendencies, and that a human group or dyad can switch rapidly from one mode  to another. An act of behaviour does not necessarily have the same meaning  in the two modes; for instance, another person escaping from oneself may be  rewarding in the agonic mode but aversive in the hedonic mode.


                   (script part 2)


   According to the yielding hypothesis, depression as yielding behaviour  evolved in relation to agonistic behaviour, and therefore it makes no  predictions about who gets their own way in relation to sexual behaviour,  nurturing behaviour, affiliative behaviour, etc.

   Of course, the agenda of the agonic mode is very conspicuously concerned  with who gets their own way. Fights for territory and dominance rank are  dedicated to setting precedents and rules which determine who gets their  way in the future, not only in matters of rank but in "hedonic" matters  such as feeding and sleeping. Moreover, when a conflict of interest arises  between two individuals in the hedonic mode, the dyad may switch to the  agonic mode to sort it out, a process which Heard and Lake (1987) called  "dissuagement". On the other hand, there are other ways of resolving  conflict in the hedonic mode, even in animals. If one monkey wants a  banana that another is about to take, he may threaten it (agonic), but he  may also beg for it (hedonic) or he may distract the other's attention  (such as by giving the alarm call) and filch it (also hedonic). In human  beings the art of persuasion is so highly developed that conflict within  groups is seldom settled in the agonic mode (except within the nuclear  family).


Signals as manipulation


Almost all signals can be seen as an attempt on the part of the sender to  manipulate the receiver (Krebs & Dawkins, 1984). This is particularly true  of agonistic behaviour. When a dominant monkey stares threateningly at a  subordinate, the latter is usually being manipulated into not doing  something that it wants to do, like mating, eating or occupying a certain  place. It is less obvious that a subordinate monkey is manipulating the  dominant when it gives a submissive signal;  the manipulation is in a very  limited area, serving only to force the dominant to stop attacking.

Submissive signals cannot force the dominant to yield rank or territory. It is interesting to note that submissive signals may be manipulative  whether they are rewarding or aversive. Most submissive signals are  probably rewarding to the recipient, but the chimpanzee has developed the  infant's distress call into its submissive repertory, and this call is  appears to be so aversive to adults that they are quick to acknowledge the  submission by giving signals of reassurance (Goodall, 1986).

   In the domain of nurturing behaviour signals are also manipulative. The  gaping mouth of the young cuckoo drives its foster-parents into a veritable  slavery of food collection. The human baby's cry forces the parent to take  some action, hopefully to feed or change it, sometimes to batter it. Much  childish behaviour is designed to coerce the parent into protecting,  feeding or in some way caring for the child. The same applies to sexual  and affiliative behaviour.

   From the above argument, we would predict that depressed patients would  fail to get their own way in the agonic mode, when by definition the dyad  or group is oriented towards agonistic behaviour (Chance, 1988);  but would  get their way as much or as little as anyone else in the hedonic mode, when  the group is oriented towards non-agonistic behaviour such as nurturant,  sexual or affiliative behaviour. This does not quite answer all the  objections, because there is an impression that depressed patients get  their own way more than others, and that they actually use the depression  to get their way.

   I think the answer to this is that  submission tends to be expressed as  a metaphor, and the metaphor may involve non-agonistic forms of behaviour. If we want to convey the message "I am weaker than you", human beings can 

simply say it, but without language it is quite a difficult message to get  across. It must be coded in symbolic form, and it is not surprising if  evolution has used symbols which are correlated with weakness;  and in  order to find suitable symbols it has had to trespass on the hedonic mode  and borrow metaphors based on childish behaviour or female behaviour. The  message then transmitted in the agonic mode is, "I am just like a child (or  a woman) and not worth bothering about;  ignore me, I am no threat to you."   But, at the same time, the message may retain its hedonic meaning which is  "Pay attention to me, I am a child in need of nurturing (or a woman in need  of loving)" and this hedonic message may be responded to as well.


                   Metaphors of submission


Let us first consider metaphors involving nurturance-eliciting behaviour. In most animals agonistic behaviour and nurturance-eliciting behaviour are  quite separate, reflecting the fact that they are parts of separate  biological systems, performing different functions, and presumably  subserved by different brain mechanisms (Gardner, 1988; Gilbert, 1989). But in the wolf (and in the rat) there is overlapping. In the wolf, the  nurturance-eliciting behaviour of the puppy has become ritualised as a  submissive signal. The defeated and submissive adult wolf rolls over on  its back and exposes its belly to the dominant animal, in the way that the  puppy presents its perineal area to the parent asking for it to be licked  clean (Eibl-Eibesfeldt, 1970). The recipient of this signal stops  attacking, but it does not do any cleaning:  the exchange of signals has  been sufficiently ritualised for the nurturant response of cleaning to be  dropped from the sequence.


The sickness metaphor


During human evolution the nurturing of sick adults has become part of our 

instinctive behaviour, possibly derived from the nurturing of children. Unlike the wolf, we do not respond favourably to adults behaving like  children. But we do respond protectively to adults who are sick. It was  Alfred Wallace, who with Darwin was the joint proposer of the theory of  natural selection, who first suggested in 1864 that care of the sick might  have been part of our evolved repertory of behaviour. Richards (1987,  p.163), a historian of science, describes how Wallace believed that  evolution of human physical characteristics ceased, perhaps as early as the  Miocene, to be replaced by the evolution of moral sentiments and reason:


  "He thought that once natural selection began to foster in those protomen reason and sympathetic feelings (especially moral sentiments which led them to care for their unfit brethren), selective pressures on their physical structures would cease." (italics added)


   It seems likely that whereas childish behaviour has become ritualised as  a metaphor of submission in the wolf, "sick role" behaviour (Pilowsky,  1988) has become ritualised as a metaphor of submission in man. The basic  vertebrate depressive yielding reaction has come in our species to be  expressed, sometimes, in the metaphor of "sick role" behaviour (Price,  1988b). Therefore it is not surprising that many depressed patients both  present themselves as, and feel themselves to be, physically ill. Sometimes the metaphor is so pronounced that depression appears secondary  and a diagnosis of hypochondriasis or somatisation disorder seems  appropriate.


   In the agonic (coercive) mode (Chance, 1988) depressive behaviour  conveys the message "I am no threat to you" and as a manipulation it has  the limited effect of stopping the attacks of conspecifics. In the hedonic  (affiliative) mode, and in the context of a caring relationship, it elicits  the same response as other forms of sickness behaviour;  namely,  nurturance. It may be that the hysterical patients discussed by Szasz have  the capacity to adopt the sick role by means of conversion symptoms,  possibly before the pressures of life get intense enough to make them  depressed. In either case their principle message is, "Stop attacking me,  I am out of action". Of course, like any other sick people they can  exaggerate their signals and become tyrants of the sick room;  although in  that case their empire is very limited, and does not lead to the  acquisition of rank or territory in the main social arena;  they are only  getting their way in the hedonic mode, not in the agonic mode in which rank  and territory are determined.

   If Wallace was right and we have an evolved tendency to nurture sick  relatives, it may well be that this tendency is only elicited by apparent  physical illness, whereas incapacitating emotional states such as  depression are perceived in other terms such as laziness or even rebellion  (in the form of failure to carry out allotted tasks). To think that one's  submissive signals were being interpreted as rebellion would be very  anxiety-provoking, and this anxiety may underly the strong motivation of  patients with, for instance, myalgic encephalomyelitis, to be categorised  as physically ill. Probably it is only in our sophisticated,  psychiatrically oriented Western society that depressed patients present  themselves as emotionally ill;  in other cultures the somatic symptoms of  depression tend to dominate the clinical picture, and depressed patients are perceived, not only by themselves and by relatives, but also by their  doctors, as physically ill.


Other metaphors of submission


Both the wolf and man have evolved metaphors of submission using care- eliciting behaviour;  in one case the submissive individual presents  himself as a child/puppy, eliciting parental care, in the other he presents  himself as a sick person, eliciting the instinctive care which in man is  given to sick relations. Perhaps this connection between submission and  care-eliciting behaviour may be set in perspective by considering a  metaphor of submission which has evolved in monkeys such as macaques and  baboons and which is not related to nurturance. Whereas the submissive  wolf says, "I am like a puppy to your adult", the submissive monkey says,  "I am like a female to your male."  He (or she) adopts the female form of  sexual presentation, following which the dominant monkey gives a brief  ritualised version of the male sexual response (mounting). Here the  submissive metaphor elicits sexual rather than nurturant behaviour, but as  with nurturance, we could say that the subordinate monkey is coercing the  dominant monkey to switch from the agonic (coercive) mode to the hedonic  (affiliative) mode. Some monkeys may use a dermal metaphor of submission,  conveying the message, "I am pale to your bright" (Price, 1989b).

   Another metaphor of submission uses the "vertical dimension" of  directiveness/receptiveness (Birtchnell, 1987). This occurs in man, monkey  and wolf, and is probably widespread among vertebrates. The submissive  individual says, "I am like a small person to your big person."  Various  forms of crawling, crouching, cringing and prostration express this  metaphor. This metaphor is not drawn from another category of behaviour,  and therefore there is no specific response by the dominant individual, equivalent to the mounting or the nurturance elicited by the other  metaphors. Therefore in using this metaphor the submissive individual does  not appear to be coercive, except in so far as the other is coerced into  stopping his attacks.

   Having considered some (possibly not all) of the submissive metaphors  which have evolved in vertebrates, it might be instructive to consider one  which has not evolved and the reason for it. No human beings use the  metaphor "I am an animal to your human being". Nor, to my knowledge, does  any other species use such a metaphor. One could imagine, for instance,  among the Canidae, the submissive individual using the metaphor of another,  "inferior", species, such as "I am like a lamb to your wolf."  This has not  evolved because submission is a component of agonistic behaviour which is  ritualised, whereas relations between species are typically unritualised,  especially when one species is the prey of the other. Therefore to use  this metaphor would be to risk losing the protection from serious injury  which is the adaptive value of ritualisation in the case of agonistic  behaviour. In contrast to the lack of animal metaphors in submissive  behaviour, they are frequent in aggressive (catathetic) behaviour, and it  is common for insults to take the form of worm, louse, rat, bitch, etc. They are also used for group aggrandisment;  for instance, the Christian  Dorze tribe has a belief that leopards are Christians whereas hyenas are  pagan, and this supports their belief that the Dorze are fine people who  eat fresh meat, whereas their pagan neighbours are scavengers like hyenas. Maynard Smith (1988), who quotes Sperber's work on the Dorze, comments  wrily that the belief does not go so far that the Dorze shepherds stop  guarding their flocks agains leopards on fast days.

   Culturally determined submissive signals are recognisable because they mean different things in different cultures. In England if I put out my  tongue at you it is an aggressive, putting-down, catathetic signal (Price,  1988). If a Tibetan puts out his tongue at someone it is a submissive,  boosting, anathetic signal. This can cause serious misunderstanding in  social relations. Another metaphor which one can sometimes detect is that  which says "I am invisible to your visible" or "I am absent to your  present";  this is a powerful metaphor because it carries the implication  of "I am in the psychological position of already having fled from your  presence."  This metaphor has been examined in a novel by Christopher  Priest (The Glamour, Sphere, 1985). The fact that the "absent" metaphor  has not evolved in animals may be due to the limited communicational  technology available to the evolutionary process. In man it is seen in  various forms of self-effacing behaviour. Superficially it might appear to  be an extreme example of the diminutive metaphor, saying, "I have become so  small that I have disappeared altogether", but this is logically different  from saying "I am so afraid of you that I have already run away from you."   The nearest that animals can get to this is actually to run away and then  come back using some other form of submissive metaphor.

   There is one important difference between the sick role metaphor and the  metaphors using female, childish or diminutive behaviour. The latter occur  at quite a "high" level of the nervous system, so that we might say that  the actors in these cases "know" they are submitting, and possibly have  some voluntary control over their actions. It is unlikely that the  subordinate male monkey thinks it is female, or the defeated wolf thinks it  is a puppy. But the sick role metaphor occurs at a "low" level of the  nervous system;  it is probably a modification of the basic vertebrate 

yielding reaction, one of agonistic and sexual inhibition, controlled in  the reptilian brain (MacLean, 1985). Therefore the individual using the  sick role metaphor does not realise he is submitting. He thinks that he is sick.

   I am not saying that the sick role metaphor is the only submissive  signal used by depressed patients. Even the wolf has more than one  submissive signal:  in addition to the puppy metaphor, and the vertical  dimension (crawling on its belly), it has a submissive signal in Darwin's  category of "antithesis" in which it exposes its vulnerable neck to the  fangs of its enemy. Similarly, the submissive message of depressed  patients may take many forms. Some use a metaphor taken from commerce, "I  am poor, even bankrupt, compared to your riches;  there is no need to  mobilise your commercial battalions against me."  Some use expressions  derived from religious behaviour, "I am sinful, compared to your  saintliness, and I am not competing with you for a place in heaven."   Depressed women may use a metaphor derived from the fashion world, "I am  plain compared to your beauty;  I am not competing with you for the  approbation of men" (Sloman et al., 1982). However, the sick role metaphor  is common, and accounts for the frequent physical symptoms of depressed  patients, the presence of somatic delusions and hypochndriacal  preoccupations, and the fact that so many depressed patients consult  doctors.

   Patients often describe their depression to doctors in terms of  metaphor. Metaphors and similes of being dead, a hibernating animal, a  defeated boxer, a nonentity and suchlike are common. A recent patient  expressed herself as a flat tyre or deflated balloon:  "Normally you get  kicked and you bounce back again, but I've been kicked too many times in  the same place, there's a hole there and all the air's gone out."  These  statements are quite different from the submissive metaphors described  above, in which the patient acts out the metaphorical part. Different  again are the submissive metaphors used in verbal flattery; e.g., "I am a  candle flame, you are the sun."


   In conclusion, the fact that depressed patients may get their own way in  the hedonic mode does not negate a hypothesis which states that depression  evolved as a losing strategy in agonic interactions. The very fact that  they use a sick role metaphor to express their submission ensures that they  obtain the care and consideration which is normally given to physically ill  loved ones. Depressed patients are sometimes said to be attention-seeking,  and it is interesting to note the opposite message about attention which is  given in the two modes. In the agonic mode the message is, "Pay no  attention to me, I am sick and therefore no threat to you."  In the hedonic  mode the message is, "Pay attention to me, I am sick and require  nurturing."





I am most grateful to colleagues who have taken an interest in the yielding  hypothesis of depression and who have discussed these ideas with me,  particularly Michael Chance, Leon Sloman, Russell Gardner, Jay Feierman,  Paul Gilbert and John Birtchnell.