British Journal of Medical Psychology, 71, 465-477, 1998

The adaptive function of mood change

by John Price

Summary

It is useful to imagine an "agonistic strategy set", containing the two alternative and mutually incompatible strategies of escalation and de-escalation. The strategy set is accessed by loss, threat, or some other form of "ranking stress". It is further suggested that ranking stress is dealt with relatively independently at three levels of the brain/mind, so that an agonistic strategy set is deployed at each of the three levels. Normally, all three levels escalate or de-escalate in unison, but sometimes lower level de-escalation is associated with middle or higher level escalation, and then the resolution of agonistic situations is delayed and psychopathology may be recognised.

Depression is based on a de-escalating strategy

In previous papers, it has been suggested that the adaptive value of depressive and manic states might be concerned with the transfer of power between individuals in a dyad or group (Price et al., 1994; Price and Gardner, 1995). Elevated mood gives the psychological resources needed for the exercise of power: self-confidence, energy, optimism and sociability. Depressed mood helps to accommodate the individual to loss or absence of power and the honours that go with it: lack of confidence, loss of interest and unsociability enable the individual to withdraw from the arena and to allow others to take control; in particular, depressive thinking about previous achievements erases the conception that there might be anything to be regained by further struggle.

In order to operationalise our connection between depression and loss (or lack) of control, we must relate it to the normal behaviour of losing. However, there is no normal psychology of losing. Perhaps, in a world of academic "winners", it has not been thought politic to study the mechanics of losing.

Fortunately, there are two areas of scientific enquiry which relate to our needs. One is the ethological study of fighting (ritual agonistic) behaviour. The other is the mathematical analysis of pair-wise contests in terms of game theory. Both of these fields have developed the concepts of escalating and de-escalating strategies.

In comparative ethology (Alcock, 1979), a fight is referred to as an agonistic encounter (or ritual agonistic encounter, to emphasise the symbolic aspect of the behaviour). For the fight to have served its function, there must be a winner and loser. Each contestant would like to be the winner, and in this they have different objectives; but they share the objective of getting the fight over quickly, because fighting is wasteful of time and effort, and makes animals conspicuous to predators. In one of the two contestants, this second desire to get the fight over takes precedence over the desire to win, and that individual then adopts, or switches to, a de-escalating strategy, while the winner can be said to adopt, or maintain, an escalating strategy, The escalating strategy increases the chances of winning the fight at the expense of increasing the potential costs of losing. It is manifested by the individual continuing to take part in the contest, and at the same time possibly threatening or attacking the opponent, or switching from a less intense to a more intense form of ritual (e.g., in stags, changing from parallel walking to horn-locking). The de-escalating strategy reduces the chances of winning and at the same time reduces the potential costs of losing. It is manifested by withdrawal from the offensive position or posture, and possibly by one or more acts of submission.

The components of the de-escalating strategy can be predicted from the fact that it is an alternative to being dead; being the outcome of ritual agonistic behaviour, it is a ritual or symbolic form of death. The individual adopting the de-escalating strategy feels "dead" in that he is in a subjective state of demobilisation and incapacity; he feels too "dead" to continue fighting, and therefore stops fighting. He also looks "dead" to his opponent, who therefore treats him as "dead" and as no threat and also stops fighting, in that one does not continue to fight a "dead" opponent.

Escalation and de-escalation are the two basic "moves" in the game theory analysis of fighting behaviour (Krebs and Davies, 1993; Archer and Huntingford, 1994). A "hawk" escalates. A "dove" de-escalates. A "bourgeois" escalates on his own territory and de-escalates on foreign soil. If a dove meets a hawk, the hawk wins. If two hawks meet, there is a terrible cost paid by both. By equating the payoff in the game to fitness in an evolutionary sense, it has been possible to determine which strategies are likely to be "evolutionarily stable" in the sense that they are resistant to infiltration by alternative strategies. A major finding has been that a mixture of hawk and dove strategies tends to be stable, and it does not matter whether the population consists of both hawks and doves, or whether each individual plays hawk and dove on different occasions, in a ratio of hawk/dove which is constant for that individual, but determined randomly on each specific occasion. Thus each individual's hawk/dove ratio becomes an important heritable character.

Apart from the difference between hawk and dove, and whether the conflict is a home or away match, there are only two variables which affect the choice between an escalating and de-escalating strategy (Krebs and Davies, 1993; Hack, 1997). One is resource-holding potential (RHP), which is an individual's estimate of its own fighting capacity (and enables it to compare itself with the estimated RHP of a rival), and the other is Resource Value, which is an estimate the individual makes about how valuable, or worth fighting for, the object of any conflict is. Clearly, the more likely an individual thinks he is to win, the more likely he is to escalate. And the more valuable the resource he is fighting over, the more likely he is to escalate. Like ownership of territory, RHP and Resource Value tend to over-ride the basic hawkishness or dovishness of an individual, so that a small "hawk" hamadryas baboon will give way (de-escalate) to a large one, unless he is defending a member of his own harem against an outsider, when he is likely to escalate. It is when RHP and Resource Value are equal that the hawk\dove difference determines the decision to escalate or de-escalate.

The problem of "active" submission

The component of losing which the ritually dead person does not manifest is the active process of submission. Most species have a ritual of submissive behaviour which requires some active performance of social skill. In our own species, a speech of submission may be elaborate, including flattery and promises of obedience in the future. This is not appropriate to the ritually dead person, and it is not the sort of behaviour which is seen in depressed patients. Depressed patients do not show active submission, nor is their depression oriented towards one or more dominant people or "winners". In this sense, they do not behave like "losers". We therefore came to the conclusion that there are at least two components or "levels" of submission - an active or "voluntary" level and a passive or involuntary level. We therefore emphasised that our theory relates depression only to passive, involuntary submission - what we have called the "Involuntary Subordinate Strategy (ISS)" (Price et al., 1994; Sloman et al., 1994).

Depression is based on a lower level de-escalating strategy

It was at this stage in our thinking that we encountered Paul MacLean's theory of the triune brain (MacLean, 1985, 1990). MacLean suggested that as our forebrains grew over the past 300 million years, the new parts did not replace the old parts, but became grafted on to them, modifying them and supplementing them; and also, that this process of modification was not continuous, but rather occurred in three main stages, leaving us with three "central processing assemblies" in which information is co-ordinated and decisions are taken in response to changes in the environment. He suggested that these three assemblies influence each other, but are relatively independent.

This theory presented us with the challenging possibility that a choice between escalating and de-escalating strategies is made not at one level in the brain, but at three separate levels, possibly at the same time. And it included the possibility that the choice made at each level might not be the same, so that one might have a brain which is escalating at one level and de-escalating at another.

(Table 1 about here)

MacLean's lowest level forebrain assembly - what he calls the reptilian brain or R-complex - is situated in the basal ganglia (mainly the corpus striatum). It deals with problems encountered during the stage of existence we shared with the ancestors of present-day reptiles. These were mainly ritual agonistic behaviour and courtship. Our reptilian ancestors needed a de-escalating strategy, and so it seems appropriate to allocate to this level the normal passive subordinate behaviour which we described as the ISS. The ISS thus is renamed as the reptilian de-escalating strategy.

MacLean's middle assembly - the paleomammalian brain - is situated in the limbic system, which is well known to be concerned with emotional behaviour. How does a choice of escalating or de-escalating strategy manifest itself at this level? Clearly the escalating strategy must be concerned with the expression of anger. Anger is the emotion of attack, which is the behavioural manifestation of escalation. And correspondingly, the de-escalating strategy at this level must involve the dysphoric emotions such as feeling chastened, social anxiety, shame, guilt and humiliation. Also included here is depressed emotion which differs from depressed mood in being focused on an object (Davidson and Ekman, 1994), and in being reactive to the changing circumstances of the object (whereas depressed mood is unfocused or self-focused).

At the level of the highest assembly, in the neomammalian brain, situated in the neocortex, we have a rational, conscious, voluntary choice between escalation and de-escalation. All the complex information available to consciousness is used at this level of decision-making. The outcome is either a decision to continue the struggle and fight on; or a decision to back off, escape, climb down, apologise, eat humble pie, admit one was in the wrong, and/or to undertake a commitment to carry out any reparations demanded by the winner.

MacLean's work has been criticised (Butler and Hodos, 1996), although it is not clear if these criticisms invalidate his concept of three central processing assemblies at different brain levels. In any case, we are not tied to a model of the triune brain. A model of the trine mind is quite sufficient, and this model has been around since Plato in The Republic suggested a "division of the soul into three parts, corresponding to the three orders in the state. Each part seems to me to have its own form of pleasure and its particular desire; and any one of the three may govern the soul".

Early Eastern scholars likened the body to a cart, whose progress was directed by a Master (who knows where he wants to go) and guided by a Driver (who knows how to get there) and drawn by a horse, who provides the energy. And subsequent thinkers have also speculated about levels of the mind to account for the different degrees of control and awareness that attend the operation of different human functions. I would argue that the triune brain not only fits in with what we know about the triune mind, but helps to clarify thinking about this complex field, beset as it is by problems of mind/body dualism, and by the use of technical language (such as "conscious" and "voluntary") which derives from language used by actors in the drama to ascribe qualities to each other, rather than from language appropriate to a scientific observer.

(Table 2 about here)

Eight combinations of escalation and de-escalation are possible from the three independent levels, and if we are concerned with depression, we are interested in the four which include lower level de-escalation. Before looking at the clinical implications, it might be useful to give an example from a simpler system, which will illustrate the general principle involved and the possible relations between the levels.

The analogy of dealing with cold

If there is a problem of cold, the higher assembly tackles it by putting on more clothes or turning on the central heating, while at a lower level a decision is made about whether to shiver or to vasoconstrict the vessels in the skin. In the case of cold, we have two main levels, and it is clear that different information is used in decision-making at the two levels. For instance, information about the integrity of the central heating system is used in deciding whether to put on more clothes, but is not used in the decision to shiver. Likewise, information about core body temperature is probably used in the decision to shiver, but is not used in the decision to turn on the central heating. Both levels are essentially making a decision about whether to increase heat production (or reduce heat loss), and usually they will make the same decision, so that the individual exposed to cold turns on the central heating as well as shivering. But in some cases, they may make different decisions. In someone who is too proud to turn on the central heating (regarding himself as hardy) the decision at the higher level may conflict with that at the lower level, which decides to shiver. Note that the higher level can pre-empt or terminate the decision at the lower level: turning on the central heating may pre-empt or terminate shivering. This represents the evolutionary tendency for the higher level to take over decision-making from the lower level. The lower level remains a failsafe mechanism in case the higher levels do not seem to be coping adequately with the problem.

Ranking stress

In the case of agonistic behaviour, I will use the term ranking stress to indicate the perception of a situation which is a threat to current rank. This could take the form of the appearance of a rival, agonistic behaviour on the part of a rival, or the disappearance or defection of an ally. Ranking stress causes the individual to "access" their "agonistic strategy sets", which, of course, consist, at each level, of the two strategies of escalation and de-escalation. We do not know whether the same types of ranking stress are effective at each level - this is a matter for empirical study. Nor do we know the precise information used at each level to decide between escalation and de-escalation. Nor do we know whether there is any overlap between the information used to trigger the accessing of the strategy set and the information used to decide between the two strategies in the set.

The components of the lower level de-escalating strategy

If we try to reduce the lower level de-escalating strategy to its basic components, we get further evidence that the capacity for mood change evolved in the context of agonistic behaviour (Stevens & Price, 1996).

There are three basic hypothetical constructs which have been found essential by behavioural ecologists in making mathematical models of fighting behaviour (Krebs & Davies, 1993; Hack, 1997). These are:

1. Resource-holing potential (RHP) which is an estimate of an individual's fighting capacity, or more generally, the capacity to control the outcome of conflictual interpersonal situations. It is used in the calculation of relative RHP, which is an individual's assessment of whether he is more powerful than a competitor. With the partial change in competition (during recent hominid evolution) from agonic competition by intimidation to hedonic competition by attraction, RHP is becoming superceded by the self-concept of social attention holding power (SAHP) which refers to an individual's ability to attract attention and investment from other group members (Gilbert et al., 1995). Whereas RHP determines the balance between attack and submit, SAHP determines the balance between self-assertion and self-effacement on important social occasions.

Lowered RHP accounts for the reduced self-esteem, inferiority and other forms of negative self-perception which are common features of depressed mood; and lowered SAHP accounts for related features such as guilt, shame, social anxiety and self-blame.

2. Resource value, which is an estimate of the value of whatever is being fought over. Resource value expresses the investment of the individual in the goals and incentives which have to be given up by a yielder. These may be particular goals such as getting one's own way over some issue, or they may be more general such as ambition and the desire for social enhancement, or they may be symbolic such as the achievement of a meaningful philosophy or religion. To the extent that rewards are dependent on social status, resource value represents ambition and all forms of status seeking.

Reduction in resource value in the depressed patient accounts for the global loss of interest, apathy and reduction in reinforcer effectiveness.

3. Ownership, which is a widespread convention in the animal world by which an owner or resident wins a contest and an intruder loses it. It applies to territories and in human beings also to possessions, and in some baboons it applies to the possession of a female consort. A non-owner has to be very much more powerful to win in a contest with an owner. Ownership is a component of the evolutionarily stable Bourgeois strategy in the Hawk-Dove-Bourgeois game (Krebs and Davies, 1993).

The sense of ownership is impaired in depression, and in extreme cases the patients feel they have no right to exist, in contrast to the patient with elevated mood, for whom "the world is his oyster". One depressed golfer expressed it in the statement, "I have no standing on the course of life". Ownership gives a sense of entitlement, and in depression the sense of entitlement is impaired.

The higher these three values are, the more the individual is likely to attack rather than back off, or to be self-assertive rather than self-effacing. Therefore, we can expect them all to be reduced in the lower level de-escalating strategy, and a fortiori to be reduced when the strategy is intense or prolonged and manifests as depressive illness. The only major component of depressed mood which is not directly related to a lowering of RHP, Resource Value or Ownership is energy. But we have ascribed the conspicuous loss of energy of depressed mood to an interaction between RHP and Resource Value (Stevens & Price, 1996). RHP represents the "can" of conflict, or the ability to overcome the opposition; Resource Value represents the "will" of conflict, or the readiness to fight. When both "can" and "will" are present in sufficient measure, then it seems likely that the motivational systems of the psyche provide the energy; but if either or both of "can" or "will" is absent, then, the ingredients for successful combat being absent, the resources of energy are not provided.

The reduction in these three basic variables can account for a considerable proportion of the manifestation of depressive states.

An illustration from English country life

Let me illustrate the normal functioning of these mechanisms by a typically English example. Lady Turvey, our heroine, is married to Sir Topsy, and lives in a fine Manor House somewhere in the Shires. One day Sir Topsy is killed in a hunting accident. This is a ranking stress for Lady Turvey, who has now become The Dowager Lady Turvey. She knows that she will now be expected to move out of the Manor House to make way for her son and for her daughter-in-law,the new Lady Turvey. She herself must go to live in the Dower House, which is considerably smaller than the Manor House.

At the higher, rational, level, the Dowager has a choice over what to do. Her de-escalating strategy is to accept the inevitable and move to the Dower House. Her escalating strategy is to stay where she is. She can boulster this strategy with many justifications: her son is not yet old enough to take on the responsibility of the Manor; her daughter-in-law is not mature enough to manage so many servants; she needs to stay to supervise some work on the house recently undertaken; and so on with her fertile mind. Truly, in spite of custom, she does not want to leave the house she has lived in for so many years.

After some weeks, it becomes clear to family and friends that she is making no plans for moving. She turns a deaf ear to rational arguments about precedent. At this stage, the middle level, or emotional strategy set is accessed. The trigger for this may well be criticism of her behaviour, or a deliberate attempt to induce middle-level de-escalation by questions such as, "Aren't you ashamed to defy custom in this way?", and "Don't you feel guilty living in that huge house all on your own?" If these comments are successful, and induce a middle-level de-escalating strategy characterised by guilt and shame, the Dowager's thinking at the higher level may be affected by these emotions, and one of two things may happen: she may switch to a higher-level de-escalating strategy and move to the Dower House; or, the shame and guilt are insufficient to overcome her wish to remain where she is, and she stays on but feels bad about it. On the other hand, her middle level assembly may select the escalating strategy, and she will become angry with these interfering busybodies who presume to tell her what to do, and in the fury of her indignation she crosses them off her visiting list and vents other such punishments upon them. She is buoyed up by a feeling of self-righteousness.

After some months, we can imagine that the new Lady Turvey is losing patience with her mother-in-law, as she is unable to make definite plans for taking over the Manor House. After discussions with the family solicitor, she sends a firm of architects into the Manor House to make provisional plans for certain alterations she wants to make.

This intrusion on her territory upsets the Dowager, and we may assume that it also triggers the lower level agonistic strategy set. If she deploys the de-escalating strategy (which we have also called the ISS), she suffers a fall in RHP/SAHP, resource value and ownership. These have a profound effect upon her higher-level thinking. Her fall in RHP/SAHP, or self-concept, is likely to make her feel inadequate to carry all the responsibility of the Manor House without the support of her husband. The fall in Resource Value weakens her attachment to the Manor House, which now seems much less attractive to her, and not worth making such a fuss over. The fall in Ownership underlines her sense that it is her daughter-in-law, and not herself, who is now entitled to the Manor House. This "depressive" change in thinking may switch her higher level assembly into de-escalation. Or it may not, and she may stay on in the manor House as a depressed person, perhaps taking the role of a chronic invalid, and it may well be on health grounds that she is finally carried on a stretcher out of her Manor House to be nursed elsewhere. In such a case, her lower level de-escalation would be expressing itself through the metaphor of physical illness (Price and Gardner, 1997).

As a final possibility, the intrusion of her daughter-in-law's firm of architects may trigger, not depression, but elevation of mood. This will increase her self-concept, make the Manor House seem even more worth defending, and strengthen her sense of ownership. Then she will have increased energy to lobby family members to enlist them on her side, to interview firms of solicitors to take the matter to court if necessary, and to spread vile calumnies about her daughter-in-law, so that all the other ladies of the shires join her in condemning the indecent haste with which the daughter-in-law appears to be usurping her position. Perhaps she will be successful, and the daughter-in-law will de-escalate, and remain happy to see out her days in the gardener's cottage. But more likely the daughter-in-law is made of "sterner stuff", and will insist on her rights. This final scenario is the stuff that family feuds are made of.

Comment

For the Dowager, moving to the Dower House was the equivalent of turning on the central heating. At any stage, she could terminate the painful experience of her middle and lower level de-escalations by the simple expedient of de-escalating at the higher level. But she was too proud, or too stubborn. If one had been called on to treat her at any stage, one would not have had much success treating "depression" or its somatic equivalents; the only logical way to proceed is to ask, "Why have you not moved to the dower house?"

We can note another similarity to the "cold" analogy. In the way that the discomfort of shivering may enhance the experience of cold and thus dispose the person to switch on the heating, so the depressive thinking of middle and lower level de-escalation enhances the experience of losing and thus dispose the person to make a voluntary submission. Guilt facilitates the acceptance of punishment. Shame facilitates the acceptance of lower rank. Dejection facilitates the giving up of unreachable goals. Likewise, depressed mood facilitates the shift from anger to dejection. People get angry when they are insulted by a lower ranking person (as Aristotle pointed out); the sense of inferiority which occurs in depression helps the person to see the insulter as higher ranking, in response to which the appropriate emotion is to feel chastened. In an egalitarian society, we are angry when we feel justified and chastened when we realise we are unjustified; and depressed mood makes us feel less justified - we feel in the wrong, and therefore guilty. We are angry when someone takes something we own, and depressed mood takes away our sense of ownership, or of being entitled to things. Depressed mood also favours de-escalation at the higher level - it takes away our expectation of victory, it takes away our confidence in our own ability, it takes away our sense of the importance of whatever we are fighting about, and it makes us feel too tired and ill to want to fight on.

The flow chart then reads: on entering a conflict situation, either fight or give in. If you fight and still don't win, become angry and fight harder, or feel chastened and give in. If you fight harder and still don't win, get elevation of mood and fight even harder, or depression of mood and give in. In most cases in the EEA, this algorithm probably led to the rapid settling of disputes. But in our complex society, there are a number of causes of "blocked voluntary yielding" which prevent agonism from being functional, and disputes from being resolved.

Sometimes the block to the resolution of conflict lies not in the submitting person, but in the "winner" whose role it is to accept the submission and move on to reconciliation. Sometimes no amount of submission at any level will induce the other to exercise the "mercy" of the winner (Weisfeld, 1977). Just suppose that the daughter-in-law has an undying hatred of the Dowager Lady Turvey, possibly because she opposed the daughter-in-law's marriage to her son on the grounds of insufficient social standing. Then, with the Dowager exposed to revenge in the absence of her husband's support, it may be that however submissive the Dowager is, the new Lady Turvey will continue to punish her by giving her the cold shoulder, or humiliating her in other ways, and so the Dowager's lower level de-escalating strategy may never remit. This occurs in situations of torture and other forms of bullying. But usually, the block to resolution lies with the loser, who is too proud or stubborn to make the appropriate higher level de-escalation. In other words, in treating depression we are often dealing with "blocked higher level de-escalation".

Manifestations of the lower level de-escalating strategy

It is the lower level de-escalating strategy with which we are mainly concerned clinically, because it is this strategy which if too intense or prolonged we recognise as clinical depression. In the past we have called it by different names, such as "the yielding subroutine" (Price & Sloman, 1987; Sloman & Price, 1987) or "the involuntary subordinate strategy (ISS)" (Price et al., 1994). Whatever its name, the strategy has four main functions:

1. It signals "no threat" to the winner, and thus underlines any higher level submission that is being made; or, if higher level submission is not being made, it reassures the winner that the sufferer is not a real threat, whatever they say, because they look too old or ill or incompetent to be effective.

2. It signals "out of action" to friends and supporters, giving them the message "do not push me out into the arena to fight on your behalf" (Price & Gardner, 1995).

3. It facilitates switching from escalation to de-escalation at the middle and higher levels, by inducing guilt, shame and inferiority, and a pervasive feeling of pessimism.

4. It puts the individual into a depressive straightjacket, by taking away energy, interest, hope, decisiveness and willpower. In this way, it forces the individual to submit, even if the higher level is being too proud or stubborn (or courageous) to back off. In such a case, the socially apparent reason for submission is likely to be physical illness or infirmity (Price & Gardner, 1995).

Causes of strategy mismatch

According to our model, the cause of depressive illness is not social adversity, or losing, or failing, because it is normal for human beings to lose and fail and to be confronted with adversity. Rather, the cause of depressive illness is the failure of the triune brain to coordinate its response to social adversity. A co-ordinated response ensures either success or successful withdrawal/submission. Often one finds that higher level de-escalation is being blocked for some reason, either in the patient or by a third party. For instance:

1. Higher level de-escalation is blocked by moral scruples.

When someone tries to maintain an impossible position, or clings on to an unrealisable goal, we call it courage or stubbornness, depending on whether or not we sympathise with the attempt. Pride, honour and moral scruples are all reasons for continuing to escalate at the higher level in spite of crippling de-escalation at the lower level. Milton took on the monarchy, Darwin took on the church, and many are the martyrs who have refused to bend the knee. Health is sacrificed to principles. We have described a case in which the patient refused to join his work colleagues in dishonesty, was punished by them and suffered depressive illness as a result (Stevens and Price, 1996). Anthony Trollope provides a paradigm of self-destructive stubbornness in his novel He Knew He Was Right. In such a case the therapist is in the position of the medical attendant of a boxer who insists on fighting more powerful opponents.

2. Higher level de-escalation blocked by ignorance or misunderstanding.

Human submission (unlike animal submission) involves obedience, or actively doing what the other person wants. Sometimes this is impossible. A dominant husband may insist on an enthusiastic sexual response, a dominant wife may require that her husband give up an involuntary tic. Sometimes the patient does not know what to do to please the other: a husband did not want his wife to work because he was afraid she would meet attractive men, but he was ashamed to confess this fear, and so he criticised her for laziness and stupidity until her depression made her unfit for work.

3. Higher level de-escalation blocked by a third party.

A wife wanted to please her husband, who insisted that she be at home on Saturday; but her mother, who was even more powerful than her husband, insisted that she visit and do chores for her on Saturday. Submission to one involved resistance to the other. She wanted to de-escalate in the two most important relationships in her life, but she was not allowed to. In this case the depressive illness remitted when her mother died.

4. Middle level de-escalation blocked by a sense of injustice.

Many patients suffer insults and wrongs which they cannot let go because of resentment or anger over the injustice of it all. Recent cases of this kind in my clinic include people sacked unfairly from their jobs, a father accused of sexually assaulting his daughter, and parents who feel the education authority has failed their children in some way. These people feel beside themselves with anger, seething with murderous rage. Even if they choose to de-escalate at the higher level, they are still escalating at the middle level of emotional reaction, and this prevents the resolution of the lower-level de-escalation.

Identifying the conflict

At the higher level, there are many areas of life in which escalating and de-escalating strategies are carried out independently; for instance, one may be escalating one's love life by pursuing an affair, while de-escalating one's parental relationship by agreeing to give up smoking. At the middle level of emotional response, there is still some variety, in that one can be depressed about one issue at the same time as being angry about another - one can even be angry and depressed about the same issue. But at the lower level, the strategy is unfocused; it is an all or nothing matter, and affects all activities and strategies. Depressive mood is pervasive. How does one decide which of the higher level escalations is producing a situation which is prolonging this lower level de-escalation? This is one of the skills of the psychotherapist. Suffice it to say that it usually concerns one of the patient's salient goals or relationships.

Therapeutic considerations

The therapeutic task is to identify the conflict, and help the patient to resolve it at the higher level. Often this will take the form of helping the patient to admit defeat, accept the inevitable, and give up unrealistic aspirations. Sometimes, however, the patient can be helped to win - in spite of the fact that he or she was losing to start with, and is further incapacitated with depression. Sometimes the adversarial context can be avoided altogether, and an outcome achieved by negotiation and compromise. Sometimes patients see conflict where there is none, others are too sensitive to threats, or their lower level agonistic strategy set is too easily accessed. It is difficult to say what proportion of patients have insight into the conflict that is causing the depression. I suspect very few. Mostly patients blame the depression on biological factors such as starting or coming off "the pill", or on the symptoms of the depression itself, e.g., "I am depressed because I can't sleep". If it is necessary to get the patient to understand the real cause of the depression, the shivering analogy (see above) is useful; e.g., the therapist says, "We are not going to waste time rubbing ointment on your shivering muscles - we are going to explore the reason why you have not been able to turn on the central heating."

The symbolic or internalised rival

The model is based on conflict with a real rival. This explains why depression is so incapacitating - it incapacitates the individual while the rival takes control. This sort of incapacity has no function in dealing with impersonal stresses such as adverse climatic conditions.

However, such is the human capacity to personalise the environment that some forms of impersonal stress are perceived as coming from a rival. Job, for instance, saw the misfortunes heaped on his family to be acts of a personal God; he became indignant with God and wanted to take him to court. In this case Job became depressed - lower level de-escalation was associated with both upper and middle level escalation. The remedy was provided by Elihu, who pointed out to Job that it was not appropriate for a man to enter into an agonistic encounter with God. Job then submitted to God and his depression lifted.

In some cases the conflict is not with an outside person, but with an internalised entity (Weiss, 1993; O'Connor, 1997). For instance, some successful people are caught in a trap between their ambition and their childhood indoctrination not to outshine their immediate family (particularly daughters are taught not to outshine brothers). Then, when they become successful as adults, they become depressed for no obvious reason. It has been suggested that their depression is caused by "survivor guilt" due to the gap between their own performance and that of other family members. But to pursue conflict into the depths of the psyche is beyond the remit of the present paper.

References

Alcock, J. (1979) Animal Behavior: An Evolutionary Approach. Sunderland, Mass.: Sinauer.

Archer, J. & Huntingford, F. (1994) Game theory models and escalation of animal fights. In: Potegal, M. & Knutson, J.F. (eds) (1994) Dynamics of Aggression: Biological and Social Processes in Dyads and Groups. Hillsdale, NJ: LEA. Pp. 3-32.

Bibring, E (1953) The mechanisms of depression. In Affective Disorders ed P. Greenacre. New York: International Press. Pp 309-316.

Butler, A.B. & Hodos, W. (1996) Comparative Vertebrate Neuroanatomy. New York: Wiley-Liss.

Davidson, R.J. & Ekman, P. (1994) Afterword: how are emotions distinguished from moods, temperament and other related affective constructs. In: Eckman, P. & Davidson, R.J. (eds) The Nature of Emotion: Fundamental Questions. Pp. 94-96. New York: Oxford University Press.

Gilbert, P. (1997) The evolution of social attractiveness and it role in shame, humiliation, guilt and therapy. British Journal of Medical Psychology, 70, 113-148.

Gilbert P, Price J & Allan S (1995) Social comparison, social attractiveness and evolution: how might they be related? New Ideas in Psychology: An International Journal of Innovative Theory in Psychology, 13, 149-165.

Hack, M.A. (1997) Assessment strategies in the contests of male crickets, Acheta domesticus. Animal Behaviour, 53, 733=747.

Krebs, J.R. & Davies, N.B. (1993) An Introduction to Behavioural Ecology, 3rd edition. Oxford: Blackwell Scientific Publications.

MacLean, P.D. (1985) Evolutionary psychiatry and the triune brain. Psychological Medicine, 15, 219-221.

MacLean, P.D. (1990) The Triune Brain in Evolution. New York: Plenum Press.

O'Connor, L.E. (1997) Survivor guilt, submissive behavior and evolutionary theory: the downside of winning in social competition. Paper read at HBES meeting, Tucson, Arizona, June 1997.

Price, J.S. & Sloman, L. (1987) Depression as yielding behavior: an animal model based on Schjelderup-Ebbe's pecking order. Ethology and Sociobiology, 8, 85S-98S.

Price, J.S. & Gardner, R. (1995) The paradoxical power of the depressed patient: a problem for the ranking theory of depression. British Journal of Medical Psychology, 68, 193-206.

Price, J.S., Sloman, L., Gardner, R., Gilbert, P. & Rohde, P. (1994) The social competition hypothesis of depression. British Journal of Psychiatry, 164, 309-135.

Sloman, L., Price, J., Gilbert, P., Gardner, R. (1994) Adaptive function of depression: psychotherapeutic implications. American Journal of Psychotherapy, 148, 401-416.

Sloman, L. & Price, J.S. (1987) Losing behavior (yielding subroutine) and human depression: proximate and selective mechanisms. Ethology and Sociobiology, 8, 99(S)-109(S).