Abstract
Facework and resilience frameworks were employed to examine threats to, and the protection of, marital partners’ identity during divorce and as they relate to outcomes associated with divorce. Divorced participants (N = 103) reported on the communicative face threats and support during divorce, reporting greater positive face threat and negative facework during divorce. The non-initiator experienced negative outcomes, including negative face threat and negative emotion, but reported no differences in stress when compared to the initiator or mutual decision makers. In addition, positive face threats and facework during divorce predicted post-divorce relationships. Specifically, low positive face threat and high positive face support were related to divorcees’ inclination to engage in positive interpersonal relationships after the divorce.
Spouses are the primary relationship for most adults (Kiecolt-Glaser & Newton, 2001). In fact, marital partners are considered both important and irreplaceable by their spouses (Cutrona & Suhr, 1994; McKenry & Price, 1991). Despite this categorization as a primary, important, and irreplaceable relationship, even marriage is not always a permanent union. Approximately one-third of first marriages end in divorce within 10 years, and 50% of all first marriages will end in divorce at some point in the lifespan (Braver, Shapiro, & Goodman, 2006; U.S. Department of Health and Human Services, 2002).
Divorce has both immediate and long-term consequences for the divorcees, extended family, and children involved. It is the most traumatic situation, having the furthest reaching implications of all life events (Amato, 2000). More recent reviews of the divorce literature continue to reveal negative outcomes for divorcees (Amato, 2010). Divorcees often experience deteriorated physical and psychological conditions for years following marital termination including guilt, depression, distress, and intimacy issues (Afifi & Hamrick, 2006; Waite, Luo, & Lewin, 2009). Leary, Koch, and Hechenbleikner (2001) found that those who experienced the loss of an intimate relationship, such as a marital relationship, could escalate into more serious issues, including substance abuse, suicidal thoughts, or suicide attempts. The well-documented negative impact of divorce speaks to the importance of understanding ways to alleviate the detrimental psychological, emotional, and physiological consequences that divorcees experience. While these effects seem grave, some individuals who go through a divorce experience positive outcomes. Individuals who experience these positive outcomes appear to be more resilient to this adverse event. There are several potential explanations for the positive outcomes.
Firstly, positivity following divorce may be a result of effective conflict management and feelings of relief. Amato (2000) argued that the results of divorce can be positive when problems between the divorcing partners are resolved effectively. Much research has focused on conflict management as one way to enhance perceptions of successful dissolution, and productive outcomes post-divorce. Spouses may experience loneliness and depression while married (Doohan, Carrere, & Riggs, 2010) and may view divorce as an exit from an already negative situation. For example, some divorcees even experience feelings of happiness, freedom, and demonstrate progress by reinventing themselves following the decision to divorce (Tashiro & Frazier, 2003).
Secondly, some individuals may be more resilient. Bonanno (2004) argued that resilience may occur as a result of other factors, including relationship quality, support, one’s worldview, and contextual factors in which the adverse event occurred (Bonanno, Wortman, & Nesse, 2004). The extant research on resilience after divorce has almost exclusively focused on the resilience of children from divorced families (e.g., Chen & George, 2005; Mulholland, Watt, Philpott, & Sarlin, 1991) or resilience of the parent–child relationship following divorce (e.g., Golby & Bretherton, 1999). Instead, this study examined face supportive communication early in the divorce process as a potential contextual and protective factor to enhance resilient responses in divorcees during and after the divorce process.
Thirdly, strategic facework, or the protection of the former spouses’ identity and feelings when requesting a divorce, may be related to perceptions of an effective divorce and resilience. Practicing facework very early in the divorce process may be associated with lower face threat, and reduced subsequent conflict between the individuals. Face sensitive divorce talks may also maintain the face of each divorcee to his/her social networks and families, setting the stage for a more amicable termination. Thus, the current study had three purposes: (1) to examine face threat and facework in divorce; (2) to differentiate between initiators, non-initiators, and mutual divorce decision makers; and (3) to examine the inter-relationship of face threats and facework and emotions, stress, and the post-divorce relationship.
Face and facework in divorce
Goffman (1967) defined face as the “positive social value a person effectively claims for himself” (p. 5). Individuals portray this constructed self to others and seek support for this identity. Individuals believe that their identity, or face, is important to them and will work to have that identity accepted (Cupach & Metts, 1994). Within every interaction, two faces are presented (i.e., positive and negative). Positive face is the desire for approval, closeness, solidarity, and liking, while negative face is a need for autonomy and freedom from imposition (Holtgraves, 2001). When identity is not supported, people experience face threats. Senders who realize that they are threatening another’s face may attempt to temper the perceived face threat through facework (Goffman, 1967). Ideally, interactants will communicate to maintain both faces of all parties involved (Goffman, 1967).
Divorce is one context where the identity of communication partners may be threatened. Kunkel, Wilson, Olufowote, and Robson (2003) argued that relational dissolution episodes are inherently face threatening, and Cupach and Metts (1994) speculated that relational dissolution was the most face-threatening interpersonal event. Divorcing couples may communicate in ways that protect their own, their partners’, or both individuals’ faces. In the event that dissatisfied spouses do not communicate to protect their partner, face threats occur. Spouses may try to temper face threats by focusing on the partner’s positive qualities, or including them in the divorce decision (i.e., practicing facework), while simultaneously dissolving the relationship. The communication of face threats or face support can occur early in the divorce process, and may set the stage for the continuing interactions between the former spouses. Only Miller (2009a) has directly examined face threats and facework in divorcing couples, finding that the face concerns reported involved parental identity. Gregson and Ceynar (2009) found that women had to separate themselves from the marriage and reconstruct a new post-divorce identity. Similarly, our study focused on the marital identity of the former spouses. During the divorce, individuals may feel that they were incompetent in the relationship, no longer liked, or that they do not have a choice, all of which may increase face threat and negative consequences. Consequently, facework strategies communicated during divorce may be related to a variety of post-divorce outcomes.
Resiliency in divorce
The theoretical framework employed here involves risk and resilience models that conceptualize complex factors contributing to positive outcomes. Appropriate communication at the time of a negative event has the capacity to improve subsequent functions. Some individuals exit divorce with relatively few negative outcomes, demonstrating resilience, defined as the ability to “maintain relatively stable, healthy levels of psychological and physiological functioning” (Bonanno, 2004, p. 20). In fact, Bonanno et al. (2002) found that resilience was more common than grief. People who are resilient exhibit fewer depression and grief symptoms (Bonanno et al., 2002). Most resilience research investigating loss has focused on death (e.g., Bonanno et al., 2002, 2004). However, resiliency should also be considered in divorce, as individuals who have experienced separation and rejection from a marital partner physiologically and psychologically may experience grief similar to those who have lost a loved one through death (Archer & Fisher, 2008; Seery, Holman, Alison, & Silver, 2010).
Risk and resilience models focus on the risk factors that affect the outcomes of individuals when faced with life stressors such as divorce (Deater-Deckard & Dunn, 1999; Seery, et al., 2010). In line with Walsh (2002), who identified communication as one of three primary factors affecting resilience, Afifi and colleagues contend that communication is one specific risk factor that can lead to either positive or negative outcomes (Afifi and Hamrick, 2006; Afifi & Keith, 2004) For example, Afifi and Keith (2004) found that for children of divorce the quality of contact, self-disclosure, and parental conflict were the communicative factors that affected their resilience to the divorce. However, Afifi and Nussbaum (2006) note that many of the studies using risk and resiliency models focused on children of divorce. Amato (2010) recognized that a variety of factors can influence the adjustment, whether positive or negative, of former spouses as well. Following Afifi and Hamrick’s (2006) and Afifi and Keith’s (2004) approach to focusing on communication, this study examined the relationship between communication processes that occur during the divorce adjustment and resilience. Instead, this study focused on resilience in former spouses.
Identifying factors that enhance risk or resilience following divorce could allow researchers and practitioners to target specific people who are in need of a post-divorce intervention or counseling. An understanding of the different responses to divorce, and who will respond in particular ways, can also inform social networks about providing comfort, since social networks are often a primary source of support following dissolution (Kincaid & Caldwell, 1991). One step toward identifying those who are at more risk begins with understanding contextual factors, as suggested by Bonanno (2004). Specifically, understanding the context in which the divorce is set into motion, beginning with the decision to divorce, and the roles individuals play in the divorce, will begin to illuminate risks that each partner may face.
Roles in divorce and resiliency
Numerous studies have revealed significant differences based on the role played in the dissolution decision (e.g., Boelen & Reijntjes, 2009; Doering, 2010; Perilloux & Buss, 2008; Thuen & Eikeland, 1998). When partners decide that they are no longer committed to continuing a relationship, they begin the process of disengaging. Despite the initiator’s decision to dissolve the relationship, Kunkel et al. (2003) argued that the person who initiates the end of the relationship is often concerned with threats to positive face, or appearing insensitive or uncaring to the partner. Thus, even for the initiator, it is often difficult to disengage. Still, outcomes may differ for initiators because they are aware that the dissolution is approaching. Bevvino and Sharkin (2003) suggest that initiators may contemplate the dissolution for approximately five years, prepare emotionally and mentally for the divorce, and find positive meaning in terminating the relationship. Consistent with that argument, additional research revealed that dissolution initiators report experiencing lower disturbance scores on grief (Boelen & Reijntjes, 2009), more happiness (Perriloux & Buss, 2008), and improved adaptation to divorce (Thuen & Eikeland, 1998).
Non-initiators, or the person who does not desire or request a divorce, but rather receives the decision from the partner, are often viewed as having the more difficult role to play in the interaction. Non-initiators suffer more emotionally, reporting greater grief, depression, and anxiety (Boelen & Reijntjes, 2009; Perilloux & Buss, 2008). Kunkel et al. (2003) reasoned that the non-initiator felt that they had not earned approval or liking from their partner. Thus, partners being left may question their individual worth and relational competence. Non-initiators evidenced lower resilience in that they experienced more negative outcomes, such as greater depression, rumination, crying, pleading (Perilloux & Buss, 2008), sadness, anger, confusion, shock, and jealousy (Leary, Springer, Negel, Ansell, & Evans, 1998).
Although research exists that delineates differences in perceptions and outcomes based on the role in the divorce decision, initiators and non-initiators exhibited some similar outcomes, including vengefulness, indifference, fear, remorse, and regret (Perilloux & Buss, 2008). Other research has failed to find significant differences in affect, adjustment, or well-being (Kincaid & Caldwell, 1991; Tashiro & Frazier, 2003). The lack of apparent differences may be attributed to the categorization of divorcees into polar opposite groups (i.e., initiator or non-initiator). To further complicate the matter, previous research has placed those who labeled the decision as mutual into the initiator category (e.g., Kiecolt-Glaser et al., 1988; Kincaid & Caldwell, 1991). More recent research revealed that 13–19% of their samples reported that the divorce was a mutual decision (Bevvino & Sharkin, 2003; Thuen & Eikeland, 1998). This proportion of mutually deciding divorcees should not be ignored or treated as unilateral initiators.
It is expected that the degree of responsibility a partner has in the dissolution decision process will guide perceptions formed about the divorce. Gray and Silver (1990) argued that the level of control in the decision was positively related to post-divorce adjustment, and negatively related to distress, preoccupation, and regret. Following this logic, it is expected that mutual decision makers will each take responsibility for the demise of the marriage, that each partner will maintain control in the divorce decision, and avoid threats to negative face. Subsequently, mutual decision makers should perceive the divorce differently than those who lose control in the termination of the marriage.
Facework and resiliency in divorce outcomes
Following dissolution, individuals often experience negative changes in affect, well-being, and overall functioning (Frazier & Cook, 1993). Divorcees may suffer from a loss of trust, lowered self-esteem, anxiety, increased worry about being hurt in future relationships, lowered confidence, preoccupation with what others think (Leary et al., 1998, 2001), negative emotions (Kam & Bond, 2008), and anger and depression (Rohde-Brown & Rudestam, 2011). However, facework has the potential to soften the blow of relational loss, alleviate negative reactions to dissolution, or deter former spouses from progressing to the more severe outcomes associated with divorce when used by partners during dissolution. Indeed, Doering (2010) found that individuals often used facework in their narratives after relationship dissolution.
Consistent with risk and resilience models, behaviors that happen early in the divorce, or even early in the marriage, can impact later outcomes between the partners. Hackney and Ribordy (1980) found that negative emotions rose sharply for individuals during the distress phase, and longitudinally continued throughout the divorce process. That is, the negative emotions that emerged before the couple completed the divorce were significant factors in the rest of the divorce process. More recently, Masheter (1997) discussed the differences between healthy and unhealthy post-divorce relationships, with preoccupation and hostility being primary qualities that differentiated those relationships. Divorcees who were highly preoccupied with, and hostile toward, their former spouse were poorly adjusted to the divorce and experienced detrimental effects on their well-being (see also Jaffe, 2011).
Divorcing couples report less constructive communication, more conflict, and more interaction avoidance than either non-distressed or distressed couples (Christensen & Shenk, 1991). Scholars have noted that in conflict, face concerns contribute to the way conflict is managed, and that perceived face threats can escalate a conversation into a conflict episode (Brew & Cairns, 2004; Cupach & Metts, 1994; Oetzel, Ting-Toomey, Yokochi, Masumoto, & Takai, 2000). Thus, the face threats that occur during the divorce decision challenge the partners’ autonomy, approval, and closeness with the former spouse, eliciting unproductive conflict behaviors throughout the entire divorce process. Taken together, face threats can generally be associated with negative outcomes, while facework is associated with positive outcomes. These findings suggest that the loss of a close, personal relationship triggers strong emotions and conflict. The emotions appear to persist throughout the process, and may continue after the divorce.
Similarly, communication between former spouses may continue after divorce. Graham and Edwards (2008) argued that divorce is not an end state to a relationship. Instead, the former spouses enter into a dynamic and changing post-marital relationship and state of being (Graham, 2003; Graham & Edwards, 2008). In this evolving state, former spouses developed new disclosure and privacy rules to maintain a harmonious relationship (Miller, 2009b) and negotiated co-parenting responsibilities (Schrodt, Baxter, McBride, Braithwaite, & Fine, 2006). Positive post-divorce relationships potentially demonstrate greater resiliency to divorce.
Resiliency in the post-divorce relationship
Post-divorce relationships may be dynamic as former spouses develop new types of relationships and sometimes enter into friendships. Although divorcees recognized that they experienced stress and conflict as a result of friendship with a former spouse, they also found the relationship supportive and beneficial (Goldsmith, 1980). Ahrons and Wallisch (1987) classified divorced couples into five categories, ranging from friendship to the elimination of all contact between the spouses. These categories seem to have an emotional foundation. Perfect pals are described as friends who are amicable in all aspects of their lives. Cooperative colleagues are successful as co-parents, but do not deal with one another in other interpersonal aspects of their lives. Angry associates are characterized as hostile, and limit their interactions with one another to avoid conflict. Fiery foes are extremely hostile and experience great conflict in both the parental and personal realm. Dissolved duos have ceased all contact with one another.
Because resiliency involves the maintenance of stable, healthy levels of functioning (Bonanno, 2004), the ability to become perfect pals or cooperative colleagues likely demonstrates a unique, and more perhaps more resilient, reaction to the divorce process. In other words, such individuals are maintaining stability and healthy functioning in post-divorce relationships. To summarize, divorce is an interpersonal event that can be compared to other interpersonal termination events where face threat is prevalent (Kunkel et al., 2003). Face threats in divorce have the potential to impact both parties involved in the dissolution of the marriage, playing a role in the emotions, stress, and type of post-divorce relationship negotiated. The emotion, stress levels, and post-divorce relationship can lend insight into those who are more resilient following marital termination. Thus, this study examined the following hypotheses and research questions: H1: Divorcees will perceive greater positive face threat than negative face threat from their former spouses in divorce. RQ1: Will divorcees perceive greater positive or negative face support from their former spouses in divorce? RQ2: How will divorce initiators, non-initiators, and mutual divorce decision makers differ in perceived face threat, face support, positive emotions, negative emotions, and stress? H2: Face threat and facework during divorce will predict emotions and stress. H3:Perceived face threat, facework, and emotion will predict the current relationship status with their former spouse.
Method
Procedures
Participants were recruited using snowball sampling techniques through large lecture classrooms, divorce support groups, Facebook divorce groups, Facebook status messages, and mass emails. Participants were instructed to contact the researcher to indicate interest and were then sent an electronic survey. 1 To assist in recall, only participants who were divorced within two years were retained for the study. Firstly, participants completed an electronic consent page. Secondly, participants responded to an open-ended prompt about the divorce (i.e., “Describe the conversation where you perceived that the decision to divorce was final”). Although there is the potential that this recollection may be distorted, Harvey and Fine (2006) argued that the way in which divorcees recall, and then describe, their accounts of the divorce can both alter and shape their present and future experiences related to the divorce. In other longitudinal studies (e.g., Hackney & Ribordy, 1980), the communication between couples early in the divorce process still influenced later interactions between the partners. The survey packet also included quantitative measures about communication behaviors that occurred during the recalled divorce conversation (i.e., face threat, facework) and post-divorce emotion and levels of stress.
Participants
Participants reported on their perceptions of the divorce (N = 103, 34 males, 66 females, and 3 who did not report their biological sex) ranging in age from 20 to 58 (M = 36.60, SD = 10.66). Participants had been married an average of 9.91 years (range = 1–29 years, SD = 8.16 years), and divorced an average of 1.05 years prior to taking this survey (range = 1 month to 2 years, SD = .66 months). Of the participants, 44.7% of the divorcees had no children, followed by 23.3% who had two children, 18.4% who had one child, 9.7% who had three children, 1.9% who had four children, and 1.9% who had six children.
Instrumentation
Role in the Divorce
Participants were asked to categorize themselves as initiators, non-initiators, or as mutual decision makers. Participants reported that they were the initiator of the divorce (n = 51), non-initiators (n = 15), or they mutually decided to divorce (n = 28), and nine participants did not report who initiated the divorce.
Face threats
Perceived face threat was measured using a scale developed by Cupach and Carson (2002). It is a 14-item scale assessing both positive (e.g., my partner’s actions strengthened the relationship between us) and negative face threats (e.g., my partner’s actions took away some of my independence) on a seven-point Likert-type scale ranging from strongly disagree (1) to strongly agree (7). The negative face threat scale has been marginally reliable in previous research (Cupach & Carson, 2002). For the purposes of this study, the scale was modified through the addition of two negative face threat items in an attempt to produce a higher reliability (i.e., my partner’s actions made me feel obligated to comply, my partner’s actions made me feel like I had no control). With the addition of these two items, the positive and negative face scales contained 10 and 6 items, respectively. See Table 1 for scale reliability, means, and standard deviations.
Relationships between demographics and divorce outcomes
Note: *p < .05, **p <. 01
Facework
Facework was assessed using a modified version of the Revised Instructional Face Support Scale (Kerssen-Griep, Trees, & Hess, 2008). The items were measured on an eight-item, seven-point Likert-type scale ranging from not at all (1) to very much (7) to indicate the degree to which both positive facework (e.g., made sure that he/she doesn’t cast you in a bad light) and negative facework (e.g., left me free to choose how to respond) was enacted. Each subscale was comprised of four items. Negative face support yielded a low reliability coefficient and could not be improved through item deletion. Negative face support was used in further analyses but the results should be interpreted with caution.
Emotion
Emotion was assessed using the Positive and Negative Affect Schedule (PANAS; Watson, Clark, & Tellegren, 1988). The scale includes 10 negative affect descriptors (e.g., guilty, nervous, upset) and 10 positive affect descriptors (e.g., inspired, enthusiastic, attentive). Following McCullough, Bellah, Kilpatrick, and Johnson (2001), participants indicated the degree to which they generally experienced this emotion on a five-point Likert-type scale ranging from very slightly (1) to extremely (5). Three additional items (i.e., love, anger, sadness) were added following a study by Sbarra and Ferrer (2006).
Given the use of an established scale and the theoretical expectations for the items, confirmatory factor analysis (CFA) was used. We adopted criteria outlined by Hu and Bentler (1999) and Browne and Cudeck (1993), who suggest that the non-normed fit index (NFI) and comparative fit index (CFI) should be above .90 to indicate a reasonably good fit, and the root mean square error of approximation (RMSEA) should be below .08 to indicate an acceptable fit. Finally, the chi-square to degrees of freedom ratio should be at 2:1 or less. However, NFI, CFI, and RMSEA are considered more appropriate primary fit indices to be used when assessing model fit (Byrne, 2001). The CFA for positive emotions revealed a poor fit to the data with approximately a 3:1 ratio, χ2 = 151.35, p < .001, NFI = .61, CFI = .67, and RMSEA = .15. The items were examined and six positive emotion items were removed from the model. The CFA of the second positive emotion model revealed an improved, and acceptable, fit to the data with approximately a 3:1 ratio, χ2 = 13.495, p = .01, NFI = .93, CFI = .95, and RMSEA = .12. Thus, the final positive emotion scale included five items (i.e., strong, inspired, active, proud, and enthusiastic).
The CFA for negative emotions revealed a poor fit to the data with approximately a 5:1 ratio, χ2 = 230.50, p < .001, NFI = .65, CFI = .70, and RMSEA = .18. The items were examined and five items were removed from the model. The CFA of the second negative emotion model revealed an improved, and acceptable, fit to the data with approximately a 3:1 ratio, χ2 = 44.05, p < .001, NFI = .88, CFI = .91, and RMSEA = .14. Thus, the final negative emotion scale included seven items (i.e., nervous, jittery, upset, ashamed, scared, afraid, sadness).
Stress
Stress was assessed using Cohen, Kamarck, and Mermelstein’s (1983) Perceived Stress Scale. The stress scale assessed levels of stress within the past month (e.g., “In the last month, how often have you felt nervous and stressed”). It is a 14-item, five-point Likert-type scale ranging from never (0) to very often (4).
Current relationship
Ahrons and Wallisch’s (1987) qualitative descriptive typology (i.e., perfect pals, cooperative colleagues, angry associates, fiery foes, dissolved duos) was provided to the participants. They indicated which description best illustrated their current relationship with their former spouse. The participants classified themselves as perfect pals (n = 13, 12.6%), cooperative colleagues (n = 41, 39.8%), angry associates (n = 27, 26.2%), fiery foes (n = 4, 3.9%), and dissolved duos (n = 17, 16.5%), and one who did not report the type of relationship they currently maintained with their former spouse.
Results
Preliminary analyses
Previous research has uncovered differences in the divorce experience based on sex (e.g., Perilloux & Buss, 2008), relationship length (e.g., Shortt, Capaldi, Kim, & Owen, 2006), time since the divorce (e.g., Kitson, 1982), and the number of children (e.g., Berman, 1985). Because of the associations of these demographics to dependent variables in previous research, preliminary analyses were used determine whether these variables should be entered as control variables. To explore sex differences, a multivariate analysis of variance (MANOVA) was conducted with participant sex entered as the fixed factor and face threats, facework, emotion, and stress entered as the dependent variables. Results indicate that there were no significant differences between the sexes, Wilk’s Λ = .96, F (1, 88) = .51, p = .83, η2 = .04.
Correlations between relationship length, time since the divorce, and number of children with face threats, facework, emotion, and stress were examined. See Table 1 for correlations among all variables. The time passed since the divorce, the number of children between the divorcees, and the length of marriage were not related to any of the variables of interest. Given the lack of significant relationships among demographics and the outcome variables, the demographics were not entered as control variables in subsequent analyses.
Hypotheses and research questions
Hypothesis one predicted that divorcees would experience greater positive face threat than negative face threat during divorce. A paired samples t-test revealed that there was a significant difference in the positive and negative face threats (r = .59, p < .001) experienced by divorcees, t (101) = 7.53, p < .001, with divorcees reporting greater positive face threat (M = 4.19, SD = 1.49) than negative face threat (M = 3.24, SD = 1.29).
Research question one inquired about the face support that participants perceived their former spouses communicating during divorce. A paired samples t-test revealed that divorcees perceived a significant difference in the positive and negative face support (r = .53, p < .001) being performed by their former spouses during the divorce, t (99) = 6.62, p < .001, with the perception that former spouses enacted greater negative face support (M = 4.28, SD = 1.39) than positive face support (M = 3.38, SD = 1.42).
Research question two inquired about the differences in perceptions of divorce based on the role of the divorcee in the process (i.e., initiator, non-initiator, and mutual decision maker). A MANOVA with the participants’ role entered as the fixed factor and positive face threat, negative face threat, positive face support, negative face support, positive emotions, negative emotions, and stress entered as outcome variables was used. Results revealed a significant model, Wilk’s Λ = .51, F (2, 82) = 4.26, p = .001, ηp 2 = .29. There were no significant differences in positive face threat, F (2, 82) = 1.56, p = .22, η p 2 = .04, negative face support F (2, 82) = 1.66, p = .20, η p 2 = .04, or in stress, F (2, 82) = .67, p = .51, η p 2 = .02.
Conversely, the role a divorcee played in the divorce process contributed to significantly different perceptions of negative face threat, F (2, 82) = 3.51, p = .03, η p 2 = .08, positive face support, F (2, 82) = 6.62, p < .01, η p 2 = .14, positive emotions, F (2, 82) = 9.14, p < .01, η p 2 = .19, and negative emotions, F (2, 82) = 4.74, p = .01, η p 2 = .11. Post hoc analyses revealed that non-initiators experienced significantly greater negative face threat than those who were initiators, but those who mutually decided to divorce did not significantly differ from either initiators or non-initiators. Mutual decision makers experienced significantly greater positive face support than those who were initiators, but non-initiators did not significantly differ from initiators or mutual decision makers. Initiators and mutual decision makers experienced significantly greater positive emotion than non-initiators. Lastly, non-initiators experienced significantly greater negative emotion than initiators or mutual decision makers. The MANOVA results for research question two are shown in Table 2.
Multivariate analysis of variance assessing differences in the divorcee role
Hypothesis two examined the ability of face threats and face support to predict emotions and stress. Three multiple regressions were used with positive face threat, negative face threat, positive face support, and negative face support entered as predictors of positive emotions, stress, and negative emotions. The regression model predicting positive emotions was not significant, F (94) = 1.62, p > .01. The regression model predicting stress was not significant, F (94) = 1.36, p > .01. The regression model predicting negative emotions was significant, F (97) = 4.75, p < .01, and accounted for 17% of the variance. The only significant predictor of negative emotions was negative face threats (β = .33, p < .01).
Hypothesis three predicted that face threat, face support, and emotion would predict the current relationship status with the former spouse. A discriminant analysis with the relationship type as the dependent variable, and face threats, facework, and emotion as independent variables, indicated that this combination of variables could accurately predict type of relationship (i.e., perfect pals, cooperative colleagues, angry associates, fiery foes, and dissolved duos) among former spouses 52.1% of the time [Λ = .51, χ2 (24, N = 103) = 58.43, p < .001]. There was one significant discriminant function (eigenvalue = .68), that accounted for 81.6% of the variance, and included positive face threat (.85, p < .01) and positive face support (–.85, p < .01). Specifically, the discriminant function was most accurate in classifying cooperative colleagues (89.5% accuracy), angry associates (54.2% accuracy), and dissolved duos (12.5% accuracy). That is, positive face threat and positive face support assist in correctly classifying participants into the types of post-divorce relationships 52.1% of the time.
Discussion
This study revealed a pattern in which facework and face threat communication during marital-ending conversations were associated with outcomes following divorce. It also identified divorcees on a recognizable continuum ranging from more positive and resilient outcomes (i.e., mutual decision makers) to more negative and vulnerable outcomes (i.e., non-initiator), with initiators situated between the two. Contrary to expectations, the communicative and emotion variables were not related to participants’ stress levels, suggesting that face threats, face support, and emotions are not a source of stress following divorce, or that other factors may be more fundamentally integrated and in need of exploration. Further, the combination of face threats, facework, and emotions correctly classified partners over 50% of the time into one of five post-divorce couple types, ranging from perfect pals to dissolved duos. These results can be explained in several ways, suggest numerous avenues for additional research on post-divorce relationships, and highlight implications for divorcees.
Both parties in dissolution are prone to experience face threat. Kunkel et al. (2003) suggested that those who initiate the decision are concerned about being perceived as insensitive or uncaring, while receivers of the dissolution message are concerned about gaining approval, liking, closeness, and being perceived as relationally competent. It was expected that the decision to divorce may be even more face threatening because of the marital connection that differentiates spouses from dating romantic partners (Cupach & Metts, 1986). Although both positive and negative face threats occurred, positive face threats were most salient. Leary et al. (1998) explained that relational devaluation is a significant cause of hurt in interpersonal rejection. Devaluing the relationship suggests lack of approval for the partner, which is threatening to positive face. Young, Paxman, Koehring, and Anderson (2008) reported that rejection messages in unrequited love were also perceived as face threatening. Taken together, feelings of rejection and unrequited love may threaten a partner’s perceived likeability, relational competence, and closeness needs, all of which comprise the positive face of a marital partner.
Although divorce is a highly face-threatening event, following the conceptual models of risk and resilience, we must recognize that all divorces are not the same. For example, Erbert and Floyd’s (2004) research on trait-like face needs may help to explain this. If, indeed, face needs are a trait-like construct linked to perceptions of facework, then this trait may affect individual’s interpretations of face-threatening events such as divorce. Secondly, the variance in face threat may also be explained by the motivation for the divorce. Relationships dissolve for a variety of reasons (e.g., infidelity, ineffective interpersonal communication) and previous research has found that perceived face threat differs by reason (Frisby, Booth-Butterfield, & Malachowski, 2010). Thus, the reason for divorcing may differentially threaten positive and negative face, and should be examined in future research.
Non-initiators of the divorce experienced greater negative face threat and greater negative emotion. Conversely, mutual decision makers and initiators of divorce emerged as most likely to be resilient, experiencing positive outcomes including greater positive emotions and less negative emotions. Sprecher and Felmlee (1997) reported that an imbalance of power is evident in decision making. That is, non-initiators in the divorce may perceive that they are powerless in the decision process, and may become painfully aware that their partner is less interested in that relationship. Gray and Silver (1990) noted that responses to termination depend on the partner’s level of control in the break up. Non-initiators lack this control, and likely experience threats to negative face.
Despite suggestions to include emotion in face threat and facework research (Craig, Tracy, & Spisak, 1986; Locher & Watts, 2005; Metts, 1992), few studies have linked face threat to emotion. Recently, Kennedy-Lightsey (2010) found that positive face threat elicited more negative emotional responses, including anger and frustration. In the current study, and consistent with predictions, face threats were also related to negative emotions, suggesting that the more face threatening the divorce, the more negative the emotional reactions. Further, the non-initiator of the divorce experienced even greater negative face threat and greater negative emotion.
Like Kennedy-Lightsey (2010), most research has focused on negative emotional responses to face threat (e.g., Cupach & Carson, 2002; Zhang & Stafford, 2008). However, Tashiro and Frazier (2003) found that some individuals experience positive emotions in response to divorce. In our study negative face threats were significantly and negatively related to positive emotions. That is, when negative face threats are low more positive emotions occur. One of the positive outcomes of divorce suggested by Tashiro and Frazier was personal growth and freedom. This finding may be indicative of low negative face threat, or gaining autonomy and independence from the relationship, as well as experiencing a positive emotional reaction to the divorce. These findings support the need to further incorporate emotion into face threat and facework research.
Our study did not find face threats, facework, or emotion to be associated with divorcees’ stress levels. Plummer and Koch-Hattem (1986) found that other factors, such as changes in the social network, income, conflict, and external family stressors, such as a child leaving home, illness, or relocation, besides the loss of the relationship through divorce, were associated with heightened stress levels. These life changes are significant, and can occur regardless of the role a former spouse played in the decision to divorce. Similarly, partners may experience turbulent emotions as they experience feelings of freedom then guilt, happiness then regret, missing the partner then elation in a new relationship, and other contradictory experiences and dialectical tensions due to divorce. Thus, despite the appearance that a mutual decision may be most beneficial, partners who engage in bilateral decisions are still as likely to experience some negative outcomes, including stress.
Importantly, this study extends research on face threats and facework by integrating those facets within a risk/resiliency model. It identifies the explanatory potential for these communicative practices as contextual indicators of resiliency and post-divorce relational negotiation. Positive face threats and positive face support were significant in classifying participants into the five post-divorce couple types approximately 52% of the time. At first glance this percentage may seem low; however, with five possible classifications the individuals could be correctly classified 20% of the time by chance. Thus, positive face threat, positive facework, and emotion increased the predictive abilities by just over 30%. Importantly, it appears that the communication which occurs during the conversation where partners decide to divorce is informative for understanding, and making predictions about, the post-divorce relationship. When former spouses pay attention to positive face by reducing positive threats and engaging in positive face support, the post-divorce outcomes are more positive.
Ours is not the only study to find such impact. Huston (2009) reviewed and summarized a 13-year longitudinal study and found that communication between two individuals as early as courtship could predict later communication and relationship success. This finding fits with the enduring dynamics model (Caughlin & Huston, 2006), which posits that the dynamics present early in marriage persist throughout the relationship lifespan. Similarly, Isaacs and Leon (1988) found that post-termination communication patterns were often continuations of the communicative patterns that happen prior to termination, especially for conflict patterns. In our study, communication patterns that developed during courtship and marriage may have continued, explaining the face-threatening and face-supporting communication between former spouses during the divorce. The communication that occurred during the marriage may also explain the stress levels and post-divorce relationship. Future research should examine communication patterns that happened during marriage as an alternative explanation for post-divorce outcomes.
Implications
When couples are in distress, it is not uncommon for them to seek help. In fact, Renick, Blumberg, and Markman (1992) argued that most couples do not seek treatment until significant relational problems are already occurring. That is, these couples are already at a significant risk for divorce. Prior clinical efforts have focused on a wide variety of communication skills (Hawkins, Blanchard, Baldwin, & Fawcett, 2008), including expressing feelings, active listening, constructive/destructive communication, problem solving, team-building, re-evaluating beliefs, and conflict management (Markman, Floyd, Stanley, & Storaasli, 1988; Renick et al., 1992). However, facework has been largely ignored in relational education courses. It is possible that facework operates at a different level than each of the communication skills included in educational programs. Specifically, facework may be a concept that subsumes individual skills, such that effective listening, constructive communication, problem solving, and conflict management are beneficial together because they communicate face support to the spouse. Facework skills may serve a dual function, improving communication to prevent divorce or communication that will make the divorce decision and process go more smoothly. Promoting face-supporting communication can help to alleviate the negative outcomes associated with divorce and, in fact, enhance positive outcomes, including the development of positive post-divorce relationships.
Limitations
The findings of this study are framed by inherent limitations. Firstly, a continuing problem in face threat and facework research is the low reliability of negative face threat (e.g., Cupach & Carson, 2002; Zhang & Stafford, 2008). Negative face threat may be more difficult to measure due to varying perceptions of autonomy, freedom, and imposition. Adding additional items in the current studies did slightly improve the reliability; however, the measure should continue to be refined. Secondly, the data are retrospective, cross-sectional, and self-report, so the results do not support causal relationships. Although Harvey and Fine (2006) argue for the value of divorcees’ perceptions as important research tools, it is important to reduce recall error in divorce research without using intrusive methods during the traumatic event. Even if participants’ ability to recall the event is accurate, they may strategically choose to report on the event in a way that would protect their own face. For example, they may claim that the divorce was a mutual decision when in fact they were compliant, coerced, or gave up on the marriage. The sample size in this study was small (N = 103) due to difficulty in recruiting divorced individuals to recall a sensitive topic. However, a strength of this study is the examination of adults who have been married and divorced, rather than dating, and college couples. Finally, a question about previous relationship education or counseling could have been added to the survey. Thus, the results should be interpreted with caution.
Conclusion
Face threats and facework were indeed contributing factors in individual well-being and relational (re)definition. The complex divorce process is fraught with opportunities to threaten, or support, partners’ face needs, which in turn, create significant turning points that allow divorcees to evolve as post-marital relational partners. Insight into this ongoing, dynamic process provides a deeper understanding of those who are evolving in a more resilient way. Enhancing resilient responses to a traumatic and face-threatening life event is essential to further development of clinical, psychological, and mandated counseling and mediation processes.
Footnotes
Acknowledgement
A version of this manuscript was presented at the 2011 annual meeting of the International Communication Association.
Funding
This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.