Brit.7. Psyc/ziat. (1977), 130, 421—31

The Making and Breaking of Affectional Bonds

II. Some Principles of Psychotherapy

The Fiftieth Maudsley Lecture (expanded version)

By JOHN BOWLBY

Summary. An account is given of how a clinician guided by attachment theoryapproaches the clinical conditions to which the theory is held to apply, whichinclude states of anxiety, depression and emotional detachment. Assessment of apatient is in terms of the patterns of attachment and caregiving behaviourwhich he commonly shows and of the events and situations, both recent andpast, which may have precipitated or exacerbated his symptoms. The problemsposed l)y relevant information being suppressed or falsified are noted.

Viewed in this perspective a psychotherapist is seen to have a number of inter@related tasks : (a) to provide the patient with a secure base from which he, thepatient, can explore himselfand his relationships ; (b) and (c) to examine with thepatient the ways in which he tends to construe current interpersonal relationships,including that with the therapist, and the resulting predictions he makes andactions he takes, and the extent to which some may be inappropriate; (d) to helphim consider whether his tendencies to misconstrue, and as a result to act misguidedly, can be understood by reference to the experiences he had withattachment figures during his childhood and adolescence, and perhaps may stillbe having.

In the first part ofthis Lecture I have given anoutline of attachment theory and its origins : andhave described some of the deviant pathwaysalong which a person's attachment behaviourmay develop, together with some of the typicalchildhood experiences that research suggests areresponsible for the development of these deviantpatterns and the various common psychiatric disorders towhich theycontribute. In thesecond partmy aim is to consider how this theoretical approachcan guide US, initially in assessing a patient'sproblems and subsequently in helping him.

First, we must decide whether the problempresented is one to which attachment theory isapplicable, an open issue still requiring muchexploration. If it seems applicable, we considerwhat pattern the patient's attachment behaviour typically takes, bearing in mind both what

421

he tells us about himself and the relationshipshe makes and also how he relates to us aspotential helpers. We also explore possibleprecipitating events, notably departures, seriousillness or death, and also arrivals, and thedegree to which the presenting symptoms canbe understood as recent or belated responsesto them. During the course of these explorationswe may begin to get some inkling of the patternsof interaction that obtain in his present home,which may be either his family of origin, or thenew family he has helped create, or (perhapsespecially in the case of women) both. Anyhistorical material that casts light on howcurrent patterns may have come into beingsharpens our perceptions.

j@ major difficulty in this process of assess

ment is that information given may omit vital

Thie On.

I@IllhliH@IIflIUI@III@I@U111111111111N2TI-KZW-8AFT

THE MAKING AND BREAKING OF AFFECTIONAL BONDS. II.422

facts or falsify them. Not only are relativesparents or spouse—apt to omit, suppress orfalsify but the designated patient may do so aswell. This, of course, is no accident. First, it isevident that many parents, who for one reasonor another have neglected or rejected a child,have threatened him with abandonment, enacted suicidal attempts, had repeated quarrelsbetween themselves or clung to a child becauseof their own desire for a caregiving figure, willbe loath for the true facts to be known. Inevitably they expect criticism and blame andthus distort the truth, sometimes unwittingly,sometimes deliberately. Similarly, the childrenof such parents have grown up knowing thatthe truth must not be divulged and perhapshalf-believing also that they themselves are toblame for every trouble, as their parents mayalways have insisted. A common method ofkeeping family disturbances secret is to attributethe symptoms to some other cause; he is afraidof boys at school (not that mother may take herlife); she suffers from headaches and indigestion(not that mother threatens to desert if she leaveshome); he was difficult from birth (not that hewas unwanted and neglected); she is sufferingfrom an endogenous depression (not that she isbelatedly mourning a father lost many yearsearlier). Time and again what is described as asymptom is found to be a response which, byhaving become divorced from the situationthat elicited it, appears inexplicable. Or else asymptom arises as a result of the patient tryingto avoid reacting with genuine feeling to a trulydistressing situation. In either case a first andmajor task is to identify the situation, or situations, to which the patient is either respondingor else inhibiting a response.

It is plainly desirable that any clinician undertaking this type of work should have at hisdisposal an extensive knowledge of deviantpatterns of attachment and caregiving behaviourand of the pathogenic family experiencesbelieved commonly to contribute to them; andhe should also be familiar with the sorts ofinformation that are frequently omitted, suppressed or falsified. Given such knowledge itmay often be evident that some piece of crucialinformation is missing and that claims of certainkinds are dubious or clearly false. Above all, a

clinician experienced in this work knows whenhe has yet to discover the facts and is preparedeither to wait for the relevant information toemerge or to probe gently into likely areas.Tyros are apt to jump to conclusions and bewrong.

In building up a clinical picture a psychiatrist is wise not to rely on traditional interviewing methods alone but, whenever possible,to engage in one or more family interviews. Noother technique is more likely fairly quickly toreveal present patterns in their true light andgive clues to how they might have developed.A large number of books on family psychiatryand family therapy are now available. Thoughthey call attention to the immense influencethat different patterns of interaction can haveon each family member and describe techniquesof interviewing and modes of intervention, theconcepts they use are not those of attachmenttheory. For purposes of this exposition, therefore,they are of limited value.

A great deal of work needs doing before wecan be confident which disorders of attachmentand caregiving behaviour are treatable bypsychotherapy and which not; and, if treatable,which of various methods is to be preferred.Much turns on a clinician's experience, capabilities and facilities. In general, we can followMalan (1963) in using as a principal criterionwhether the designated patient and/or membersof his family show a willingness to explore theproblem presented along the lines described:whether or not this is so usually emerges duringthe course of our assessment. Sometimes bothdesignated patient and relatives respond, readilyor reluctantly, to the notion that the problem orsymptoms complained of seem to make sense interms of the events and family disturbancesthey are describing. Not infrequently such ideasare unpalatable to one or more, and on occasionthey are rejected as irrelevant and absurd.Depending on these reactions we decide ourtherapeutic strategy.

There is not space here to consider the usesand limitations of the many possible patternsof therapeutic intervention, either with parentsand offspring (of all ages) or with married pairs,that have now become established practice.Joint interviews, individual interviews, alterna

JOHN BOWLBY 423

tions of the two, all have their place, and sohave prolonged sessions lasting several hours:but we are a long way from knowing whichpattern is likely to be best for a given problem.There are, however, certain principles that arerelevant to any of these therapeutic procedures.For ease of exposition I take the case of individual therapy; though note that it is possibleto rephrase each paragraph so that it refers tothe members of a family instead of to a singleperson.

As I see it a therapist has a number of interrelated tasks, among which are the.following:

(a) first, and above all, to provide the patientwith a secure base from which he can exploreboth himself and also his relations with all thosewith whom he has made, or might make, anaffectional bond; and simultaneously to make itclear that all the decisions as regards howbest to construe a situation and what action isbest taken have to be the patient's, and thatgiven help we believe him capable of makingthem;

(b) to join with the patient in such explorations, encouraging him to consider both thesituations, in which he nowadays tends to findhimself with significant persons and the. partshe may, play in bringing them about, and alsohow he responds in feeling, thought and actionwhen in those situations;

(c) to draw the patient's attention to the waysin which, perhaps unwittingly, he tends toconstrue the therapist's feelings and behaviourtowards him, and to the predictions he (thepatient) makes and the actions he takes as aresult; and then to invite him to considerwhether his modes of construing, predictingand acting may be partly or wholly inappropriatein the light, of what he knows of the therapist;

(d) to help him consider how the situationsinto which he typically gets himself and histypical rCactions to them, including what maybe happening between himself and the therapist,may be understood in terms of the experienceshe had with attachment figures during hischildhood and adolescence (and perhaps maystill be having) and of what his responses tothem then were (and may still be).

Although the four tasks outlined are conceptually distinct, in practice they have to be

pursued simultaneously. For it is one thing forthe therapist to do his best to be a reliable,helpful and continuing figure, and another forthe patient to construe him and trust him assuch. The more unfavourable the patient'sexperiences with his parents were, the less easyis it for him to trust the therapist now and themore readily will he misperceive, misconstrueand misinterpret what. the therapist does andsays. Furthermore, the less he can trust thetherapist the less will he tell him and the moredifficult will it be for both parties to explorethe painful or frightening or mysterious eventswhich may have occurred during the patient'searlier years. Finally, the less complete andaccurate the picture available of what happenedin the past the more difficult the patient'spresent feelings and behaviour are for bothparties to understand, and the more persistentare his misperceptions and misinterpretationslikely to be. Thus we find each patient isconfined within a more or less closed system andonly slowly, often inch by inch, is it possible tohelp him escape.

Of the four tasks the one that can best wait isconsideration of the past since its only relevancelies in the light it throws on the present. Thesequence may often be for the therapist andpatient, working together, first to recognize thatthe patient tends habitually to, respond to aparticular type of interpersonal situation in acertain self-defeating way, next to examinewhat kinds of feeling and expectation suchsituations commonly arouse in him, and thereafter to consider whether the patient may havehad experiences, recent or long past, whichhave contributed to his responding with thosefeelings and expectations in the situationsconcerned. In this way memories of relevantexperiences are evoked, not simply as unhappyoccurrences but in terms of the pervasiveinfluence they are exerting in the present on thepatient's feelings, thoughts and actions.

It is evident that a great many psychotherapists, irrespective of theoretical outlook,habitually address themselves to these tasks, sothat much of what I am saying will have longbeen familiar to them. In traditional terminology the tasks are referred to as providingsupport, interpreting the transference, and

THE MAKING AND BREAKING OF AFFECTIONAL BONDS. II.424

constructing or reconstructing past situations.If there are any new points of emphasis in thepresent formulation they are:

(i) giving a central place, not only in practicebut also in theory, to our role of providing apatient with a secure base from which he canexplore and then reach his own conclusionsand take his own decisions;

(ii) abjuring interpretations which postulatevarious forms of more or less primitive phantasyand concentrating instead on the patient's realexperiences;

(iii) directing attention particularly to thedetails of how the patient's parents may actuallyhave behaved towards him, not only during hisinfancy and childhood but during his adolescence and up to the present day as well; and alsoto how he has commonly responded;

(iv) utilizing interruptions in the course oftreatment, especially those imposed by thetherapist, either routinely as in the case ofholidays or exceptionally as in the case ofillness, as opportunities first to observe how thepatient construes a separation and responds toit, then to help him recognize how he is construing and responding, and finally to examinewith him how and why he should have developed so.

An insistence on the principle that a patient'sattention should be directed to consideringwhat his real experiences may have been, andhow these experiences may still be influencinghim, often gives rise to a misunderstanding.Are we doing no more, it may be asked, thanencouraging a patient to lay all the blame forhis troubles on his parents? And, if so, what goodcan that do? First, it must be emphasized that,as therapists, it is not our job to determine whois to blame or for what. instead, our task is tohelp a patient understand the extent to which hemisperceives and misinterprets the doings ofthose he is fond of or might be fond of in thepresent day, and how, in consequence, he treatsthem in ways that have results of a kind heregrets or deplores. Our task, in fact, is to helphim review the representational models ofattachment figures and of himself that withouthis realizing it are governing his perceptions,predictions and actions, and how those modelsmay have developed during his childhood and

adolescence, and, if he thinks fit, help him tomodify them in the light of more recent experience. Secondly, inasmuch as a patientmay be quick to blame, we may be able topoint to the emotional difficulties and unhappy experiences his parents may perhapshave had and thus invite his sympathy. Bearingin mind our medical role, we must approachwhat may be the deeply regrettable behaviour ofthe patient's parents in as objective a way aswe try to approach those of the patient himself.Our role is not to apportion blame but to tracecausal chains with a view to breaking them orameliorating their consequences.

This is a good moment to refer to familytherapy, since during the course of familyinterviews it may be possible to get a muchlonger perspective on how the current difficulties have come into being. By using suchoccasions to draw a detailed family tree, vitaldata may be unearthed for the first time,especially when grandparents are included. Asa colleague remarks, ‘¿�Itis amazing to see theeffects on a patient of hearing his grandparentstalk about their grandparents.'

Although I believe the same principles applyin family therapy as in individual therapy, thedifferences in application are too many to bedealt with here and deserve a full discussion oftheir own. One difference may, however, bementioned. A main aim of family therapy is toenable all members to relate together in such away that each member can find a secure basein his relationships within the family, as occursin every healthily functioning family. To thisend attention is directed to understandingthe ways in which family members may attimes succeed in providing each other with asecure base but at other times fail to do so,for example by misconstruing each other'sroles, by developing false expectations of eachother, or when forms of behaviour that wouldbe appropriately directed towards one familymember are redirected towards another. As aresult, during family therapy less time is likelyto be given to interpreting the transference thanin individual therapy. A main benefit is that,when therapy proves effective, it can often beterminated sooner and with less pain anddisturbance than can individual therapy, during

JOHN BOWLBY 425

the course of which a patient may easily cometo regard the therapist as the only secure basehe can ever imagine having.

Let us return now to speak again in terms ofindividual therapy.

I have already emphasized that, in myview, a major therapeutic task is to help apatient discover what the situations are, currentor past, to which his symptoms relate, be theyeither responses to those situations or else theside effects of trying not to respond to them.Since it is the patient who has been exposed tothe situations in question he is in a sense alreadyin possession of all the relevant information.Why then does he need so much help to discover it?

The fact is that much of the most relevantinformation refers to extremely painful orfrightening events that the patient would muchprefer to forget. Memories of being held alwaysto be in the wrong, of having to care for adepressed mother instead of being cared foryourself, of the terror and anger you felt whenfather was violent or mother was utteringthreats, of the guilt when you were told yourbehaviour would make your parent ill, of thegrief, despair and anger you felt after a loss, ofthe intensity of your unrequited yearningduring a period of enforced separation. No onecan look back on such events without feelingrenewed anxiety, anger, guilt or despair. Noone, either, cares to believe that it was his veryown parents, who at other times may have beenkind and helpful, who on occasion behaved insome most distressing way. Nor are parentslikely to have encouraged their children toregister or to recall such events; all too oftenindeed they have sought to disconfirm theirchildren's perceptions and have enjoined themto silence. For parents, on their part, to considerin what ways their own behaviour may havecontributed and perhaps still be contributingto their child's current problems is equallypainful. In all parties, therefore, there arestrong pressures towards forgetting and distorting, repressing and falsifying, exonerating oneparty and blaming another. Thus, we find,defensive processes are as frequently aimedagainst recognizing or recalling real life eventsand the feelings aroused by them as ever they

are against becoming aware of unconsciousimpulse or phantasy. Indeed, it is often onlywhen the detailed course of some disturbed anddistressing relationship has been recalled andrecounted that the feeling aroused by it and theactions contemplated in reply come to mind.I well remember how a silent inhibited girl inher early twenties given to allegedly unpredictable moods and hysterical outbursts athome responded to my comment ‘¿�itsounds tome as though your mother never has reallyloved you'. (She was the second daughter, to befollowed in quick succession by two muchwanted sons.) In a flood of tears she confirmedmy view by quoting, verbatim, remarks madeby her mother from childhood to the presentday, and the despair, jealousy and rage hermother's treatment roused in her. Discussionof her profound belief that I also found herunlovable and that her relations with me wouldbe as hopeless as they were with her mother,which accounted for the sulky silences whichhad been impeding therapy, followed naturally.

The technique developed for helping bereavedpeople illustrates well the principles I amdescribing. In this work, the events in questionand the feelings, thoughts and actions arousedby them are recent and, compared to childhoodevents and responses, likely to be more clearlyand accurately remembered. Painful feeling,moreover, is often either still present or at leastmore readily accessible.

Those counselling the bereaved (e.g.Raphael, 1975) have found empirically that,if they are to be of help, it is necessary toencourage a client to recall and recount, ingreat detail, all the events that led up to theloss, the circumstances surrounding it and herexperiences since;* for it seems only in this waythat a bereaved person can sort out her hopes,regrets and despairs, her anxiety, anger andperhaps guilt, and, just as important, reviewall the actions and reactions that she had it inmind to perform and may still have it in mindto perform, inappropriate or self-defeatingthough many of them might always have beenand would certainly be now. Not only is it

* For demographic reasons the development of tech

niques of bereavement counselling has been mainly withwidows; hence the gender in this paragraph and the next.

426 THE MAKING AND BREAKING OF AFFECTIONAL BONDS. II.

desirable for a bereaved person to revieweverything surrounding the loss but to reviewalso the whole history of the relationship, withall its satisfactions and deficiencies, the thingsthat were done and those that were left undone.For it appears that only when she has been ableto review and reorganize past experience doesit become possible for her to consider herself asa widow and her possible futures with theirlimitations and opportunities, and to make thebest of them without subsequent strain orbreakdown.

So far I have not mentioned advice. Experience of bereavement counselling shows thatuntil the bereaved has had time to progresssome distance in her review of the past and herreorientation towards the future advice doesfar more harm than good. Furthermore, what aperson needs much more than advice is information. For a widow's situation in life is verydifferent to what it was. Many familiar coursesof action are now closed and she may welllack information about those now open to her,with the advantages and disadvantages ofeach. Providing her with, or guiding hertowards, relevant information and helpingher review its implications for her future, whilstleaving her to take the decisions, may in duecourse be very useful. Hamburg has repeatedlyemphasized tht great importance of a personseeking and utilizing new information as anecessary step in coping with any stressfultransition (Hamburg and Adams, 1967). Assisting a patient to do so at the right time and in theright way thus constitutes a fifth task for thetherapist.

When helping a psychiatric patient the tasksto be undertaken and the techniques forachieving them are, I believe, no different inkind to counselling the bereaved. Such differences as exist are due to the fact that the patient'srepresentational models and the patterns ofbehaviour based on them have been so longentrenched, that many of the events which ledto their development occurred long ago, andthat the patient and members of his family mayhave a deep reluctance to look at things afresh.As a consequence, when helping a psychiatricpatient explore his world and himself, a therapisthas a complex role to fill.

Thus, he must encourage his patient toexplore even when he is resistant to doing so,and also help him in the search by drawingattention to features in the story that seemlikely to be relevant and away from those thatseem irrelevant and distracting. Often he willcall a patient's attention to his reluctance evento consider certain possibilities and, perhapssimultaneously, sympathize with the bewilderment, anxiety and pain that to do so mightentail. In all this, it will be noted, I am inagreement with those who believe a therapist'srole should be an active one. Yet, to be effective,he must recognize that he cannot go faster thanhis patient, and that by calling attention topainful topics too insistently he will arouse hispatient's fear and earn his anger or deepresentment. Finally, he must never forget that,plausible—even convincing—though his ownsurmises may seem to him, compared to thepatient he is ill-placed to know the facts, andthat in the long term it is what the patienthonestly believes that must be accepted as final.

Here we touch on the immensely importantissue of the therapist's own outlook and valuesin relation to the patient and his or her problems; for whatever the therapist's outlook andattitudes may be they are bound to influencethe patient's own attitudes, if only through thelargely unconscious process of observationallearning (identification). In this process thepatient's experience of the therapist's behaviourand tone of voice and how he approaches atopic are at least as important as anything hesays. Thus, with attachment theory in mind, atherapist will convey, largely by non-verbalmeans, his respect and sympathy for his patient'sdesires for love and care from her relatives, heranxiety, anger and perhaps despair at herwishes having been frustrated and/or denigrated,not only in the past but perhaps also in thepresent, and the distress and grieving to whichperhaps a childhood bereavement may havegiven rise; and he will indicate his understanding that similar conflicts, expectationsand emotions may be active in the therapeuticrelationship as well. As much through nonverbal as through verbal communication alsowill a therapist convey respect for and encouragement of his patient's desire to explore

JOHN BOWLBY 427

the world and reach her own decisions in life;while at the same time he recognizes that shemay have a deep-seated belief, derived fromwhat others have insisted, that she is incapableof doing so. In these everyday exchanges acertain pattern of conducting interpersonalrelationships is, unavoidably, demonstrated bythe therapist, and this cannot but influence insome degree his patient's outlook. For example,in place of what may have been a pattern offault-finding, punishment and revenge, or ofcoercion by induction of guilt, or of evasionand mystification, he introduces a pattern inwhich an attempt is made to understandanother person's viewpoint and to negotiateopenly with him. At some points in therapydiscussion of these different ways of treatingpeople, and the probable consequences of each,can be useful. During such discussions a therapistis likely both to raise questions and to provideinformation while, once again, leaving thepatient to take the decisions.

Clearly, to do this work well requires of thetherapist not only a good grasp of principles butalso a capacity for empathy and for toleratingintense and painful emotion. Those with astrongly organized tendency towards compulsiveself-reliance are ill-suited to undertake it and arewell advised not to.

In discussing earlier the therapist's four basictasks it is emphasized that, though conceptuallydistinct, in practice they have to be pursuedsimultaneously. How far therapy can andshould be taken with any one family or patientis a complex difficult question. The main pointperhaps is that a restructuring of a person'srepresentational models and his re-evaluationof some aspects of human relationships, with acorresponding change in his modes of treatingpeople, are likely to be both slow and patchy.In favourable conditions the ground is workedover first from one angle then from another.At best progress follows a spiral. How far atherapist goes and how deeply involved hebecomes is a personal matter for both parties.Sometimes one or a few sessions enable a patientor a family to see problems in a new light, orperhaps confirm that a point of view, rejectedand ridiculed by others, is indeed plausible andcan with advantage be adopted. (See accounts

and examples by Caplan, 1964; Argles andMackenzie, 1970; Lind, 1973; Heard, 1974).A special value ofjoint family interviews is thatthey enable each member of a family to discoverhow each of the others views his family lifeand to move together in reappraising it andchanging it. Often, too, it enables all familymembers to learn, often for the first time, of theunhappy experiences that one or other parentmay have had in years past, to the consequencesof which current family conflict may quiddybe perceived as due. (An excellent example, inwhich a current marital crisis is traced to thepersisting consequences of failed mourning afterchildhood loss, is described by Paul, 1967.)There are many other cases, however, especiallyin patients who have developed a highlyorganized false self and become compulsivelyself-reliant or given to the caretaking of others,in which a much longer period of treatment maybe necessary before change of any kind isseen.

Nevertheless, however short or long thetherapy, evidence is clear that, unless a therapistis prepared to enter into a genuine relationshipwith a family or individual, no progress can beexpected (Malan, 1963; Truax and Mitchell,1971). This entails that a therapist should, so

far as he can, meet the patient's desire for asecure base, while recognizing that his bestefforts will fall short of what a patient desiresand might well benefit from; that he shouldenter into the patient's explorations as a companion ready either to take the lead or to beled; and that he should be willing to discuss apatient's perceptions of him and the degree towhich they may or may not be appropriate,which is sometimes not easy to determine; andfinally, that he should not pretend otherwiseshould he become anxious about a patient orirritated by him. This is especially importantfor those patients whose parents have persistently simulated affection to cover deepseated rejection of them. Guntrip (i@@) haswell described the therapist's job: ‘¿�Itis, as I seeit, the provision of a reliable and understandinghuman relationship of a kind that makes contactwith the deeply repressed traumatized child ina way that enables [the patient] to becomesteadily more able to live, in the security of a

428 THE MAKING AND BREAKING OF AFFECTIONAL BONDS. II.

new real relationship, with the traumaticlegacy of the earliest formative years, as itseeps through, or erupts into consciousness.'

When he adopts a stance of this kind a therapist risks certain dangers of which it is as well tobe aware. First, a patient's eagernesss for asecure base and his tormenting fear he will berejected may make his claims insistent anddifficult to deal with. Secondly, and far moreserious, in exerting these claims a patient mayapply to the therapist the very same methodsthat a parent may have used on him when hewas a child. Thus, a man whose mother whenhe was a boy inverted the relationship bydemanding he should care for her, and who usedthreats or guilt-inducing techniques to forcehim to do so, may during treatment applythese very same techniques to his therapist.Plainly it is of the greatest importance that thetherapist should recognize what is happening,trace the origin of the techniques being usedand resist them, i.e. set limits. Yet the moresubtly guilt-inducing the techniques are and themore eager the therapist is to help the greateris the danger of his being drawn in. A sequenceof this sort, I suspect, accounts for many of thecases described by Balint (1968) as exhibiting‘¿�malignantregression' and classified by othersas borderline. The clinical problems to whichthey can give rise are well illustrated by Main(m@@') and also by Cohen et al@The latter group point to the danger of atherapist not recognizing when a patient'sexpectations are becoming unrealistic, becausewhen it becomes clear they will not be met thepatient may suddenly feel totally rejected andso despair.

Because attachment theory deals with so manyof the same issues as are dealt with by othertheories of psychopathology—issues of dependency, object-relations, symbiosis, anxiety,grief, narcissism, trauma and defensive processes—it is hardly surprising that many of thetherapeutic principles to which it leads shouldbe long familiar. Some of the overlaps betweenideas I have advanced and those of Balint (1965,1968), Winnicott (1965) and others have beendiscussed by Pedder (1976) in connection withthe treatment of a depressed patient with a‘¿�falseself'. Other overlaps, for example the

equivalence of Winnicott's concept of play(Winnicott, 1971) and what is here termedexploration, have been noted by Heard (inpress). Overlaps with the ideas of therapistswho have drawn special attention to the partplayed in the genesis of episodic depressionsand many other neurotic symptoms by thefailure to mourn a parent lost during childhoodor adolescence (e.g. Deutsch, 1937; Flemingand Altschul, 1963) or to come to terms witha parent's attempted suicide (Rosen, 1955) willbe evident. Yet, though these overlaps are realenough, there are significant differences also,both of emphasis and of orientation. Theyturn partly on how we conceive the place ofattachment behaviour in human nature (or, bycontrast, what use we make of the concepts ofdependency, orality, symbiosis and regression),and partly on how we believe a person acquirescertain disagreeable and self-defeating ways ofinteracting with those close to him, or misplaced beliefs, such for example that he isinherently incapable of doing anything useful oreffective.

All those who think in terms of dependency,orality or symbiosis refer to the expression ofattachment desires and behaviour by an adultas being the result of his having regressed tosome state believed to be normal during infancyand childhood, often that of a suckling at hismother's breast. This leads therapists to talk toa patient about ‘¿�thechild part of yourself' or‘¿�yourbaby need to be loved or fed', and torefer to someone tearful after a bereavementas being in a state of regression. In my view allsuch statements are mistaken both for theoreticaland for practical reasons. As regards theoryenough has been said to make it clear thatI regard the desire to be loved and cared for asbeing an integral part of human nature throughout adult life as well as earlier and that theexpression of such desires is to be expected inevery grown-up, especially in times of sicknessor calamity. As regards practice, it seems highlyundesirable to refer to a patient's ‘¿�babyneeds'when we are trying to help him recover hisnatural desires to be loved and cared for which,because of unhappy experiences earlier in hislife, he has endeavoured to disclaim. Byconstruing them as childish and referring to

JOHN BOWLBY 429

them as such, a patient can easily interpret ourremarks as disparaging and reminiscent of adisapproving parent who rejects a child seekingto be comforted and calls him ‘¿�sillyand babyish'.An alternative way of referring to a patient'sdesires is to refer to his yearning to be loved andcared for which we all have but which in hiscase went underground when he was a child(for reasons we may then be able to specify).*

A second area of difference concerns how wesuppose a person comes to apply to spouse andchildren, and sometimes also to therapist,certain disagreeable pressures, for examplethreats of suicide or subtle modes of inducingguilt. In the past, though the problem has beenrecognized, no great attention has been given tothe possibility that the patient learned how toexert these pressures through having sufferedthem himself when a child and, consciously orunconsciously, is now copying his parent.

A third area of difference concerns the originof prolonged despair and helplessness. Traditionally this has been traced, almost solely, to theeffects of unconscious guilt. The view I favour,which is in keeping with Seligman's studies oflearned helplessness (Seligman, 1975) and isalso compatible with the traditional view, is thatsomeone who is readily plunged into prolongedmoods of hopelessness and helplessness hasbeen exposed repeatedly during infancy andchildhood to situations in which his attempts toinfluence his parents to give him more time,affection and understanding have met withnothing but rebuff and punishment.

Finally, we may ask, what evidence is therethat therapy conducted according to the principles outlined is effective and, if so, in whattypes of case? The answer is that there is nodirect evidence because no series of patientshas been treated along exactly these lines, sothat no investigation of results has been possible.The most that can be said is that certain indirectevidence is hopeful. It comes from investiga

* The distinctions I am making are identical with

those made by Neki (1976), who contrasts the value setby Indian culture on ‘¿�stronginterdependent affiliativeattachments fostered and carried over into adulthood'with the Western value of ‘¿�achievement-oriented independence'. His discussion of how these divergent idealsaffect therapy in these respectsfollowslines closelysimilarto those sketchedhere.

tions of the efficacy of brief psychotherapy and ofbereavement counselling.

For many years Malan (1963, 1973) has beenexamining the results of brief psychotherapy(defined arbitrarily as no more than 40 sessions)and has concluded that a group of patients canbe specified who are likely to benefit from acertain type of psychotherapy, the features ofwhich can also be specified. The patients likelyto benefit are those who, during the first fewinterviews, show themselves able to face emotional conflict and are willing to explore feelingsand to work within a therapeutic relationship.The technique that proved effective was one inwhich the therapist felt able to understand hispatient's problems and to formulate a plan;and in which he attended to the transferencerelationship and interpreted it boldly, payingspecial attention to the patient's anxiety andanger when the therapist set a date for termination.

During the course of a replication studyMalan and his colleagues reached the sameconclusion. In addition they found evidencethat ‘¿�animportant therapeutic factor is thepatient's willingness to involve himself in away that repeats a childhood relationship' withone or both of his parents and his ability, withthe therapist's help, to recognize what ishappening (Malan, 1973). A further study bythe same group, this time of patients whoimprove after no more than a single interview,presents further evidence in support of thatconclusion (Malan ci at, 1975).

Although the theory of psychopathology usedby Malan and his colleagues differs in somerespects from the one outlined here, there areimportant similarities. Furthermore, as will benoted, there is considerable similarity betweenthe principles of technique he finds effectiveand those advocated here.

Evaluation of the efficacy of bereavementcounselling for widows thought to have a badprognosis also points in a hopeful direction.Among widows who received the form ofcounselling described above, significantly morewere found, at the end of thirteen months, tohave progressed favourably than among thosein a control group who received no counselling(Raphael and Maddison, 1976).

430 THE MAKING AND BREAKING OF AFFECTIONAL BONDS. II.

It must, of course, be recognized that to outline principles of therapy is a great deal easierthan to apply them in the ever varying conditions of clinical practice. Furthermore, thetheory itself is still at an early stage of development and a great deal of work is still to bedone. Among priority tasks are to determineboth the range of clinical conditions to whichthe theory is relevant and the particular variantsof technique best suited to treat them.

Meanwhile, those who adopt attachmenttheory believe that both its structure and itsrelation to empirical data are now such that itsusefulness can be tested systematically. In thefields of aetiology and psychopathology it canbe used to frame specific hypotheses whichrelate family experience to several forms ofpsychiatric disorder and also, it may be, to theneurophysiological changes that accompanythem as Hamburg and his colleagues (i@@4@)believe. In the field of psychotherapy it can beused to specify therapeutic technique, to describetherapeutic process and, given the necessarytechnical developments, to measure change.As research proceeds the theory itself will nodoubt be modified and amplified. This giveshope that, in due course, attachment theorymay prove useful as one component within thatlarger corpus of psychiatric science whichHenry Maudsley did his utmost to foster.

Acx@iowzoasasrrs

In preparing this Lecture I am grateful for commentsand suggestions on early drafts from a number of colleagues,especially DrJohn Byng-Hall and Dr Dorothy Heard.

REFERENCES

Part H

Aaoi@s, P. & MACKENZIE, M. (‘97°)Crisis interventionwith a multi-problem family: a case study. Journd ofChild P@ychologyand P@ychiatrj,ii, 187-95.

BAUNT, M. (1965) Primaiy Love and P@yc/zoanaljtic Technique. London: Tavistock.

(i968) The Basic Fault. London: Tavistock.CAPI.@u4,G. (1964) Principles of Preventive P@ychiatrj. New

York: Basic Books; London: Tavistock.Coiir.i@i, M. B., BAKES, G., Cosrass, R. A., Fao@sse

RF.ICHMANN,F. & WEIGERT, E. (i@4) An intensivestudy of twelve cases of manic-depressive psychosis.Psychiatry, ‘¿�7,103—37.

DEV'rscfl, H. (i937) Absence of grief. PsychoanalyticQparterly, 6, 12—22.

Futaweo,J. & Ax.@rscuuz@,S. (1963) Activation of mourningand growth by psychoanalysis. International Journal ofPsychoanalysis, 44, 419-31.

GUNTRIP, H. (1975) My experience of analysis withFairbaim and Winnicott. International Review ofPsychoanalysis, 2, 145—56.

HAMBURG, D. A. & AnAMS, J. E. (1967) A perspective oncoping behaviour. Archives of General Psychiatry, 17,277—84.

—¿� HAMBURG, B. A. & BARCIIAS,J. D. (1974) Anger and

depression in perspective of behavioural biology.In Parameters of Emotion (ed. L. Levi). New York:Raven Press.

H@iw, D. H. (,974) Crisis intervention guided byattachment concepts: a case study. Journal of ChildPsychology and Pjychwtry, 15, 111-22.

—¿� (in press) From object relations to attachment: a

framework for family therapy. British Journal qfMedical Psychology.

LIIqE), E. (i@7@) From false-self to true-self functioning:a case in brief psychotherapy. British Journal ofMedical Psychology, 46, 381-9.

Mam, T. F. (‘957) The ailment. British Journal of MedicalPsjchology, 30, 129-45.

MALAN, D. M. (1963) A Study of Brief P@hotherapy.London: Tavistock Publications.

—¿� (1973) Therapeutic factors in analytically-oriented

brief psychotherapy. In Support, Innovation andAutonomy (ed. R. H. Gosling). London: TavistockPublications.

—¿� H@m, E. S., BACAL, H. A. & BALPOUR, F. H. G.

(@7@) Psychodynamic changes in untreated neuroticpatients: II. Apparently genuine improvements.Archives of General Psychiatry, 32, 110-26.

Nasa, J. S. (1976) An erarnination of the cultural relativism of dependence as a dynamic of social andtherapeutic relationships. Parts i and 2. BritishJournal of Medical Psychology, @,1-22.

PAUL, N. L. (1967) The role of mourning and empathy inconjoint marital therapy. In Family Therapy andDisturbed Families (ecis G. H. Zuk and I. BoszormenyiNagy), pp 186-205. Palo Alto, California: Scienceand Behavior Books.

PEDDER, j. (i@76) Attachment and new beginning.

International Review of Psychoanalysis, 3,491-7.RAPHAEL, B. (1975) Management of pathological grief.

Australian and Xew Zealand Journal of Psychiatry, 9,173—80.

—¿� & MADDISON, D. C. (1976) The care of bereaved

adults. In Modern Trends in Psychosomatic Medicine(ed. 0. Hill). London: Butterworth.

ROSEN, V. H. (@95@) The reconstruction of a traumaticchildhood event in a case of derealization. Journal ofthe American Psychoanalytical Association, 3@ 211-21.Reprinted in Survivorsof Suicide(ed. A. C. Cain).Springfield, Illinois: C. C. Thomas, 1972.

SELIGMAN, M. E. P. (1975) Helplessness: on Depression,Development and Death. San Francisco: W. H. Freeman.

JOHN BOWLBY 43'TRUAx, C. B. & MITCHELL, K. M. (‘97') Research on

certain therapist interpersonal skills in relation toprocess and outcome. In Handbook of Psychotherapyand Behavior Change (eds A. E. Bergin and S. L.Garfield). New York: Wiley.

WINNIC0TF, D. W. (ig6@) The Maturational Processes andthe Facilitating Environment. London: Hogarth Press.

(1971) Playing and Reality. London: Tavistock

Publications.

John Bowiby, M.D., F.R.C.P., F.R.C.Psych.,Honorary Consultant Psychiatrist, Tavistock Clinic, Belsize Lane,London XW3 5BA