FORMER SPOUSES WHO ARE FRIENDS:

THREE CASE STUDIES

Journal of Social and Personal Relationships 1997, Vol. 14 no. 2: ps 207-222.

Carol Masheter

University of Utah

ABSTRACT

Within the larger context of modified analytic induction, each of the three case-study couples in this article was selected to address questions that emerged during the research process about the nature and development of friendships between former spouses. The cases confirmed earlier descriptive findings on these friendships, the importance of perceived intentions, and de-escalation as a path to friendship after ro- mantic involvement. The cases also yielded new findings. Modification and de-escalation emerged as additional paths to friendship. Partners of one couple disagreed whether they were friends, yet they were able to co-parent cooperatively. Also, focus of friendship varied; the partners of one couple fo- cused on their children, whereas the partners of the other two couples focused on each other. These findings bridge the di- vorce and friendship literatures and have relevance for other types of friendships.

KEY WORDS 8 case studies 8 friendships 8 post-divorce relationships

In spite of the fact that divorce has become commonplace, researchers have only begun to map the contours of the ex-spousal relationship without ac- tually knowing its intricacies (Ambert, 1989). In mapping some of these contours, several possibilities for relationships between former marital partners have come to light, including friendship. This article presents three case studies on post-divorce friendships that build on a previously pub- lished case study (Masheter & Harris, 1986). The case studies also confirm several previously published findings and present new findings about the nature of friendships between former spouses and how they develop. In doing so, these case studies bridge the friendship and divorce literatures, which historically have been separate.

Several scholars have considered the nature and meaning of friendship. In their recent review of the friendship literature, Blieszner & Adams (1992) concluded that the ways people sustain friendship usually revolve around continued similarity, rewarding communication and interaction patterns, and positive feelings. The same could be said of many sexual, ro- mantic, and marital relationships. An obvious question is how these re- lationships compare with friendships. Earlier writers (e.g. Davis & Todd, 1982) suggested that most people like their friends but love a single ro- mantic partner and experience passion for that person, which includes fas- cination, exclusivity, and sexual desire.

More recent studies have considered relationships that include elements of friendship and sexual/romantic involvement. The love-style research (Hendrick & Hendrick, 1989) has suggested that 'storge' relationships in- clude both friendship and love feelings. More recent research has con- firmed the importance of friendship as well as passion in young adults' romantic relationships (Hendrick & Hendrick, 1993). Schwartz (1994) re- cently has explored the benefits and risks of being married to one's best friend. Other examples of overlap between friendship and love are evident in heterosexual cross-sex friendships and same-sex friendships between gay men and between lesbians. Whereas heterosexual men and women tend to keep friendships and sexual relationships separate (Sapadin, 1988), some males, both gay and heterosexual, view sex as a primary means to achieve intimacy (Nardi & Sherrod, 1994). Compared to lesbians, gay men are more likely to have sex with casual and close friends but not best friends; whereas lesbians are more likely to say their best friend was once their lover or their best friend is their current lover (Nardi & Sherrod, 1994). Thus, blurring of distinctions also is evident in friendships between partners for which romantic or sexual involvement is an option.

Although first ignored and then pathologized (Kressel et al., 1978), schol- ars have recently described several forms of friendship between former spouses with fewer negative judgments. The form mentioned most often is the post-divorce family network that includes former spouses and new part- ners at family celebrations and holidays. Stacey (1990) has conducted in- depth case studies, including one on a post-divorce network in which the ex-wife's former husband served as photographer at her wedding to her second husband. Ahrons' (1994) 'perfect pals' and some of Hobart's (1991) remarried families developed similar networks.

Among Ahrons' (1994) divorced co-parents, 'perfect pals' included ad- ditional friendship possibilities. According to Ahrons, these former spouses called themselves good friends. Though their divorce usually had not been amicable, it had not been adversarial. They cooperated in co-parenting their children and usually spoke with each other at least once or twice a week. They asked about each other's work, activities, and even how each other was feeling. They consulted each other about decisions regarding fi- nances, occupation, needs of aging parents, or their own retirement plans. Some former spouses exchanged sex-typed favors, such as a home-cooked meal for home repairs. A few confided in each other about new intimate re- lationships (Ahrons & Rodgers, 1987). Ambert (1989) reported similar findings plus several financial ones; most mutually friendly divorced couples had children, could co-parent cooperatively, were homogeneous regarding SES (Socio Economic Status), whether higher or lower, and the ex-wives experienced no or little downward mobility.

I asked whether mutually beneficial and satisfying friendship between former spouses was possible without dysfunctional emotional dependency and without children as a shared interest. Thus, a 'test case' (Masheter & Harris, 1986) (Masheter's own term) was used to examine this possibility; the couple selected had no children, and both partners were involved in new intimate relationships for at least 1 year at the time of their interviews.

The test case not only confirmed that a mutually beneficial and satisfying friendship was possible, it also yielded other findings. The partners were anything but friends during the last years of their marriage, yet, after they divorced, they transcended their failed marriage and created a close friend- ship for the first time. Their relationship history was more complex than Adams & Blieszner's (1994) progression from stranger to acquaintance to friend. This couple's path included marriage, estrangement, divorce, per- sonal growth while apart from each other, and then co-creatiofi of a close friendship. Also, they were not 'just friends', in the sense of a secondary, less important relationship than a romantic one. Instead, they treasured their friendship as close, enduring, and irreplaceable, much like other kinds of close friendships (Wright, 1974).

The research just cited, including the author's, considered mutually friendly post-divorce relationships. Another study has examined accounts from individuals who described unilateral post-divorce friendships (one partner regarded the other as a friend, but the friendship was not recipro- cated), but these individuals' partners' accounts were unavailable (Masheter, 1994). However, Ambert (1989) collected data from both part- ners of divorced couples and has examined both congruent (e.g. each part- ner regarded the other as a friend) and incongruent relationships (e.g. one partner regarded the other as a friend, but the other had different feelings, such as preoccupation, dislike, or indifference ). Ambert reported that when one ex-spouse harbored negative feelings for the other, co-parenting was not likely to succeed.

Regarding prevalence, Ambert (1989) reported that for 13 percent of couples in her study both partners reported having a 'generally friendly re- lationship' 6 years after separation. Ahrons (1994) reported that, 5 years after divorce, 12 percent of co-parents in her study were 'perfect pals', e.g. they called themselves good friends, they 'trusted one another, asked for advice, and helped each other, as friends do' (p. 53). In a recent survey, I (Masheter, in press) found that nearly one-third of respondents reported having contact with the ex-spouse, because 'we are friends', 1 year after divorce.

Differences in these cross-sectional data have suggested that menoly feelings between former spouses diminish over time (Ahrons & Rodgers, 1987; Ambert, 1989). However, other factors may influence former spouses' involvement more directly than simply passage of time. Ahrons (1994) has suggested that friendly feelings diminish as the partners' lives differentiate, such as when one or both partners remarry, or children leave home. Ambert (1989) reported that parents were more likely to be either friendly or unfriendly, whereas non-parents more likely to be indifferent or have no relationship. However, friendships exist between former spouses without children and with new intimate relationships (Masheter, 1991, 1994; Masheter & Harris, 1986). They warrant study in order to understand friendship dynamics apart from the potentially confounding influences of children and remarriage.

Friendship between ex-spouses still is a relatively unexplored phenom- enon, and less has been published on how these friendships develop. Ambert (1989) emphasized personality characteristics and interactional history. If both partners were 'reasonably adaptable' and had 'mature per- sonalities', they were more likely to have a mutually friendly relationship after divorce. Couples who had less conflict during their marriage also were more likely to have friendly feelings toward each other than couples with \ more intense pre-divorce conflict. For Ahrons (1994), interactional history r played a more directly positive role; 'perfect pals' often reported having been best friends during their marriage and still called themselves good friends after divorce. Yet how former spouses got from failed marriages and inimicable divorces to friendship still was unclear .

As mentioned earlier, the test-case (Masheter & Harris, 1986) partners escalated to friendship for the first time after divorce. However, research on premarital romantic relationships has suggested that partners 'de-esca- late' to an earlier 'script' for friendship, which existed before or during the romance (Metts et al., 1989). Other scholars have reported that some couples use specific strategies, such as relationship talk (Baxter, 1987), e.g. 'this relationship isn't working; let's talk', and specific requests for de-esca- lation (Banks et al., 1987), e.g. 'let's just be friends', that increase the like- lihood of a post-romance friendship. Thus, de-escalation and escalation invited further exploration as paths to friendship between former spouses.

Following from these previous findings, the case studies presented in this article are intended to explore variations in meaning, function, and path, ad- dressing some of the cognitive, affective, and interactional features of dyadic friendships within Adams & Blieszner's (1994) conceptual framework.

Method

Because friendships between former spouses are relatively unexplored, ad- equate theories to study them currently do not exist. Rather than rely on exist- ing theories and assumptions in the divorce literature that may pose inappropriate questions ( e.g. in what ways are friendships between former spouses dysfunctional?), researchers who use modified analytical induction (Bogdan & Biklen, 1992) often begin with initial, tentative questions, similar to researchers who use grounded theory (Gilgun, 1995). Then, in the first stage of research, researchers select one or more cases, and collect and analyze data to address the initial questions. During or after this first stage, new questions often arise. For example, are escalation and de-escalation the only paths to friendship between former spouses? In the second stage, researchers may modify case selection criteria, data collection and analytical methods to address the new ques- tions. In turn, the second and additional stages may generate more questions and modifications to be incorporated in subsequent stages.

Also, researchers who use modified analytic induction actively seek to dis- confirm emerging hypotheses through 'negative-case' analysis, that is, analysis of cases that may disconfirm emerging hypotheses and that can add variability to the sample (Gilgun, 1995). For example, one emerging hypothesis from the test case on a mutual friendship was that partners must cooperate to carry out effective desired intentions (Masheter & Harris, 1986). A negative case could examine a unilateral friendship to see whether former spouses who disagree about the nature of their relationship can cooperate. Using recurrent question generation and negative case analysis, proponents of modified analytic induc- tion have de-emphasized universality and causality and have emphasized the development of descriptive hypotheses that identify patterns of behaviors and interactions, unlike previous approaches to grounded theory (Gilgun, 1995). In turn, these descriptive hypotheses and patterns generate new theories and con- cepts grounded in qualitative data about the phenomenon under investigation.

Within the wider context of modified analytic induction, I selected one cou- ple for each of the three stages of research to address questions that emerged in the previous stage. For example, one of the tentative questions that guided the selection and analysis of the test case was: can former spouses co-create a mutually beneficial friendship apart from a shared interest such as their children? Analysis of the test case answered this question in the affirmative and raised new questions about the nature of unilateral friendships and whether partners in such friendships could cooperate. Thus, the first-stage couple was a negative case selected to address these new questions. Findings and research questions generated from the first-stage analysis then guided the selection and analysis of the second-stage couple, findings and questions which in turn guid- ed the selection and analysis of the third-stage couple. Thus, each couple was selected sequentially; the findings and research questions from each stage pro- vided the selection criteria for the next stage's case.

Several researchers have noted the need for interview data from both partners of divorced couples (e.g. Ambert, 1989; Ahrons & Rodgers, 1987). Accordingly, only couples for which both partners volunteered to participate in interviews without the partner present were selected. The partners of these couples were members of two larger surveys (Masheter, 1991, in press) in which all respon- dents (N =265 + 232) were invited to participate in additional interviews; a sub- set of respondents accepted this invitation (n = 91 interviewees from both surveys, including 9 couples of whom both partners were interviewed). Each of the three couples presented in this article was selected from this larger group of couples, because each selected couple addressed questions abol1t friendship that emerged during the three-stage research process described later in this article. The specific questions and criteria are described within their respective stages.

With each interviewee's written permission, I conducted tape-recorded research interviews that lasted 1-2 hours. Each interview consisted of two parts: (1) a narrative and (2) episode analysis. Three questions guided the narrative: (1) How did you come to be divorced? (2) What has that experience been like for you? (3) If you have contact with your ex-spouse, what has that been like? Interviewees were encouraged to answer these questions with few interruptions except points of clarification. Because friendship has different meanings for dif- ferent people (Blieszner & Adams, 1992), the six interviewees described 'friendship' with the ex-spouse in their own terms without restrictions, such as whether the friendships were mutual, beneficial, or enduring.

Using episode analysis, a methodology informed by symbolic interaction (Blumer, 1969), developed by communication theorists (Cronen et al., 1982), and adapted for research on post-divorce friendships between ex-spouses (Masheter & Harris, 1986), I invited the interviewees to recount conversations or episodes with the ex-spouse from before and after the divorce. As described elsewhere (Masheter, 1994), the interviewees were invited to interpret each episode, line by line, in terms of their own intentions and those of the ex- spouse, effectiveness of each intention, satisfaction with each intention, and sig- nificance of each episode for the relationship. For the first-stage couple, the rationale for eliciting two episodes was to compare samples of interaction before and after divorce. For the second- and third-stage couples, the rationale for eliciting up to four episodes, from 'good and bad times' before and after di- vorce, was to determine whether friendship existed during the couple's mar- riage as well as after their divorce. Previous research using episode analysis (Masheter & Harris, 1986) has demonstrated that partners cooperate in order to carry out effective desired intentions, such as promoting a satisfying mar- riage. In contrast, effective undesired intentions, such as perceived effective re- jection of one partner by the other, have negative implications for the couple's relationship, as do ineffective desired intentions, such as perceived failure of one partner to enlist the other's cooperation.

Student volunteers and paid assistants transcribed each interview verbatim, following Riessman's (1993) retranscription method, which includes all utter- ances (such as repeated words, false starts, and 'urns') and detectable nonverbal communications (such as pauses, laughter, weeping, sighs) from both the interviewee and the interviewer, However, to balance accuracy with brevity, the quotations frominterviews included in this article follow Riessman's (1993) rough transcription method, which uses the interviewee's own words but ex- cludes communications that obscure rather than clarify the interviewee's mean- ing (see Table 1 for examples of retranscription and rough transcription).

TABLE 1

Retranscription and rough transcription

Rough transcription

Bob: How is our relationship now? She picks up my mail. Maybe it's just taking advantage of each other to our advantage. If it's advantageous, then we'll take advantage of each other because it is allowed. On both sides.

Retranscription

Bob; Urn, urn, urn, urn. How is our relationship now? (pause) Well, she picks up my mail. I, well, I don't know. Maybe it's just, urn, taking advantage of each other to our advantage.

Interviewer; Ah. That's a nice way to put it.

Bob; Maybe that -if it's advantageous then we'll take advantage of each other because it is allowed. On both sides.

I then used Jones' (1984) approach to narrative analysis to identify themes in each interviewee's transcribed narrative and to cluster these themes into more general concepts. For example, alcoholism and violence were themes that were clustered into the more general concept of one ex-wife's perceived causes for her divorce. I then took a meta-perspective of the interviewees' analyses of the partners' recounted episodes to compare pre- and post-divorce episodes and to compare episodes from both partners of the same relationship. Findings from these three cases were then compared with findings from previous research.

Stage 1: cooperation in a unilateral friendship

The couple for the first stage of this article's research was analyzed as a nega- tive case to challenge the findings from the test case (Masheter & Harris, 1986) which illustrated escalation to a mutually close, beneficial friendship between former spouses who had no children. Criteria for selection of the negative case included: (1) the partners explicitly described their friendship as unilateral, and (2) the partners were parents of minor children. Based on a first reading of each partner's transcribed interview data, Ann and Amos (pseudonyms, as are all names for interviewees), respondents from a larger survey (Masheter, 1991), met these criteria.

Analysis of this negative case addressed the following questions: (1) Can partners in a unilateral friendship cooperatively co-parent? (2) Did the friendly partner escalate or de-escalate to friendship? and (3) Would the friendly part- ner be dysfunctionally dependent on the other partner?

Both Ann and Amos were in their late forties, had three children, and div- orced after 27 years of marriage. Both had held low-paying jobs during and after their marriage. Their second child, a daughter, developed a life-threatening condition that left her unable to walk after age 7.

Amos' interview took place several months before Ann's. According to Amos' narrative, they 'did what everybody did in the 50s and got married right out of high school'. He claimed that he had not had a drink for the past 15 years, though he reported completing a 30-day program of treatment for alco- holism after the divorce. Amos described several instances in which he ex- pressed his anger to Ann before and after the divorce and blamed Ann's increasing independence for the divorce -'she used to follow my lead, (then) she wanted her freedom, her own space'. In his pre-divorce episode, Ann had said that she no longer wanted to have dinner late, after Amos had returned from 'huntin' or fishin' or whatever I was doin' " because 'she got too hungry waiting' for him. In response, he yelled, 'You dumb bitch, why didn't you say so!', then he 'stormed out' and 'roared off' on his motorcycle. Amos felt effec- tive and satisfied in expressing his anger -he added: 'I can use my eyes like daggers' - but he was dissatisfied with what he perceived to be Ann's effectively 'doing what she wanted'.

In his post-divorce episode, Amos claimed that Ann had said,"I have you to thank for getting me to stand on my own feet', which he interpreted as a com- pliment. However, Amos was unsure whether her compliment was effective, because he responded, 'you turned it around and threw me out! I had no choice!'. He intended this remark to be a mixture of anger, resentment, and humor, which he was unsure was effective. Amos had mixed feelings of satis- faction and dissatisfaction about both his own intentions and his perceptions of hers. Regarding their post-divorce relationship, Amos said, 'She says we're friends. Hell, she's no friend of mine! She threw me out! How can we be friends?!'. Yet Amos mentioned several family gatherings at which 'everybody got along okay', and he expressed admiration for Ann's dedicated care of their disabled daughter .

In her narrative, Ann claimed: '1 followed all the rules in ladies' magazines: please your husband in bed, keep a nice, neat home, dress attractively, don't make waves. I was very angry when these rules didn't work for me. I felt lied to by society. The rules said nothing about communication of our real wants and needs. Amos and I knew nothing about communication. Neither did our par- ents. We were very ignorant about marriage'.

Ann eventually divorced Amos because of his drinking, which he tried un- successfully to stop, and his abusive behavior; once he threw a knife at her. After their daughter's diagnosis, 'things were better for a while. Amos did try. But most of the worry about [our daughter] fell to me'. The marriage deterio- rated further, Ann became exhausted, depressed, and wanted to die. Amos told Ann that he had started drinking again, because he knew that would make her divorce him, yet he angrily resisted the divorce. Though Ann was fearful about her future as a divorced mother with a low-paying job, a disabled child and two other children, she eventually filed for divorce.

In her pre-divorce episode, Ann described a scene in which Amos criticized their younger adolescent daughter's friends, the daughter 'talked back', and an argument escalated between father and daughter, just as Amos' mother tele- phoned. Ann told Amos' mother 'I'll get back to you', intending to protect her nuclear family's privacy. However, before she could hang up, Amos grabbed the receiver, thrust it into his screaming daughter's face, then took it back and said: 'See, Mom, this is what I've been telling you about. This is what I have to live with'. Ann felt that she effectively prevented physical violence by standing behind her daughter, putting her hands on the daughter's shoulders, and say- ing: 'Let's go for a walk'. However, Ann claimed that this episode was typical of her inability to convey effectively her fear and exhaustion, which dissatisfied her, as did Amos' effective expressions of anger.

For her post-divorce episode, Ann recounted a conversation with Amos during a 2-hour drive to help their younger daughter move home from college. Ann was nervous about being alone with Amos for the first time in over a year , but she wanted to give a graduation party for this daughter and wanted Amos to ask his family to help, intending to be honest about her own needs. Amos re- sponded: 'Sure. I don't blame you for not wanting to do everything'. Ann felt effective and found Amos effectively cooperative, both of which satisfied her. She also reported that this episode was atypical compared to the 'tense' inter- actions during and just following the divorce.

Ann ended her interview by saying: '[After the graduation party], I realized that relationships do not have to end. ...It's not necessarily the fantasy you grew up with. But you can work on the relationship even after divorce, and have something of substance. In some ways it's better for us now than when we were married. Amos and I, even after all we have been through, are still [our children's] mother and father. We still have that connection'. Ann then spoke of improvements in Amos' relationships with their children.

From the author's perspective, Ann and Amos disagreed about many of the major features of their relationship. For Ann, marriage was a series of trials - her physically challenged daughter's need for care and her husband's alco- holism and violence -whereas for Amos, marriage was a battle of wills. Ann and Amos also disagreed about the causes for their divorce. Ann claimed Amos' alcoholism and abuse compelled her to divorce him, but Amos claimed that he had stopped drinking (though his account was inconsistent) and blamed Ann's increasing independence for their divorce. Finally, Annand Amos dis- agreed about the very nature of their post-divorce relationship -whether they were friends.

Returning to the research questions, Amos and Ann's unilateral friendship included several incidents of cooperative co-parenting that each partner de- scribed independently. These findings contrasted with those from Ambert's (1989) study, which reported that incongruent post-divorce relationships tended not to be cooperative. Regarding path to friendship, Ann appeared to have escalated to a friendship with Amos after their divorce centered on their children. Like the test-case couple (Masheter & Harris, 1986), Ann trans- cended a failed marriage and difficult divorce. In contrast, Amos neither esca- lated nor de-escalated to friendship; instead, over nearly 3 years after divorce, he remained angry, resentful, and confused.

Earlier literature has suggested that friendly feelings toward the ex-spouse are evidence of dysfunctional dependency (Kressel et al., 1978). However, as the friendly partner, Ann showed no evidence of this. In fact, she appeared to be less dependent emotionally on Amos than he was on her. Amos still was in- vested in being angry at Ann, which can be a kind of dependency (Johnson & Campbell,1988).

Stage 2: friends who describe their friendship differently

The test case (Masheter & Harris, 1986) and the negative case showed two dif- ferent possibilities for friendships between ex-spouses. The test-case partners experienced their pre-divorce episode very differently, but their interpretations of their recounted post-divorce episodes were nearly identical, and they were able to co-create a mutually beneficial friendship without children and dys- functional dependency. In contrast, Ann and Amos disagreed about many major features of their relationship, yet they cooperated in important ways for their children. Though both dyads cooperated, the focus of their relationships differed. The test-case partners focused on each other. In contrast, the nega- tive-case partners focused on a shared interest, their children, outside their dyad. The differences between these two couples invited further questions: (1) Could partners differ about some post-divorce events and/or interpretations and still have a mutually beneficial friendship? and (2) Would their focus be on each other or on a shared interest?

To address these questions, a second couple was selected, based on a first reading of the partners' transcribed interviews in which each claimed to be friends with the other, but their descriptions of their friendship differed. Like Ann and Amos, Betty and Bob were volunteers for interviews in a divorce sur- vey (Masheter, 1996).

Bob and Betty were in their late thirties and were employed intermittently in low-paying jobs before and after their divorce. Like Ann and Amos, Betty and Bob had financial and parenting difficulties. Unlike Ann and Amos, Betty had adolescent fraternal twins, a son and daughter, from a former marriage who lived with her and Bob during their marriage, but Betty and Bob had no children from their union. Bob was interviewed several weeks before Betty.

In his narrative, Bob claimed that they agreed to divorce after Betty 'went out on me [had sex with another man] and I couldn't deal with it. It messed up our sex life. If I didn't have good feelings, I couldn't just turn on like a ma- chine'. Bob also claimed that conflict over Betty's children contributed to their divorce, Bob reported: '1 am really into honesty, and I gave the children per- mission to speak openly' as they did with their friends, of which Betty disap- proved. Also, Bob claimed that the daughter was jealous of him, because Betty 'would talk to her daughter like an adult' before Bob and Betty married.

Later in his interview, Bob claimed that Betty had accused him of commit- ting homosexual acts after they married. He had told her that he was gay before they married, but he claimed, '1 was true to her, even though I had opportunity not to be. One of my friends that she had sex with kept bringin' that up to her [Bob's alleged homosexual acts], when I wasn't around. And now I found out that it was because he wanted to have sex with her',

For his pre-divorce, 'good-times' episode, Bob recounted 'blowing money' with Betty, intending to give her a sewing machine that she wanted, which they did effectively and about which he felt satisfied. For him, this episode was typi- cal of the good times in their marriage. For his pre-divorce, bad-times episode, Bob described the time when he smashed Betty's television. He explained that he was 'television addicted' and did not like having it on, but Betty's children watched a lot of television. 'One time I totally lost it. I felt it was too much tele- vision. So I started to say, "look, I want the TV off, I'd like some silence". And they'd say "yea, yea, yea", and nothing would ever happen. Betty'd usually side with them. And I beat that TV up and killed it. It was pretty out of character'. Bob claimed that he felt ineffective in getting Betty's children to turn off the television but effective in expressing his anger, both of which were not satisfy- ing to him.

When asked about his post-divorce relationship with Betty, Bob responded: 'We decided to be friends. ., It's taking advantage of each other to our advan- tage because it is allowed. On both sides. We can talk to each other in ways that she doesn't feel she can express herself to her current boyfriend or other people. She can say anything she wants, I don't care, and she knows. I can say anything to her, and that's just what developed. After the divorce, we sat down together, and we talked about I miss you, and I know it's over, but I really need a friend, and I really enjoy your friendship. It was a mutual thing, and it's not a sexual thing'.

For his post-divorce episode, Bob recounted a typical scene in which he and Betty both said they missed each other and hugged. For him, these actions sig- nified effective expressions of their friendship, which he found satisfying. Bob added: '1 still love her as a person and friend. ..I'd do nearly anything [for her], but not sexual. ..I give her [emotional] support. ..We were friends long before we had sex'. Bob also described several examples of how he and Betty help each other through exchange of goods and services. He adopted her dog when she no longer could care for it, they shared a post-office box that he paid for, she cooked his lunches, he contributed food and food money, he let her use his freezer and washer, and he may move in with her and her boyfriend as 'friends, not for sex'.

In her narrative, Betty reported that they agreed to divorce, because she could not tolerate his homosexual behavior after their marriage and Bob and her daughter did not get along. Bob had told her before they married that he was gay but wanted to 'go straight'; Betty thought that she could help him. However, Betty's male friends told her that Bob had sex with them after the marriage, and her son told her that Bob made a 'pass' at him.

Regarding Betty's children, Betty claimed that her daughter tried hard to get close to Bob, but he first rejected her, then physically abused her. Bob wanted to give Betty's children 'freedom of speech', which meant they could use 'cuss words'; Betty disapproved. Bob 'smoked weed' with them without telling her. When Betty's daughter became pregnant at age 14, Bob wanted her to get an abortion, but Betty disagreed. In an argument, Bob once kicked the daughter in the thigh so hard that she had trouble walking for several days.

Betty reported that she and Bob both 'smoked weed' before and after the di- vorce, she had difficulties with depression and panic attacks and currently was unable to hold a job. Betty claimed that Bob had Klinefelter's syndrome (an abnormality of the twenty-third chromosome, XXY), was 'brilliant' in some ways, 'retarded' in others, and still used 'crack'.

For her pre-divorce, good-times episode, Betty mentioned several enjoyable drives, cook outs, times when she, Bob, and her children 'all danced around the house like crazy people'. The episode Betty chose to analyze was a time when Bob intended to brush something from her face, but she interpreted his gesture as intent to strike her. She recoiled sharply from his gesture then threatened to kill him if he ever hit her. They had a long discussion about her abusive first husband, during which Bob effectively reassured her that he would never hit her, which was very satisfying to her. Betty regarded this episode as a typical example of their marriage when it was going well.

For their pre-divorce, bad-times episodes, Betty (like Bob) recounted the time Bob smashed her television. Betty saw herself and her children as using television to relax. Though she had effectively expressed fear and anger in re- sponse to Bob's effective expression of anger, Betty was dissatisfied. She noted that this episode typified the bad times in their marriage, though Bob's violent behavior was unusual. The incident also seemed to be a turning point in their relationship. 'That's when we started arguing, because I started bringing it up [Bob's alleged homosexual behavior]. He started putting my kids down for watching TV so much and me allowing it, and if he's going to throw shit in my face, I'm fixin' to throw some right back in [his]'.

For her post-divorce, Betty described a typical conversation with Bob about her depression, panic attacks, and inability to work. Betty felt effective in ex- plainingthese problems to him and regarded Bob as effective in supporting her emotionally, both of which were satisfying. When describing their post-divorce friendship, Betty claimed that she valued Bob's friendship 'dearly' and he was 'the best friend I ever had. ..[M]e and Bob will probably always be friends. It wouldn't surprise me if we don't end up sharing a place together or something someday', though sexual relations with Bob no longer interested her.

Returning to the stage-two research question about partners' agreement con- cerning their relationship, Betty's and Bob's general descriptions were the same, but the specifics differed considerably. Each partner accused the other of infi- delity, but Bob denied being unfaithful, whereas Betty did not mention Bob's al- legations of her infidelity. Both reported problems with Betty's children and Bob's 'honesty', but Bob claimed that the daughter was jealous of him, whereas Betty claimed that Bob abused her daughter and made sexual advances toward her son. Both Betty and Bob valued their friendship after divorce, which they both claimed was no longer sexual. However, Betty emphasized the emotional support Bob gave her, whereas Bob also mentioned their freedom to say any- thing to each other and their mutual exchange of goods and services.

Returning to the second-stage question about focus, Betty and Bob argued about Betty's children during their marriage, but after their divorce, their friendship focused on each other. Regarding focus, they were more similar to the test-case couple and contrasted with Ann and Amos, whose post-divorce focus was on their children and family celebrations. Yet Betty's and Bob's friendship differed from the test-case partners' financial independence and em- phasis on personal growth (Masheter & Harris, 1986). Betty and Bob inter- acted frequently and depended on each other in more day-to-day ways than the test-case partners, who contacted each other rarely but valued the mutual con- firmation that their friendship provided. Though the focus of these two mutual friendships was within each dyad, the manifestation of this focus differed.

Stage 3: former spouses who became confidantes

Former spouses who have no children yet regard each other as confidants and confirm each other's personal growth have received less attention in the litera- ture than post-divorce co-parents and family networks. Thus, the third-stage couple was selected to address the following questions: (1) Did these partners escalate or de-escalate to their current friendship? and (2) Are the partners dysfunctionally dependent on each other? Selection criteria for the partners in- cluded: (1) they had no children and (2) both partners reported being close friends who confided in each other and confirmed each other's growth. Like the partners of the other two couples, Cathy and Carl were respondents in a larger survey (Masheter, in press), and a first reading of their transcribed interviews indicated that they met the above criteria.

Cathy and Carl were both professionals in their mid-thirties. Cathy was in- terviewed several months before Carl. According to Cathy, they divorced be- cause Carl 'just closed down emotionally and interactively. ..I'd like people to share their feelings, what's really going on, and I think he was having difficulty even knowing what those were'. Cathy met a man who gave her the attention she did not get from Carl and had an affair with him. For Cathy, the affair was a cry for help. However, when the affair failed, Cathy and Carl decided to di- vorce and to be friends who were committed to their own and each other's per- sonal growth -'to be all we could be'. During and after their divorce, Cathy and Carl continued to travel together, an activity that they enjoyed during their marriage and that promoted personal growth.

For her pre-divorce, good-times episode, Cathy described a trip in which they both intended to be spontaneous and have fun, unusual behavior for the usually reserved Carl. Cathy felt effective and satisfied with their intentions. For her pre-divorce, bad-times episode, Cathy recounted an interaction that typified her frustration about Carl's avoidance. Cathy told Carl that he was 'in- visible. ..like a whisper' in social settings, intending to point out his avoidant behaviors, which she did effectively. Cathy perceived Carl's response, a de- scription of how a waitress at a restaurant 'didn't see' him, as effective ex- pression of his hurt and shock about his avoidant behaviors. She perceived his next remark about his envy of the social attention she received as his effective expression of honesty. Cathy was satisfied with both her own and Carl's effectiveness in carrying out their respective intentions, but she was dissatisfied with what she viewed as 'his old pattern' of recurring avoidant behaviors.

In her post-divorce episode, Cathy recounted a conversation about her de- sire to spend time alone with her current boyfriend and time alone with Carl as friend and confidant. She also described a long-distance telephone conversa- tion during which Carl confided in her about a new, failed relationship. For Cathy, this episode typified their irreplaceable, mutually beneficial friendship in which they effectively confided and supported each other's personal growth.

In Carl's narrative, he reported that Cathy wanted more emotional closeness than he did, they traveled well together, and after their divorce they had de- cided to be close emotional friends who confirmed each other's personal growth and could confide in each other about new relationships. However, Carl felt 'devastated' when Cathy moved in with a male friend after her affair failed. Carl had formed a strong emotional bond with another woman, and he filed for divorce. Carl concluded: 'when it was clear we couldn't be together married, we decided to be friends. I think we're lucky. Cathy thinks we were together in past lives; we both do. I've learned a lot from her. She's a great teacher. We ident- ified the "ground relationship", our eternal relationship'.

For his pre-divorce, good-times episode, Carl described a trip to the desert in which he encouraged Cathy to paint, an activity she enjoyed though she lacked confidence as an artist. He felt effective and satisfied with his intentions to support Cathy's personal growth. For his pre-divorce, bad-times episode, Carl recounted a typical post-divorce episode in which Cathy expressed won- der and delight upon seeing a butterfly. In response, Carl effectively expressed skepticism and avoided Cathy by remaining silent, which was dissatisfying to him. In his post-divorce, doing-well episode, Cathy and Carl congratulated each other's personal growth, and Cathy confided in Carl about a failed relationship with a new man. Carl reported that both he and Cathy were effective in carry- ing out these desired intentions, which was satisfying to him.

Cathy's and Carl's narratives were strikingly similar, especially since their in- terviews occurred 6 months apart and Carl then lived in another continent. Though their specific recounted episodes differed, Cathy and Carl indepen- dently described serving as each other's confidant and supporting each other's personal growth. The major differences in their narratives were the significance of Cathy's affair and who decided to divorce.

Returning to the third-stage research question about path to friendship, neither de-escalation nor escalation described this couple's relationship devel- opment adequately. Instead, they modified a pre-existing friendship. They kept some elements of their pre-divorce friendship, such as traveling together and promoting personal growth, and they added new elements, such as confiding in each other about new intimate relationships. Returning to the research ques- tion about dependency, neither Cathy nor Carl was financially or emotionally dependent on the other. However, both manifested a kind of secure attach- ment, using each other as a 'safety base' to explore other relationships and then return to each other for confiding and confirmation.

Discussion and future research directions

These case studies supported several previously published findings, yielded new findings, and raised additional questions that invite further research and conceptualization. The first-stage couple confirmed the importance and resilience of post-divorce networks (Ahrons, 1994; Stacey, 1990). The sec- ond-stage couple demonstrated that Metts et al.'s (1989) de-escalation to friendship is possible after divorce as well as after premarital romantic in- volvement. Episode analysis from all three couples confirmed earlier find- ings (Masheter & Harris, 1986) on the importance of partners' cooperation for carrying out effective desired intentions and the negative implications of effective undesired intentions and ineffective desired intentions for part- ners' relationships.

This article's case studies also yielded new findings. The third-stage couple illustrated a manifestation of friendship between former spouses, mutual confirmation of personal growth, that other researchers have not described. This couple also showed that close emotional friendship be- tween former spouses is possible without shared parenting responsibilities and can persist after one or both partners establish new intimate relation- ships. Though Ambert (1989) concluded that partners in unilateral friend- ships tended not to co-parent cooperatively, the first-stage couple provided a vivid exception. Also, these two couples differed in the focus of their friendship; the partners of one couple focused on each other, whereas the partners of the other focused on a shared interest outside their dyad. In ad- dition to Metts et al.'s (1989) de-escalation, this article presents evidence for two other paths to friendship after divorce: (1) escalation and (2) modi- fication. Thus, different paths to friendship and focuses within and outside the dyad suggest new aspects of friendship not explicitly addressed in ex- isting conceptual frameworks (e.g. Adams & Blieszner, 1994).

Though the findings presented in this article shed some light on friend- ships between former spouses, several limitations warrant caution. First, the small number of cases presented here undoubtedly does not reflect maximum variation. Second, generalizations based on this small number of case studies are unwarranted. Third, samples of couples in which both part- ners volunteered to participate in research interviews may not represent adequately the population of couples who have friendly relations after di- vorce.

However, these limitations, as well as the findings, suggest several direc- tions for further research. More extensive sampling could map more sys- tematically the spectrum of friendships of which the cases presented in this article have provided only intriguing glimpses. Such sampling could include unilateral and mutual friendships, partners who focus on each other and partners who focus on a shared interest outside their dyad, partners with and without children, partners with and without (new) intimate relation- ships, partners whose friendship developed through escalation, de-escala- tion, or modification, partners who simply 'get along' (Le. do not fight anymore), partners who behave civilly at family gatherings, partners who exchange goods and services, partners who confide, support, and confirm. After determining which cells actually contain members and characterizing these cells, more quantitative approaches may be appropriate, such as sur- veys and analysis of archival data, such as letters and diaries.

Additional questions invite further investigation through qualitative and quantitative approaches. Specifically, what does being 'friends' with one's former spouse or lover mean to people of various subcultures, cohorts, and life circumstances, such as education, income, parental status, and sexual orientation? Why do some post-divorce friendships persist, even without children and after formation of new intimate relationships? Are unilateral friendships inherently unstable? Do different manifestations, focuses, and paths to friendship reflect different needs or personal ideologies about re- lationships? What distinguishes beneficial friendships from dysfunctional ones? Could clinicians and mediators teach former spouses -and partners of other dyads-to interact constructively rather than destructively? True to the tradition of many types of qualitative research, this article has raised more questions than it had addressed. However, further exploration of these questions could extend and clarify existing theoretical frameworks for friendship (e.g. Adams & Blieszner, 1994).

Ambert (1989) concluded that expectations for a civil and cooperative post-divorce relationship may be unrealistic and mere wishful thinking. Indeed, friendship may not be realistic for many divorced couples. However, friendship between ex-spouses, while hardly the new paradigm for post-divorce relationships, is not as rare as formerly assumed and can provide instructive examples of complex relationship development. This article's findings have indicated that at least for some divorced persons 'being friends' with the former spouse is quite possible and sometimes ben- eficial.

REFERENCES

Adams. R.G. & Blieszner. R. (1994) 'An Integrative Conceptual Framework for Friendship Research', Journal of Social and Personal Relationships 11: 163-84.

Ahrons, C.R. (1994) The Good Divorce. New York: HarperCollins.

Ahrons, C.R. & Rodgers. R.H. (1987) Divorced Families. New York: W.W. Norton.

Ambert, A-M. (1989) Ex-Spouses and New Spouses: A Study of Relationships. Greenwich. Cf: jAl Press.

Banks, S.P., Altendorf, D.M., Greene, j.O. & Cody, M.j. (1987) 'An Examination of Relationship Disengagement: Perceptions, Breakup Strategies and Outcomes', Western Journal of Speech Communication 51: 19-41.

Baxter, L.A. (1987) 'Cognition and Communication in the Relationships Process', in R. Burnett, P. McGhee & D. Clark (eds) Accounting for Relationships: Explanation, Representation, and Knowledge. London: Methuen.

Blieszner, R. & Adams, R.G. (1992) Adult Friendship. Newbury Park, CA: Sage.

Blumer, H. (1969) Symbolic lnteractionalism: Perspective and Method. Englewood Cliffs, Nj: Prentice'Hall.

Bogdan, R.C. & Biklen, S.K. (1992) Qualitative Research for Education. Boston, MA: Allyn and Bacon.

Cronen, V.E., Pearce, W.B. & Harris, L.M. (1982) 'The Coordinated Management of Meaning: A Theory of Communication', in F.E.X. Dance (ed.) Human Communication Theory. New York: Harper & Row.

Davis, K.E. & Todd, M.J. (1982) 'Friendship and Love Relationships', in K.E. Davis & T.O. Mitchell (eds) Advances in Descriptive Psychology, pp. 79-122. Greenwich, Cf: JAI Press.

Gilgun, J.F. (1995) 'We Shared Something Special: The Moral Discourse of Incest Perpetrators', Journal of Marriage and the Family 57: 265-81.

Hendrick, C. & Hendrick, S.S. (1989) 'Research on Love: Does It Measure Up?' Journal of Personality and Social Psychology 56: 784-94.

Hendrick, S.S. & Hendrick, C. (1993) 'Lovers as Friends', Journal of Social and Personal Relationships 10: 459-66.

Hobart, C. (1991) 'Relationships Between the Formerly Married', Journal of Divorce and Remarriage 14: 1-23.

Johnson, J.R. & Campbell, L.E.G. (1988) Impasses of Divorce. New York: Free Press.

Jones, C.S. (1984) 'Training Manual for Thematizing Interview Protocols Phenomenologically, Technical Report #1 " Phenomenological Psychology Research Group, H.R. Rol1io (Director).

Kressel, K., Lopez-Morillas, M., Weinglass, J. & Deutsch, M. (1978) 'Professional Intervention in Divorce: A Summary of the Views of Lawyers, Psychotherapists, and Clergy', Journal of Divorce 2: 119-55.

Masheter, C. (1991) 'Post-Divorce Relationships Between Ex-Spouses: The Roles of Attachment and Interpersonal Conflict', Journal of Marriage and the Family 53: 103-10.

Masheter, C. (1994) 'Post-Divorce Relationships Between Ex-Spouses: Examples of Dialectic Relationship Development', in R. Conville (ed.) 'Structure' in Communication Studies, pp. 83-101. Westport, Cf: Praeger.

Masheter, C. (in press) 'Healthy and Unhealthy Friendship and Hostility between Ex- spouses', Journal of Marriage and the Family.

Masheter, C. & Harris, L.M. (1986) 'From Divorce to Friendship: A Study of Dialectic Relationship Development', Journal of Social and Personal Relationships 3: 177-89.

Metts, S., Cupach, W.R. & Bejlovec, R.A. (1989) '"I Love You Too Much To Ever Start Liking You": Redefining Romantic Relationships', Journal of Social and Personal Relationships 6: 259-74.

Nardi, P.M. & Sherrod, D. (1994) 'Friendship in the Lives of Gay Men and Lesbians', Journal of Social and Personal Relationships II: 185-99.

Riessman, C.K. (1993) Narrative Analysis. Newbury Park, CA: Sage.

Sapadin, L.A. (1988) 'Friendship and Gender: Perspectives of Professional Men and Women', Journal of Social and Personal Relationships 5: 387-403.

Schwartz, P. (1994) Peer Marriage. New York: Free Press.

Stacey, J. (1990) Brave New Families. New York: Basic Books.

Wright, P.H. (1974) 'The Delineation and Measurement of Some Variables in the Study of Friendship', Representative Research in Social Psychology 5: 93-6.

This research was partially funded by grants from Sigma Xi and the University of Utah. The author wishes to thank student volunteers, Wendy Cole, Chris Erickson, and Naomi Shumway, for transcription of the tape-recorded interviews for this research. Also, the author wishes to thank David Demo, the journal's reviewers, and colleague, Irwin Altman, for com- ments that helped strengthen this article.

Address correspondence to Carol Masheter, Family and Consumer Studies, 228 AEB,

University of Utah, Salt Lake City, UT 84112, USA. [email: masheter@alfred.fcs.utah.edu]