When Dissociation Violates the Self, Shame Follows — Even Alone

Dissociation is one of psychology’s most misunderstood phenomena. It spans experiences of detachment, such as depersonalization, derealization and distortions of time, and compartmentalization, such as amnesia, flashbacks and dissociative identities. These disruptions to memory, identity, attention and awareness are far from rare, and decades of clinical observation have tied them to intense negative emotions, above all shame. Shame is an exceedingly aversive affect associated with cognitions of being flawed, weak and inadequate, with a sense of falling short and feeling excluded. It is relationally triggered when there is a subjectively meaningful emotional disconnect within a significant relationship, and elevations in shame can themselves evoke a reactionary dissociative response that may serve an acute regulatory function, reducing the intensity of the shame experience. The relationship between shame and dissociation appears bidirectional, and a new study published in Current Psychology by Alix Harding, Martin J. Dorahy and Kumar Yogeeswaran of the University of Canterbury builds on recent work investigating elevations in shame following dissociative experiences, asking a deceptively simple question: when does dissociation actually make us feel ashamed?

Previous research offered conflicting answers. Using vignettes describing dissociative experiences in three different relationship contexts — in the presence of an acquaintance, a close friend, or alone — McKeogh and colleagues found in 2018 that shame was elevated only in the close friend context, suggesting that more intimate relationships may moderate dissociation-induced shame. Yet another vignette study by Dorahy and colleagues showed only tentative support for intimacy moderating the connection between dissociation and reactive shame, instead supporting shame as a more general reaction to dissociative experiences regardless of context. Behavioural responses added another layer: those who dissociated with a close friend wished to stay present and talk about what had happened, whereas those who dissociated with an acquaintance wished it had happened when they were alone. These approach and avoidance motivations are common responses to shame, and whether the self can be repaired or needs protection from further harm is thought to determine which motivation emerges.

The Canterbury team suspected the missing variable was self-discrepancy. Self-discrepancy theory, proposed by E. Tory Higgins in 1987, posits three primary domains of self: the actual self, one’s perception of one’s actual attributes; the ought self, the attributes one feels obligated to have; and the ideal self, the attributes one would like to have. These domains operate from two standpoints — one’s own view and the perceived view of an other. A difference between two self-state representations is a self-discrepancy, for example a feeling that one is falling short of one’s own sense of who one should be. Higgins argued that shame is induced when there is a discrepancy between one’s actual and ideal self-state representations. This claim was challenged by June Tangney and colleagues in 1998, who demonstrated that most people do not distinguish between their ought and ideal selves, or between their own evaluations and those of significant others, theorising instead a general discrepancy between the actual self and some perceived optimal self. The present study therefore focused on an own/actual–own/ought discrepancy, examining whether dissociation that conflicts with how an individual believes they ought to be evokes greater shame than dissociation consistent with this self-guide.

The experiment used hypothetical vignettes written in second-person narrative to maximise immersion and self-referential processing. Participants aged between 20 and 72 years (N = 310), recruited through the Cloud Research crowdsourcing platform, were randomly assigned to one of four conditions: no discrepancy with a close friend, discrepancy with a close friend, no discrepancy alone, or discrepancy alone. A priori power analysis using G*Power estimated a required sample of at least 293 participants to detect a medium-sized interaction effect in a 2 × 2 between-subjects MANOVA with alpha = 0.05 and power = 0.99. The twelve vignettes, adapted from previously published work and pilot tested on seven postgraduate psychology students, described three dissociative experiences: amnesia, in which the person became disorientated, drifted off, and returned unaware of what had happened; a flashback, in which the person lost contact with the world and was transported back to a past traumatic experience; and depersonalization/derealization, in which the person felt detached from their physical body and perceived objects in the room as a long way away. In the no-discrepancy conditions, the person behaved in a way true to their ought self during dissociation; in the discrepancy conditions, the person behaved in a way that violated their ought self. Key words and phrases were bolded to emphasise the dissociative experience, the relationship context and relevant characteristics of the ought/actual self.

Shame was assessed along two qualitatively different dimensions. Direct recognition of shame was captured by a single-item measure in which participants rated the extent to which they felt shame, with shame and embarrassment combined into one rating as members of the same affect family. Indirect shame was measured with the five-item state shame subscale of the State Shame and Guilt Scale, whose items describe shame phenomenologically without naming it — for example, feeling small or wanting to sink into the floor and disappear. The scale showed excellent internal consistency in this study, with a Cronbach’s alpha of 0.94. Participants who reported any shame also rated seven potential explanations, including feeling exposed, feeling flawed, feeling a loss of control over oneself, and fears of being judged or rejected if someone saw them like that. Behavioural responses, adapted from McKeogh’s earlier work and developed in part from Nathanson’s Compass of Shame, ranged from talking to the close friend about it to quickly leaving the room, hiding the head in the hands, getting annoyed with oneself, or sitting calmly. Trait shame was controlled using the 25-item Experience of Shame Scale (alpha = 0.97) and trait dissociation with the 22-item Detachment and Compartmentalisation Inventory (alpha = 0.97), both containing embedded validity checks that excluded inattentive or socially desirable responders.

The results were technically revealing. Controlling for trait shame, the multivariate analysis showed a significant effect for relationship context on single-item emotions, with ashamed and guilty ratings significantly higher in the close friend context than when alone, while surprised and anxious ratings were significantly higher when alone. There was also a significant multivariate effect for discrepancy status: ashamed, surprised, guilty and anxious ratings were all significantly higher in the discrepancy context, while relaxed ratings were significantly higher when no self-violation occurred. Crucially, the state shame measure told a different story about context. The ANCOVA showed a significant main effect for discrepancy status, with state shame significantly higher in the discrepancy context, but no significant main effect for relationship context and no interaction. The expected amplification — that shame would be highest overall when self-discrepant dissociation was witnessed by a close friend — was not supported, as the interaction between relationship context and self-discrepancy was not significant for either shame measure. Supplementary analyses confirmed that participant age did not account for these effects.

The shame explanations proved equally informative. Discrepancy status significantly influenced how participants attributed their shame: those experiencing self-discrepant dissociation attributed their feelings significantly more to being exposed, feeling flawed, and feeling a loss of control over themselves. This aligns with research indicating that people feel shame when a self-flaw is revealed — in this case, dissociative experiences inconsistent with their self-view. Behavioural responses showed significant effects for relationship context: the desire to talk about the experience was significantly higher with a close friend than when alone, while the desire to leave the room was significantly higher when alone. This contrasts with the classical view that shame is associated with avoidance; if there is an opportunity to soothe shame by talking to a friend, it may be preferred over avoidance. In the discrepancy context, ratings for hiding and getting annoyed with oneself were significantly higher, while sitting calmly was significantly higher when dissociation did not violate the ought self.

The divergence between the two shame measures is itself scientifically interesting. The fact that overt shame, assessed directly via the word shame, was higher in the close friend context supports previous research suggesting social connectedness moderates the dissociation–shame relationship. Shame is associated with social rejection, and the risk of shame-evoking rejection increases as intimacy rises. When alone, there is no risk of real social rejection, reflected in lower overt shame scores. Yet subtle shame indicators showed no differences between being alone and being with a close friend. One possibility is that relationship context particularly influences the explicit appraisal of an experience as shameful: having dissociation witnessed by a close friend may heighten social-evaluative concerns, making shame more readily identifiable without necessarily increasing the broader phenomenological features captured by the State Shame Scale. Alternatively, the divergence may reflect differences in the measurement properties of the two approaches. The findings cannot distinguish between these possibilities but suggest that how shame is assessed may be important when examining the influence of relationship context. It is also possible that close relationships have competing effects: while being witnessed may increase concerns about social evaluation, a close friend may simultaneously represent safety, acceptance or an opportunity for support.

The study has limitations worth noting. Participants imagined dissociative symptoms rather than experiencing them, and perceived realism and immersion were not assessed. There were no direct manipulation checks confirming that participants perceived the intended self-discrepancy, omitted to avoid making the study aims transparent and introducing demand characteristics. The characteristics of the imagined close friendship, such as duration or perceived intimacy, were also not measured, and the three dissociation types were collapsed into a single composite variable, which does not account for the heterogeneity of dissociative experiences. Only the own standpoint of self-discrepancy was used; future research should include vignettes violating the perceived standpoint of others, particularly relevant for understanding the psychology of victims of interpersonal trauma. Nevertheless, the clinical implications are potentially valuable. Where dissociation occurs during therapy, shame may be heightened if the client’s dissociative behaviour violates their ought self. Therapists could examine clients’ perceptions of how they are and how they ought to be, and normalise dissociative symptoms where appropriate, to reduce negative self-appraisals and associated shame. Given that shame and dissociation have been implicated in self-concealment, interventions that reduce shame and dissociation, or their relationship, may ultimately benefit therapeutic safety and progress — though these applications require empirical validation in clinical populations. In summary, the risk of shame as a response to dissociation may be especially heightened when the dissociative experience feels discordant with the individual’s self-representation of who they ought to be or what they ought to experience.

Subject of Research: Psychology

Article Title: The effects of self-discrepancy and relationship context on dissociation-induced shame

Article References: Harding, A., Dorahy, M. J., & Yogeeswaran, K. (2026). The effects of self-discrepancy and relationship context on dissociation-induced shame. Current Psychology, 45(20), Article 1625. https://doi.org/10.1007/s12144-026-10175-9

Image Credits: AI Generated

DOI: 10.1007/s12144-026-10175-9

Keywords: dissociation, shame, self-discrepancy theory, social psychology, Current Psychology, mental health, trauma, effects, self-discrepancy, relationship, context, dissociation-induced