Psihološka obzorja /Horizons of Psychology, 19, 3, 89-102 (2010) © Društvo psihologov Slovenije 2010, ISSN 1318-187 Znanstveni empiričnoraziskovalni prispevek Shame-proneness and its correlates in couples Tomaž Erzar'*, Matej Torkar2 and Katarina Kompan Erzar1 'University of Ljubljana, Faculty of Theology, Franciscan Family Institute, Ljubljana, Slovenia 2 Franciscan Family Institute, Ljubljana, Slovenia Abstract: The study assessed the relationships between three TOSCA-3 (Test of Self-Conscious Affect; Tangney, 1990) subscales (shame-proneness, guilt-proneness, and externalization), and the following variables: stress, anxiety, depression, fear of intimacy, and attachment style in a sample of 68 heterosexual couples in committed relationships. Within-subject and within-dyad correlations were computed. Results confirmed a low to moderate connection for depression, anxiety, fear of intimacy and secure attachment. Shame-proneness, externalization, and guilt-proneness were not correlated within couples. The findings also provided further evidence for a differential understanding for some of the variables in each gender. Key words: shame, heterosexual couples, connectedness, dyadic effects Občutljivost za sram in njeni korelati pri parih Tomaž Erzar', Matej Torkar2 in Katarina Kompan Erzar' ' Univerza v Ljubljani, Teološka fakulteta, Frančiškanski družinski inštitut, Ljubljana. 2Frančiškanski družinski inštitut, Ljubljana. Povzetek: Raziskava preučuje povezanost med tremi lestvicami instrumenta TOSCA-3 (občutljivost za sram, občutljivost za krivdo ter eksternalizacija) ter stresom, depresivnostjo, strahom pred intimo in stilom navezanosti pri 68 heteroseksualnih parih v trajnih zvezah. Izračunane so bile korelacije za posameznike po spolu ter za par kot enoto. Čeprav so rezultati potrdili šibke do zmerne povezave za depresivnost, tesnobo, strah pred intimo ter varni stil navezanosti pri posameznikih, niso potrdili povezav med občutljivostjo za sram, občutljivostjo za krivdo in eksternalizacijo pri parih. Rezultati so med drugim pokazali, da moramo pri preučevanju nekaterih povezav upoštevati različno razumevanje teh vsebin pri ženskah in moških. Ključne besede: sram, heteroseksualni pari, povezanost, diadični učinki CC = 2360 * Naslov / Address: Tomaž Erzar, Frančiškanski družinski inštitut, Prešernov trg 4, 1000 Ljubljana, tel: 01 200 6760, Fax: 01 200 6766, e-mail: tomaz.erzar@guest.arnes.si In the last 20 years, research on excessive shame-proneness in adults has revealed numerous positive associations between this emotion and various forms of maladaptive behavior and mild to severe psychopathology (Tangney & Dearing, 2002). Studies have confirmed the relationship between shame and depression (Tangney, Wagner & Gramzow, 1992), shame and fear of intimacy (Lutwak, Panish, & Ferrari, 2003), shame and self-concealment (Pineles, Street, & Koenen, 2006), shame and poor conflict-resolution skills (Lopez et al., 2005), shame and loneliness, low self-image, submissiveness, co-dependency, and insecure attachment style (Wells, Glickauf-Hughes & Jones, 1999). Despite the considerable impact the proneness to shame has on intimate relationships, this role has so far been underinvestigated. One of the reasons this might be so is the elusive nature of this emotion and the fact that each gender uses different strategies to cope with and express shame. Our study explored differential relationships of shame-proneness, guilt-proneness, and externalization to stress, anxiety, depression, fear of intimacy, and styles of attachment in committed couples. Shame: the basic emotion of connectedness Research on shame in interpersonal contexts stems from the hypothesis that shame (including the entire family of emotions connected with shame, such as embarrassment, humiliation, discomfort, shyness, dishonor, mortification, and degradation) represents the core affect of human relatedness and the basic social emotion (Scheff, 1997, 2000). Avoiding shame, checking the image one fears may arise in others, and concealing one's true self behind a mask of compliance, submissiveness, and conformity - all of these forms of social behavior point to the tight relation between concerns related to connectedness and maintenance of relationships, and the deepest experience of one's own self as acceptable, lovable, and desired, or unacceptable, undesired, and unloved. Historically, attachment research has explored in detail how a child's perception of self reflected the way parents perceived and behaved towards the child (Bowlby, 1988; Main, Kaplan & Cassidy, 1985). Whereas securely attached individuals believe they are lovable and desired and have a positive view of both themselves and others, insecurely attached individuals constantly seek reassurance, fear they might be abandoned, and have a negative view of themselves and/or others. Subsequent research on adult attachment patterns established the value of early attachment bonds with parents for romantic relationships (Hazan & Shaver, 1987). Attachment patterns influence people's mental wellbeing and personal growth through stable internalized schemes or internal working models and the level of intimacy and confidence established in current adult romantic relationships (Pielage, Luteijn & Arrindell, 2005). Research tapping into intergenerational risk factors for shame has found strong links between shame-proneness in adulthood, codependent behavior in intimate relationships, and parentification in early childhood, which represents a serious form of family dysfunction (Wells, Glickauf-Hughes & Jones, 1999; Wells & Jones, 2000). Several authors vividly described how living in a shame-based family (families with alcohol dependence, families with maltreatment and abuse, families with narcissistically wounded, rejecting parents, with family secrets, etc.) affects the adult potentials of children and leads to retraumatization, longstanding loss of contact with one's own emotions, emotional emptiness and rigidity, dependent behavior, pathological perfectionism, eating disorders, personality disorders, violence and criminal activity, and difficulties with affect regulation and maintaining stable intimate relationships (Bradshaw, 1988; Earley & Cushway, 2002; Golomb, 1992; Mills, 2005; Nathanson, 1987). Overt and hidden shame in couples Social contexts and interpersonal relationships in which individuals experience rejection and unwantedness cause them to feel more shame and, consequently, more distress and insecurity (Gruenewald, Kemeny, Aziz & Fahey, 2004). Recent studies have found that the level of cortisol in blood peaks in situations in which people fear negative evaluation from others or expect to feel ashamed (Dickerson, Mycek & Zaldivar, 2008). For those that tend to make negative attributions to themselves, such situations may be exacerbated to the point where they cannot recognize whether negative feelings were evoked by some external source, or by some internal source, usually the pervasive sense of being a failure and ruminative thoughts about their inner value (Tracy & Robins, 2006). Due to constant avoidance of shame and excessive shame-proneness, these individuals experience their intimate relationships as permanent sources of danger, which in turn fuels their sense of loneliness, social isolation, non-acceptance, and rejection. In order to prevent these feelings from being too frequent and unpleasant, shame-prone individuals start ignoring and avoiding them. In her seminal book on shame, H. B. Lewis (1971) distinguished between overt and covert shame, indicating that overt shame encompasses behavior such as flushing, excessive or inappropriate laughter, touching or covering the face, gaze aversion, and rapid or barely audible talking, which all tend to hide and protect the person from the gaze and evaluation of others. In contrast, covert shame reveals itself only through subtle cues such as biting lips, sustained, tense laughter, rapid talking or stuttering, and a motionless face, which all tend to control outer signs of inner psychic states and redirect attention from these states. These differences can be traced at the verbal level as well: whereas the statement "I feel ashamed" clearly reflects overt shame, the statement "I feel uncomfortable" usually indicates covert or unrecognized shame. In addition, shame is very often covered by expressions of guilt. In couples the problem arises when unacknowledged shame in one partner through seemingly shame-free interaction induces shame in the other partner, who starts to feel shame and cannot hide it (Balcom, Lee & Tager, 1995). In this way the interpersonal bond or connection between partners becomes an arena of recurring shame cycles in which the experience of shame is strengthened and imprinted onto the couple's identity (Kaufman, 1980), regardless of whether shame is experienced by both partners or by one partner alone. Shame-based couples are characterized by intolerance toward changes and differences, lack of separation and differentiation, constant blaming, negative attributions, chronic conflicts, never-ending fights for emotional survival, use of collusive defenses, global mistrust, and manipulative rendering of past events (Rhodewalt & Eddings, 2002). Because of this mutual dynamic, researchers assume that dysfunctional couples develop some sort of common negative identity as a result of insufficient self-differentiation, individual proneness to shame, and the ongoing dynamic of blaming (Harper & Hoopes, 1990). Theoretical underpinning for this idea was found in the intergenerational theory of Murray Bowen (1978), which states that individuals find mates that are at the same level of differentiation from their families of origin. Another promising starting point for investigating couple's conflictual dynamic is the concept of shame-proneness because it taps into what makes people seek or avoid proximity. Design and hypotheses Despite the frequent use of self-report instruments such as the TOSCA (Test of Self-Conscious Affect; Tangney, 1990) for measuring self-conscious emotions, to date no study has explored shame-proneness in couples. The first aim of our study was thus to examine associations between shame-proneness and other subscales of the TOSCA instrument in couples. We expected to find in couples a similar susceptibility to shame (Hypothesis 1). In order to compare shame-proneness with other indices of interpersonal malfunctioning and personal stress, we used several additional measures: the DASS (Depression, Stress Anxiety Scales; Lovibond & Lovibond, 1995), FIS (Fear-of-Intimacy Scale; Descutner & Thelen, 1991), and RQ (Relationship Questionnaire; Bartholomew & Horowitz, 1991). We expected to find moderate to high within-subject correlations between proneness to shame and the DASS and FIS measures (Hypothesis 2), and low to moderate within-dyad correlations on all DASS measures and the FIS measure (Hypothesis 3). We further assumed that in couples the shame-proneness of the TOSCA measure would correlate with a negative view of self and others, implicated in the fearful avoidant attachment styles (Type B) of the RQ measure, and that externalization of the TOSCA measure would correlate with the dismissing avoidant attachment style (Type D), which includes a negative view of others and a positive view of self (Hypotheses 4 and 5). In order to explore cross-relationships between attachment styles, shame- proneness, and fear of intimacy in couples, we classified couples into three groups according to their self-reported attachment style - namely, secure-secure (S-S), insecure-secure (I-S), and insecure-insecure (I-I) - with an additional subgroup formed by distinguishing between secure (wife)-insecure (husband) and secure (husband)-insecure (wife) dyads (S-I). One-way ANOVA and a post-hoc test were computed for these groups with shame-proneness and fear of intimacy as the between-group factor. We predicted that there would be significant differences between groups, with both secure pairs scoring low, both insecure pairs scoring high, and insecure-secure pairs scoring in between (Hypothesis 6). Method Participants The participants in the study were 68 heterosexual couples in permanent relationships. They were recruited from several couple-support groups and via personal acquaintance, with relationships lasting from approximately one year to 40 years (M = 9,4). The age range for females was 21-60 years, mean 34,5, and for males 23-67 years, mean 37,6. Instruments The Tests of Self-Conscious Affect (TOSCA-3) consist of 16 brief scenarios (11 negative and 5 positive) depicting commonplace life situations. Respondents are asked to rate a series of associated responses on a 5-point rating scale ranging from 1 (not likely) to 5 (very likely), including descriptions of affective, cognitive, and behavioral features associated with shame and guilt. For example, "I'm inconsiderate" would be a shame-prone response to realizing you stood a friend up for an appointment, whereas "You would try to make it up to him as soon as possible" would be the associated guilt-prone response. Thus, items designed to assess the construct of shame focus on negative evaluation of the entire self, whereas items assessing guilt focus on specific behaviors. The latent factors underlying these scales are: shame, guilt, externalization, alpha pride, and beta pride. The a internal reliabilities in the current study were in the acceptable range for this instrument: .80 for shame, .73 for guilt, and .72 for externalization. Subscales with alpha coefficients under .70 were omitted from further analysis. The Depression, Anxiety, Stress Scales (DASS) is an instrument with 42 items measuring symptoms of depression, anxiety, and stress as experienced by respondents in the last week (Lovibond & Lovibond, 1995). Each scale corresponds to 14 items. Respondents answer on a 4-point scale (0 - Did not apply to me at all, 1 - Applied to me to some degree, or some of the time, 2 - Applied to me to a considerable degree, or a good part of time, 3 - Applied to me very much, or most of the time). The Slovenian version of the DASS shows reliabilities of .91 for the depression scale, .85 for the anxiety scale, and .87 for the stress scale. The intercorrelations between the scales are: depression-anxiety r = .65; anxiety-stress r = .73; and depression-stress r = .492. Principal component analysis shows that three factors explain 45.3% of variance (Erzar & Torkar, 2007). The Fear of Intimacy Scale (FIS) is a 35-item self-report inventory that measures individuals' anxiety about close, dating relationships (Descutner & Thelen, 1991; Doi & Thelen, 1993). Items are rated on a 5-point Likert-type scale, ranging from 1 (not at all characteristic of me) to 5 (extremely characteristic of me). Approximately one half of the items are reverse scored to mitigate response bias. Higher scores are indicative of greater fears. The reliability of the Slovenian version of the test is .94 (Erzar & Torkar, 2007). The Relationship Questionnaire (RQ) is a single-item measure made up of four short paragraphs, each describing a prototypical attachment pattern as it applies in close adult romantic relationships. Participants are asked to rate their degree of correspondence to each prototype on a 7-point scale (Bartholomew & Horowitz, 1991). Despite its initial prototypical approach, the RQ is based on a two-dimensional construct of adult attachment implicating a model of the self and a model of others. Thus Type A attachment style (secure) is defined as representing a positive model of self and a positive model of others, Type B (preoccupied) as representing a negative model of self, combined with a positive model of others, Type C (fearful avoidant) as representing a negative model of self and a negative model of others, and Type D (dismissing avoidant) as representing a positive model of self and a negative model of others. The RQ, a widely used measure of adult attachment, shows adequate reliability with observer-based ratings of behavioral and personality characteristics (Bartholomew & Horowitz, 1991; Griffin & Bartholomew, 1994). Results Table 1 presents the means, SD, and t-tests by gender on the three TOSCA subscales, and all of the DASS, RQ, and FIS scales. As found in other studies, females scored significantly higher on measures of shame-proneness and guilt-proneness, whereas males scored higher on dismissive attachment style. Table 2 presents the correlations by gender of the three TOSCA subscales (shame-proneness, guilt-proneness, and externalization) with the DASS, FIS, and RQ measures. The correlations appear to be moderate to low and gender-specific. Hypothesis 2 was confirmed for both genders regarding the relationship between shame-proneness and all of the DASS measures, but was disconfirmed regarding fear of intimacy. Type B attachment style (fearfulness, negative view of self and others) was found to be positively correlated with shame-proneness only in males (Hypothesis 4 partially confirmed). Type D attachment style (dismissiveness, positive view of self, and negative view of others) was positively correlated with externalization in females and negatively correlated with externalization in males (Hypothesis 5 also partially confirmed). Table 1. Descriptive statistics and t-tests for couples Males Females M SD M SD t Test of Self-Conscious Affect (TOSCA) Shame 35.59 11.29 40.55 9.94 -2.89** Guilt 59.19 11.34 64.99 6.45 -3.61*** Externalization 33.97 9.05 37.13 7.30 -2.19* Depression, Anxiety, Stress Scales (DASS) Stress 12.60 6.52 14.69 7.90 -1.59 Anxiety 6.28 4.93 7.71 6.61 -1.59 Depression 8.29 6.74 9.72 8.34 -1.20 Fear of intimacy (FIS) 74.29 23.79 71.61 21.15 0.80 Relationship Questionnaire (RQ) Secure Type A 4.09 1.96 4.49 1.91 -1.35 Fearful avoidant Type B 3.63 1.80 3.78 2.00 -0.75 Preoccupied Type C 3.14 1.58 3.64 1.90 -1.82 Dismissing avoidant Type D 3.05 1.71 2.94 1.61 2.28* * p < .05, ** p < .01, *** p < .001 Table 3 presents within-dyad correlations of all the subscales used previously. Contrary to our Hypothesis 1, shame-proneness was not positively correlated within pairs. On the other hand, the results confirmed low but significant correlations on measures of anxiety, depression, fear of intimacy, and secure attachment style, as predicted in Hypothesis 3. Figure 1 presents the scores of the TOSCA shame-proneness and fear of intimacy subscales for the four groups of pairs combined according to their attachment style. As hypothesized, both secure pairs (S-S) had the lowest scores on both measures. However, both insecure pairs (I-I) scored the highest only on fear of intimacy. Surprisingly, on the measure of shame-proneness insecure (male)-secure (female) (I-S) couples outscored both other groups (Hypothesis 6 only partially confirmed). 13 CD 'fe =1 CJ O o