For Better or Worse: Interpersonal Relationships and Individual Outcome

In these days of biological reductionism, with its brain disease metaphor and emphases on descriptive diagnoses and psychotropic medications, it is easy to forget that life is lived in relationships, and the quality of those relationships has much to do with how life turns out. Curiously, the minimization of relationships has occurred despite an increasing emphasis on relationships in major theories of growth and development, including psychoanalysis, attachment theory, and cognitive approaches to psychotherapy. At the broadest level, the interpersonal approach to psychiatry focuses on the interplay between psychopathology and the individual's relationships with significant others. At its center this perspective holds that relational structures—the more or less enduring patterns of interaction—either facilitate or impede the continued maturation of the participants. It is important to note that the relationship between an individual and his or her relational system is not linear; rather, individual characteristics influence system properties, and these properties shape individual characteristics.

Although this presentation focuses on marital relationships and their impact on the participants, it is my belief that what has been learned applies equally well to all central adult relationships, including enduring heterosexual and homosexual alliances, deep friendships, and other dyadic relationships that form life's crucial context. The current status of the interpersonal perspective is illustrated with the use of selected data from three sources: the Timberlawn studies of well-functioning marriages and families, data supporting the idea of healing marriages, and information regarding the role of marital factors in the onset and course of depressive syndromes. Before turning to these areas, however, I will describe a small group of basic relationship concepts. These concepts are derived from the literature (1–7) and my research and clinical work with marital couples.

BASIC RELATIONSHIP CONCEPTS

1. Attachment. Each individual seeks a central relationship in which to find security, satisfaction, and meaning.

2. Connection and separation. Each individual brings to the central relationship a greater or lesser proclivity for both connection (closeness, intimacy) and separateness (independence, autonomy), the strengths of which are determined by the complex interaction of biological, developmental, psychodynamic, social, and cultural variables.

3. Negotiation. The early stages of relationship formation involve a complex negotiation between the partners of the balance of connection and separateness that is to prevail in the relationship and that cannot be predicted solely on the basis of each individual's characteristics.

4. Unconscious fears. Because for many individuals unconscious fears of connection, separateness, or both that arise from childhood experiences play an important role in the negotiation, each individual attempts to achieve a relationship balance that provides both maximum satisfaction and freedom from fear.

5. Power. In the negotiation of the relationship structure, including, centrally, the balance of connection and separateness, each individual attempts to influence the other to accept a balance congruent with his or her proclivities and fears.

6. Maintenance of balance. Once established, the balance is maintained by day-to-day interactions around issues such as space (actual and metaphorical), affection, work, relationships with extended families, parenting responsibilities, and other everyday issues. These interactions become patterned and usually operate out of awareness.

7. Changes in the balance. Although the balance is more or less enduring, changes may occur as a consequence of normative challenges (e.g., parenthood) and stress (e.g., loss of a child). Changes in the balance result from either relatively conflict-free negotiation or crisis and conflict.

8. Health-facilitating balances. Although any balance of connection and separateness may be relatively satisfactory to both participants, certain balances facilitate the continued and healthy development of the participants.

9. The optimal balance. Although contextual factors (e.g., poverty, chronic illness) may limit that which is possible, health-facilitating systems are usually characterized by high levels of both connection and separateness.

10. Values. A system's characteristic interactions reflect underlying shared beliefs about the nature of humankind (e.g., how dangerous it is to be close to others) and other existential concerns (e.g., whether the world is orderly or capricious). The system's characteristic interactions also memorialize the underlying shared beliefs. Cultural factors often influence the values that predominate in a marital or family system.

THE TIMBERLAWN STUDIES

For 30 years my colleagues and I have reported findings from three different studies, two of which were cross-sectional and descriptive and one of which was longitudinal (8–14). Basing our work on the value judgment that the cardinal tasks of the family are the facilitation of the continuing maturation of the spouses and the provision of an emotional climate conducive to the healthy development of the children, three groups, consisting of 128 individuals, were studied. Two of the groups, a middle-class Caucasian and a lower-income African American, contained adolescent children. The third group involved young, predominately middle-class couples expecting their first child at the start of the study. Multiple observational and self-report measures were used at the level of the individual, the couple, and the family in all three study groups.

For the purposes of this presentation, the focus is on the characteristics of the parents' marital relationship that facilitated healthy development of all family members. Despite the absence of overt psychiatric disturbances in the participants, the groups contained many couples and families who had overtly dysfunctional relational structures. Reasonably healthy people can construct very dysfunctional relationships.

First, it is clear that the marital relational structure formed the template for the family system. Although on the surface some couples may look either better or worse with the children present, independent ratings of marital and family videotapes are highly correlated.

Another within-group difference concerns what can be called style. Some competent couples have a charged, passionate bond, and others present a more even and bland exterior. For some, a vibrant sexuality is present, while for others, the sexual relationship is more peripheral. Some competent couples share a strong emphasis on career; for others, careers are valued not for themselves as much as for what marital and family benefits they provide. Religion is another difference. For some, neither beliefs nor practices are important, while for others, religion provides an important foundation for the marital relationship.

These and other differences are unrelated to the competence of the marital relationship in facilitating the maturation of the spouses and the healthy development of the children. What, then, is common to couples who so often seem very different in style? Characteristics of well-functioning marriages include the following.

1. Power is shared. It is clear that in the basic process of defining the relationship structure, including how much of what types of connection and separateness are to prevail, both partners have had significant influence. Many couples come to a relatively equal sharing of power after considerable trial and error in which each tries to get his or her definition of the relationship accepted. The complex negotiations occur around everyday issues such as money, sexuality, families of origin, and the like, but imbedded in these everyday matters is the more fundamental issue of how connected (in agreement) and how separate (different) the spouses are to be.

2. There are high levels of both connection and separateness. Competent marriages encourage both connection and separateness. The spouses are strongly committed to each other (no other relationship is as important), demonstrate high levels of closeness (sharing values and interests), and may communicate at the in~timate level (the reciprocal sharing of vulnerabilities). They also demonstrate high levels of separateness (firm self-identities, autonomy, and the capacity for generative solitude).

3. Respect for subjective reality is prominent. The ability to accept and respect each other's opinions and feelings results, in part, from shared power. One is better able to share openly and to both listen carefully and respond respectfully to the experiences of an equal. Differences are not usually the source of power struggles and conflict in well-functioning marriages; rather, they are often valued. There is no preoccupation with the “truth” unless such is called for by the context (“Our car is stalled on the railroad tracks! Is the train coming?”)

4. Affects are openly expressed. The unspoken rules that govern who can express which feelings under what circumstances are very liberal. Perhaps, again, in response to the shared power and respect for subjective reality, a wide range of feelings are freely expressed. The mood of the relationship is affectionate and optimistic under usual circumstances. Empathy is found more frequently than it is in less competent couples.

5. Problem solving is highly developed. Once again, equal power and respect for differences sets the stage for effective problem solving. Competent couples explore problems thoroughly, use outside experts and data when such are needed, and show the ability both to negotiate and to reach compromises. In laboratory problem-solving exercises these couples are not only effective in the sense of reaching closure (unlike many less competent couples) but are seen as clearly working together.

6. Conflict, although inevitable, is not chronic and infrequently escalates. Chronic conflict often focuses on a particular topic or topics, but it is frequently related to the failure of the partners to agree on a basic definition of their relationship. In competent marriages the basic definition of the relationship is acceptable to both partners. When conflicts do occur they often involve those situations in which one partner is not emotionally available to the other. The couple's usual mechanisms of repair fail, and conflict ensues. In competent marriages escalation (increased affective intensity) and generalization (broadening the scope of the initial conflict) may occur, but at a lower frequency than in less competent couples.

These six features of competent marriages can be understood as the foundation of a relational structure that facilitates emotional maturation and healing. Both the strength of the connection and the ability to repair it after the inevitable periods of lack of synchrony, in combination with the encouragement of autonomy, play crucial health-inducing roles. To be healthy means to be able both to connect and to stand alone.

HEALING MARRIAGES

Many years ago and before our studies of well-functioning marriages and families had started, my attention was called to the possible healing impact of new connections by spending a day with each of 35 former intravenous drug users who had stopped using drugs without formal treatment. My motivation for this project was the question of how persons are able to emerge from an addiction that is notoriously difficult to treat. When asked how they had done so, most of the former addicts talked about “hitting bottom” or used other metaphors of despair. When asked, however, to describe what was going on in their lives before and during the period of giving up intravenous drugs, each subject described a new and intense connection to either a person or a belief system (usually religious). What seemed important about their descriptions was the intensity of the new connections.

As a result of this experience with these young men and women, my interest in the ability of some adult relationships to transform the lives of some individuals increased. At this point involvement with research volunteer couples and families revealed that, for some, there was a history of childhood abuse, neglect, or abandonment—much the same as is usually found in clinical samples—but in these spouses there was clear evidence of competent marriages and healthy individual functioning. The interview data suggested that the marital relationship had been an experience of individual transformation, a healing process.

In couples therapy it was useful to understand marital dysfunction as failed healing (15). It was usually possible to identify a central problematic relationship of childhood that was either being vigorously defended against or being reenacted in the marriage. The issue can be framed as why some people repeat the pattern of the past (why transference prevails or, in Ackerman's evocative words, why love is “the language of scars”) (16, p. 79), whereas others work through the internalized childhood relationship and experience a healing process.

These experiences with research and clinical samples led to a literature review and a growing list of studies suggesting healing marriages. I describe below a few representative studies.

Quinton et al. (17) reported on 94 women who had been raised in institutions and a control group of 51 women who had been raised by their parents. The institution-raised girls had been given up by their parents because of their behavioral difficulties. In young adulthood these women had a greater prevalence of both poor psychological functioning and parenting difficulties. About one-fourth of the institution-raised women had good outcomes that were associated with entering stable marriages with healthy, supportive men. The authors noted that the women with stable marriages and better outcomes did not have fewer or less serious adolescent problems, and that the mechanisms underlying healing marriages are unknown.

Women who were abused as children are more likely to abuse their own children. An emotionally supportive relationship with a husband or boyfriend has been shown to protect such women from abusing their children (18, 19).

Some of the delinquent boys studied by the Gluecks in the 1940s turned out well in adult life (20). The two processes that reversed the delinquent life trajectory were a steady job with an employer who valued them and a strong, stable marriage.

The Berkeley Guidance Study followed four generations of subjects. Caspi and Elder (21) described the interplay of individual problems, marital dynamics, and parenting skills in the generational production of psychopathology. Women who had demonstrated problems with anger and inadequate behavioral controls as children were likely to marry passive men and to manifest continuing problems as adults. They were described by their children as mean and ill-tempered. If such women, however, married more assertive men, they did not continue to manifest behavioral problems in adulthood and were not described negatively by their children. Once again, the authors indicated that although marriage can be a corrective experience, its mode of operation is not known.

Paris and Braverman (22) presented qualitative data about marriages that appear to interrupt the course of borderline pathology in young women. Older, caretaking spouses who provide an accepting environment may reverse the course of mild to moderate borderline pathology.

There is also evidence that women who describe insecure attachments with their mothers during childhood (a predictor of insecure attachments with their own infants) may enter into secure attachments with their infants if they have experienced a “corrective attachment experience” in marriage or psychotherapy (23).

These and other studies demonstrating discontinuity in development have been examined (24, 25), with the conclusion that “corrective emotional experiences” with spouses, teachers, and therapists provide the most common mechanisms for healing. Thus, it is believed that although we have evidence that adult relationships transform lives, we know little about how such healing relationships come about and what their dynamics might be. At the descriptive level, it appears that some such relationships may be brought about through the mechanisms of emotional support and affirmation (empathy, warmth, and genuineness) and others through providing a specific relationship ingredient that is needed (assertiveness rather than passivity). At this stage of our knowledge, however, we do not know whether the crucial factors are to be found in the personalities of the individuals or their partners, the structure of the transforming relationships, or both. Contemporary psychoanalytic theory suggests that we might well focus on higher-order internalization processes. Blatt and Behrends (26), for example, suggested that growth resulting from psychoanalytic treatment involves the same processes as normal development, and they focused on the establishment of a strong affective bond and repair of its disruptions as the necessary precursors of internalization in both situations.

A final comment about the dynamics of healing relationships: it seems likely that whatever the crucial dynamics turn out to be, they must be understood as powerful enough to block the strong inclination to repeat—the ubiquitous presence of transference.

MARITAL RELATIONSHIPS AND DEPRESSION

Although Vaillant (27) has written that in the longitudinal study of men's lives, a stable marriage is synonymous with freedom from all serious individual psychopathology, here the focus will be only on the relationship of marriage and depression. This is because there are more data that assist in a beginning clarification of this complex interface.

From a clinical perspective, couples are seen when a depressed spouse fails to respond to antidepressants and individual psychotherapy, and the referring clinician suspects that there is a connection between the refractoriness to treatment and the patient's marriage. Several types of relational structures are commonly seen. In my experience the most common is a dominant-submissive marriage in which the depressed patient has played the less powerful role. He or she (more usually, she) complains about the spouse's controllingness, emotional inaccessibility, and lack of investment in the marriage. The depressive symptoms can be understood as attempts, usually unconscious, to change the relational structure. They may reflect one or several themes. Depressive symptoms may be a plea for help, an attempt to redress the power imbalance, or a vengeful effort to punish.

A second type of relational structure involves the very distant marriage. Neither spouse is particularly dominant; rather, each participated in the negotiation of a relationship involving high separateness and little closeness and intimacy. Often, the onset of the depressed spouse's symptoms follows the loss of a relationship with a child, parent, or friend that provided her (or him) with closeness and intimacy. The unconscious messages carried by the symptoms may also involve needs for closeness, power, and vengeance.

A third marital relational structure often seen in the marriages of depressed individuals is chronic conflict. The couple has never been able to agree on a definition of their relationship, and the depressive symptoms are understood as another set of unconscious tactics in the chronic struggle to define the relationship.

In all three types of marital relational structures, the core conflict involves the struggle to define (or redefine) the basic structure of the marriage, particularly the amount and quality of closeness and intimacy. It needs to be emphasized, however, that these relational variables should not be considered the only factors involved in the etiology of the depression. Frequently, there is evidence for both family loading for depression and individual personality characteristics that predispose to depression. There is also nothing about the three relational structures that is specific to depression. The same patterns are found in couples referred because one spouse is resistant to treatment for alcohol abuse, phobic symptoms, or other individual syndromes. A brief clinical vignette may give a better sense of some of the issues involved at the interface of depressive syndromes and relational structures.

When the focus is turned from clinical to research observations, there is much that is congruent with the clinical observations as previously outlined. I will briefly describe several representative studies. Although most of these studies showed the adverse influence of diminished marital quality on the course of depression, several reported the buffering effect of a close, confiding relationship on the onset of depression. Thus, some studies showed that women or men at high risk for depression are less apt to develop depression if they participate in a supportive, confidential relationship (28, 29). In a similar vein, women who had lost their mothers in childhood were less likely to show elevated levels of state or trait depression if they received high levels of affection from their husbands (30). Weissman (31) documented a 25-fold increased risk of depression in both men and women who reported unhappy marriages, and Barnett and Gotlib (32) concluded on the basis of their extensive review that marital distress is both a consequence and an antecedent of depression.

A negative course of both depression and bipolar disorder has been associated with high marital expressed emotion (hostile, critical comments by the spouse) (33, 34). Long-term marital conflict was found to be one of a small group of variables associated with treatment-refractory depression (35). A rare prospective study of women with major depression demonstrated that women's reports of high levels of spousal support or high levels of marital conflict were the strongest pretreatment predictors, respectively, of positive or negative outcome of treatment (36).

Waring and colleagues (37) reported qualitative data suggesting that it is the presence or absence of intimacy in a marriage that is crucial in determining the severity of depressive symptoms.

NARROWING THE FOCUS

After the review of representative studies that suggest the impact of relational system characteristics on healthy individual development, the evidence for the healing impact of some adult relationships, and the role of marital factors in depressive syndromes, the question about crucial relationship processes can be asked. I suggest that a promising focus for future exploration is the factors that lead to the establishment of strong affective bonds and the clarification of the processes of repair when such bonds are temporarily dissolved.

Affective Bonds, Intimacy, and Empathy

In The Stone Diaries Carol Shields (38) writes that our lives are our life stories, and we need important others to listen to them (“Life is an endless recruitment of witnesses”) (p. 36). Toni Morrison (39) has Sixto explain in Beloved that he spends his weekend walking to see the Thirty Mile Woman because she is a “friend of his mind,” a person who helps him better understand what he thinks. These modern novelists speak to the importance of having someone who listens to our experiences and helps us to sort them out. To be able to enter into such a relationship usually requires the development of an affective bond. The prerequisites include a genuine and reciprocal liking for each other, mutual respect, and a two-way valuing and affirmation.

It can be argued that intimacy, the reciprocal sharing of vulnerabilities, is the hallmark of the strongest affective bonds. Its presence in a relationship grows out of repetitive conversations in which the following characteristics are found. 1) One partner discloses an emotionally charged experience with self, the other partner, or an outsider. 2) The other partner reacts by listening carefully and assisting in the exploration of the experience. He or she often responds empathically to the feelings expressed and does not change the subject, direct the conversation, impose meanings, give advice, or pass judgment. Rather, he or she attempts to understand and/or immerse himself or herself in the partner's subjective reality. 3) Such conversations often result in greater self-exploration and self-understanding and feelings of intense closeness. The participation of both partners may lead to what Weingarten (40, 41) has called the co-creation of meaning.

Most often, the experience being disclosed involves feelings of vulnerability. The speaker must take the chance of being hurt by the listener's response. Thus, some level of trust is essential, except in intimacy with strangers with whom one anticipates no further contact and, as a consequence, little likelihood of being hurt.

My understanding of an individual's empathic ability and its biological, developmental, and social antecedents has been described in some detail in other publications (42–44). The ability to accurately recognize and respond to another's feelings and, at times, to share those feelings is understood as movement into the other's subjective reality. Empathy, however, is not without its dangers, since for some it provokes underlying fears of engulfment. Havens (45, 46) has taught us that there is a language of empathy just as there is a language of distance and objectivity. Indeed, he suggests that in our choice of language we regulate the metaphorical distance between us and others. Anderson (47) has written of collaborative language and the importance of the avoidance of premature conclusions.

During the last few decades, instructional techniques from my seminars with beginning psychotherapists have been incorporated into my work with couples (15). As the brief clinical vignette indicates, my experiences with beginning therapists and couples in therapy suggest that most persons' empathic abilities can be enhanced. In those instances (with couples) in which the treatment fails, it is usually because of my inability to intervene successfully in the intense blaming-projecting process, a prerequisite for learning how to have intimate conversations.

Even under the best of circumstances, partners often fail to be empathic, and the conversation produces disappointment, distance, and lack of synchrony. There are no data that inform us about how often a spouse or therapist needs to respond empathically for the relationship to be successful. In one study of normal mothers, however, it was reported that a 30% empathic success rate was usual (48). It would be interesting to know whether a similar rate of empathic responding is associated with competent marriages and successful psychotherapy. Regardless of what the needed frequency of empathic responses may turn out to be, it is clear that couples must deal with empathic failures, and that the processes of repair are important in the establishment of an intimate relationship.

Repair or Conflict

There are a number of factors that influence whether repair or conflict follows an empathic failure. Although the failure can best be understood as interactional (i.e., involving both participants), some factors are more closely associated with one or the other participant. To begin with, the nature of the disclosed experience itself has an important influence on the outcome of the interaction. Experiences with and feelings about oneself or an outsider are easier to respond to empathically than are feelings about the listener. In the latter instance there is an increased likelihood of a defensive, distancing response. In teaching beginning therapists and couples, it is important to begin the learning process with feelings that are not directed at the listener. After the participants have learned something of the pragmatics of intimate conversations, we can then turn to feelings they have about each other with a lesser likelihood of failure.

Another factor involves the context in which the conversation occurs. More often than not, it is the partner who wishes to share the experience who selects the time and place. If inappropriate contexts are repetitively selected, there often are unconscious factors at work. In couples therapy an agreed-upon time and place characterized by privacy and an absence of distractions, when both partners are emotionally available to focus on the task, is important.

Although the storyteller is often the initiator of the interaction, the listener may take the first step by responding to or asking about the affective component of the experience. “How do you feel about that?” is often a response that signals the listener's availability to explore the storyteller's experience at the level of feelings. Thus, the failure of the interaction to move toward the intimate conversational level is understood as a shared failure.

The clarity with which the experience is related also is an important factor in determining the outcome. In enduring relationships partners may come to know that one or both introduce painful experiences indirectly and with much tentativeness and, as a result, they do not seize the initial story as necessarily the one that needs exploration. The “real” story is often the result of the interaction itself; it evolves out of the conversation rather than existing in final form before the conversation begins.

Another factor involves how readily available the storyteller's feelings of vulnerability may be. Often, the hurt or fear is hidden behind anger. Some individuals appear relatively fixated at the angry level, while others are much more in touch with the underlying hurt or fear. The responses of the listener may be crucial; denial (of the hurt or fear) often requires the assistance of a compliant other. In work with couples it is often necessary to take the lead in early sessions in helping the storyteller move to the level of vulnerability. Indeed, this approach is the major intervention used in the attempt to moderate conflict between the partners and to stop the blaming-projecting process.

Most of the experiences we wish to share involve relationships with self, partner, or an outsider. In some couples there is the gradual recognition of central painful relationship patterns. Such central relationship themes have been the focus of empirical research into essential processes of psychotherapy and, more recently, with nonclinical populations (49, 50). It appears that there may be a finite number of such themes, perhaps a dozen or so. In couples therapy it is often useful to help each partner become aware of his or her central relationship pattern and that of the other partner.

There is a tendency for the processes of either repair or conflict to become patterned. Couples who seek therapy often present with a pattern of conflict that can be understood as emanating from failed efforts both to connect with each other and to repair the disconnections. Generalization and escalation of the conflict are common. Such couples need to learn conflict management mechanisms, including techniques to prevent escalation. These can be taught in the marital therapist's office and may generalize to the real world. When they do not, and couples continue to live a life of conflict, it is safe to assume that one or, usually, both spouses may have underlying fears of closeness and intimacy, and explorations of this possibility are required. If this part of couples therapy does not succeed, individual therapy may be in order.

In recent years the importance of the establishment of a strong affective bond and the successful repair of its inevitable breakdowns has received empirical support from studies of marital interactions in which physiologic markers have been used as outcome measures (51–54). The studies focus on two relationship processes, confiding and conflict, that are closely related to the establishment of a strong affective bond (confiding) and the failure to repair its disruption (conflict). Taken as a whole, confiding and conflict have opposite effects on physical well-being. Confiding has positive effects on both hemodynamic and cellular immune functioning, whereas conflict has a negative impact on both.

CONCLUSIONS

Clinical observations and empirical research that support the role of interpersonal relationships in normal development, in healing relationships, and in the course of individual psychopathology have been briefly described. The focus of this presentation has been on those aspects of interpersonal relationships that can be observed directly, that is, on the interactional rather than the transferential. It seems increasingly clear that relationships can be altered by changing the way people talk to each other. It is possible to teach some couples intimate communication, with its emphasis on empathic processes and exploratory skills and the processes of repair of the all-too-frequent disconnections that are a part of life with important others. Indeed, a promising hypothesis for the development of a superordinate science of interpersonal relationships might well focus on the evidence that the repair of disrupted bonds is a common thread in successful infant-mother interactions, healthy marriages, and effective psychotherapy. Finally, the data presented in this article point to the need for a greater emphasis on relationship in our system of diagnostic classification.

In closing, here is the way in which the depressed attorney and his scientist wife ended our last session. He said that they did not wish to schedule any further regular appointments but preferred to move to an as-needed format. “I'm feeling so much better,” he said, “and our relationship is so different.” “Yes,” his wife added, “we're more connected and each of us is less alone.” “You may not need a therapist so much,” he concluded, “if you talk with each other sometimes like you do in therapy—kind of explore things. You become, well, sort of each other's therapist.”

Presented as the Benjamin Rush Award Lecture at the 150th annual meeting of the American Psychiatric Association, San Diego, May 17–22, 1997. Received July 7, 1997; revision received Sept. 29, 1997; accepted Oct. 17, 1997. From the Timberlawn Research Foundation. Address reprint requests to Dr. Lewis, Timberlawn Research Foundation, P.O. Box 270789, Dallas, TX 75227.