Self-invalidation in borderline personality disorder: A content analysis of patients’ verbalizations

Abstract

Objective:

The ability to trust one’s own perceptions is crucial for psychological well-being and growth. The relevance of its opposite, self-invalidation (SI), to the psychopathology of borderline personality disorder (BPD) is emphasized in many contemporary theories of evidence-based treatments for BPD. Empirical research on this topic remains scarce, however. This study aimed to describe manifestations of SI in individuals with BPD during a 40-session psychoeducational intervention based mainly on schema therapy.

Method:

Transcripts of videotaped group sessions were analyzed inductively using qualitative content analysis.

Results:

SI emerged as a recurrent, ubiquitous phenomenon. The content analysis yielded three core categories of SI: (1) a self-critical and harsh attitude towards the self (subcategories reflected punitive internalizations that could engender fear-based inertia, self-erasing, submissive coping behavior, and temporal fluctuation in SI), (2) a deficient sense of normalcy, and self-doubt, and (3) self-stigma. We also found an association of SI with various dimensions of BPD, including difficulty in the identification of emotions, secondary emotional reactions such as guilt, shame, anger, and resentment, self-related and interpersonal problems, and suicidal urges.

Conclusions:

SI is a detrimental cognitive-emotional process relevant to BPD that merits treatment. Efforts to reduce self-stigma, a pernicious aspect of SI, are imperative.

Keywords:

• Borderline personality disorder
• self-invalidation
• self-stigma
• qualitative research
• content analysis

Clinical and methodological significance of this article: Despite the hypothesized centrality of self-invalidation for borderline personality disorder in many contemporary psychotherapies, research on this topic is scarce. Applying inductive content analysis to a large data set consisting of 80 hours of videotaped group sessions, this study found that individuals with borderline personality disorder strongly experienced self-invalidation. Self-invalidation adversely affected the processing of emotions and self, and interpersonal interaction.

Introduction

Although individuals with borderline personality disorder (BPD) may be well-known for being convinced of their ideas, they are also prone to self-invalidation (SI). SI refers to doubting or questioning the authenticity of one’s feelings, ideas, or experiences (Livesley, 2017).

It can be hypothesized that SI encompasses different dimensions, such as implicit and explicit facets, along with temporal fluctuation. With respect to temporal fluctuation, Linehan (1993) describes how individuals with BPD may initially observe themselves accurately but thereafter discount their perceptions owing to self-mistrust. Thus, Linehan refers to the ability to maintain trust in what was observed (and apparently at least tentatively validated) a moment ago. However, due to the absence of empirical research, the dimensions of SI, including its temporal fluctuation, deliberateness, or the level of awareness individuals have over their SI, remain poorly understood.

SI can manifest itself in the form of overt judgmental thoughts, such as “I am a bad person,” or “I don’t deserve to feel better” (Manning, 2019). Other manifestations include trivializing one’s distress (Livesley, 2017) and oversimplifying the ease of problem solving, expressed by denying one’s problems or blaming oneself for them (Miller et al., 2017). The individual’s internal representations of the self may be polarized and distorted (i.e., “all-bad”) (Kernberg, 1975). S/he may turn against the self with self-blame and self-hatred (Koerner, 2012), or in the belief that s/he deserves to die, even want to commit suicide as self-punishment (Miller et al., 2017). SI may also allow individuals to avoid dealing with their emotions, thereby functioning as a regulatory mechanism (Linehan et al., 2002).

Dialectical behavior therapy (DBT; Linehan, 1993) posits that SI stems from an invalidating environment that fails to teach individuals when to trust their own emotional and cognitive responses as reflections of valid interpretations of individual or situational events. Individuals then adopt the characteristics of the invalidating environment. Mistrusting their own internal states, they rely instead on the environment for clues on how to respond (Linehan, 1993).

Schema therapy (ST; Young et al., 2003) describes a severe self-punitive state, the so-called punitive parent mode (Online Supplement 1), in which individuals condemn themselves as bad and evil or as deserving of punishment (Arntz et al., 2005). This mode is hypothesized to gestalt an internalization of one or both parents’ rage, hatred, loathing, abuse, or subjugation of the individual as a child (Young et al., 2003). The punitive parent mode affects information processing in different phases. As Valkonen (2018) states, in this mode, it seems as if self-observation is executed mainly from an extremely critical observer position, or from an observer position that is occupied by an internalized other, that is, a punitive parental introject. Being performed by an abusing other, self-observation manifests an extremely negative or harshly critical tone that renders it highly problematic (Valkonen, 2018). Besides affecting self-observation, the punitive parent mode is characterized by specific patterns of thought that typically concern the invalidity of one’s own opinions/wishes/emotions along with beliefs that one has no right to express these (Arntz et al., 2005). ST also identifies the so-called mode of compliant surrender. According to ST theory, this submissive mode, hypothesized to be driven by fear, attempts to protect the individual from further exposure to invalidation, rejection, conflict, or abuse, thereby functioning as safety behavior (Arntz et al., 2005; Young et al., 2003).

In a similar vein, mentalization-based treatment (MBT; Bateman & Fonagy, 2004) assumes that due to traumatic experiences in the attachment relationship, individuals with BPD feel that they are evil or hateful because they have internalized evil into part of the self (the “alien self”). More specifically, persecution from the maltreating person is experienced from within; part of the self-structure feels a desire to destroy the rest of the self (Bateman & Fonagy, 2004). This can be understood as an extreme manifestation of SI.

SI is not without consequences. It is proposed that the ability to evaluate one’s behavior nondefensively and to trust one’s own self-evaluations is crucial to growth (Linehan, 1993). Conversely, it is hypothesized that insidious doubt concerning one’s own perceptions (Livesley, 2003), responding to one’s emotional states with negative secondary emotions such as shame, disgust, and anger (Miller et al., 2017), and a tendency to look for external sources of validation all hinder the development of an adaptive self-system (Safer et al., 2009). Moreover, this self-invalidating cognitive style may hinder self-understanding (Livesley, 2017) and the establishment of personal goals and the development of a sense of agency (Livesley, 2003).

To the best of our knowledge, empirical research on SI is scarce. However, overlapping concepts, such as self-criticism (Shahar, 2015; Werner et al., 2019) and self-silencing (Jack, 1991), have been investigated. Self-criticism refers to the tendency to set oneself unrealistically high standards and to adopt a punitive stance towards oneself once these standards are not met (Shahar, 2015). Empirical evidence supports the conceptualization of self-criticism as an important vulnerability factor for psychopathological development (Werner et al., 2019) and the role of parental criticism (Muralidharan et al., 2015) and rejection (Campos et al., 2013; Sobrinho et al., 2016), attachment avoidance and anxiety (Dagnino et al., 2017; Martins et al., 2015), and childhood maltreatment (Falgares et al., 2018; Glassman et al., 2007; Sachs-Ericsson et al., 2006) in the development of self-criticism. In addition, psychotherapy research indicates that compared to their less self-critical counterparts, patients with high levels of self-criticism show a less favorable response to psychotherapy (e.g., Marshall et al., 2008). Compared to self-criticism, self-silencing has received less empirical attention. Using a phenomenological, descriptive approach and a longitudinal study design, Jack gave voice to 12 depressed women (Jack, 1991). These women described an inner dialogue between their authentic, first-person part and another, internalized voice that relentlessly judged them harshly. They often responded to the latter by self-sacrificing and self-negating behaviors. Overall, the condemning voice confused these women, obscuring what they knew by experience and creating inner division. A measure, the Silencing the Self Scale (TSSS; Jack, 1991, 2017) that assesses womens’ beliefs about the self in intimate relationships, was also developed and validated in a larger, heterogeneous sample of women (Jack & Dill, 1992).

Study Aims

Despite the hypothesized centrality of SI for BPD in many contemporary psychotherapies, few researchers have addressed this phenomenon. The present study contributes to filling this gap. This qualitative study comprised individuals with BPD attending a predominantly ST-based psychoeducational intervention. As our approach to the data was inductive, meaning that we allowed relevant themes to emerge freely from the data, the only research question was: what emerges as the most pertinent phenomenon from this data set comprising 40 transcribed group sessions? Since this turned out to be SI, we explored its manifestation in group discussions.

Method

This study is part of a larger process-outcome study which is described in detail elsewhere (Koivisto et al., 2021). The study involved a community mental healthcare services center (henceforth the center) in the City of Jyväskylä, Central Finland.

Participants

Research Subjects

A total of eight outpatients were included in the study; seven of them were female. Patients were aged 23–42 (mean 30, median 26) at study start. At baseline, the participants’ mean Borderline Personality Disorder Severity Index (BPDSI; Arntz et al., 2003) score was 31.1, indicating moderate to severe symptoms. On average, the participants suffered from a marked functional impairment, as shown by the fact that only two were working or studying at study entry. One patient was attending a work try-out as occupational rehabilitation and five were receiving disability payments. No structural assessment of functioning was performed.

Researchers

The present authors are psychiatrists and cognitive-integrative psychotherapists specialized in the treatment of BPD. TM has a PhD and is also a psychodynamic psychotherapist. SL, MD, and PhD is a professor in psychiatry and one of the developers of the intervention, while MK and TM, who analyzed the data, had no involvement in either the development of the intervention or the organization that delivered the treatment.

Instruments

BPDSI Interview

The Borderline Personality Disorder Severity Index interview (BPDSI; Arntz et al., 2003) was administered three times: before entry into the study, at treatment end, and at the 12-month follow-up. It is a clinical interview assessing the frequency and severity of BPD symptoms during the previous three months. The purpose is to provide a quantitative index of current symptom severity. The BPDSI is based on the DSM criteria for BPD and consists of 70 items organized into nine subscales: (1) abandonment, (2) unstable relationships, (3) identity disturbance, (4) impulsivity, (5) parasuicidality, suicide plans and attempts, (6) affective instability, (7) emptiness, (8) outbursts of anger, and (9) paranoid ideation and dissociative symptoms. The frequency of occurrence of each item over the previous three months is rated on an 11-point scale from 0 (never) to daily (10). Answers are then scored from never (0 points) to daily (10 points) or rated on a 5-point Likert scale (Giesen-Bloo et al., 2010). Previous research has found a cutoff score of 15 between patients with BPD and controls, with a specificity of 0.97 and a sensitivity of 1.00 (Giesen-Bloo et al., 2010). Recovery is defined as achieving a score of less than 15. Reliable change, which reflects individual clinically significant improvement, is achieved when the improvement is at least 11.7 points (Nadort et al., 2009). The present study used the translated Finnish version of the BPDSI (Leppänen et al., 2013).

Procedures

Setting and Recruitment

Participants were recruited from the center, whose services form part of Jyväskylä’s municipal secondary, specialized psychiatric services. Professionals working at the center were approached, informed about the study, and asked to refer patients aged 18–65 years with BPD symptoms for potential recruitment. The professionals, as part of their routine work, then informed patients with BPD symptoms of the possibility to participate in the study. The intervention which formed part of the study is routinely offered to individuals with BPD being treated at the center and thus was not controlled for in the study. Hence, patients were recruited for both the study and the group treatment simultaneously. We assessed potential participants in order of referral. Owing to financial constraints, we could study only one treatment group. Therefore, when the number of eligible participants reached eight, recruitment ceased.

The inclusion criterion was a BPD diagnosis based on the Diagnostic and Statistical Manual of Mental Disorders, 5th Edition (DSM-5; American Psychiatric Association, 2013). Exclusion criteria were a DSM-5 diagnosis of a psychotic disorder or a substance use disorder necessitating pretreatment detoxification. Exclusion criteria were assessed clinically only, with no other structured evaluations. The referred patients were assessed for eligibility based on the Finnish version of the BPDSI (Leppänen et al., 2013), with no other diagnostic evaluations. All the eligibility assessments except one (performed by SL) were performed by MK.

Treatment

Group Intervention. The intervention, comprising 40 weekly 2-hour psychoeducational group sessions and implemented between August 2017 and June 2018, was developed to meet the needs of public mental health services (Leppänen et al., 2016). The group was facilitated by two experienced psychiatric nurses who delivered the treatment as part of their routine work at the center. The framework integrates elements drawn from cognitive and behavioral treatment models designed to treat BPD. One of the main components of the intervention is patient education using the concept of schema modes (Online Supplement 1). The intervention also includes education in the development of BPD and DBT skills.

Adjuvant Treatment. In addition to group treatment, all patients continued their preexisting treatment as usual (TAU) at the center. TAU consisted of weekly individual sessions provided by psychologists or psychiatric nurses as well as medication. TAU was not linked to the group intervention, and therefore we did not control for it. While some of the individual therapists were familiar with BPD treatment or the treatment model studied, others were not.

Data Collection

The present data consist of 80 hours of videotaped group sessions (40 videotaped group sessions each lasting two hours). All group sessions were delivered at the center, conducted in Finnish, and videotaped with two cameras to ensure that all participants would be visible simultaneously.

Data Analysis

The data were analyzed using inductive content analysis. The data sensitivity of content analysis informed our choice of method, since we wished to remain close to the participants’ lived experience by employing description but a relatively low level of interpretation (Sandelowski & Barroso, 2003). Moreover, inductive content analysis is suitable for the exploration of phenomena uncovered in existing research (Kyngäs, 2020).

MK and TM watched 80% of the videotaped data in each other’s company. MK transcribed verbatim all the videotaped group sessions. Due to a lack of additional transcribers, parts consisting solely of psychoeducation were left untranscribed. This yielded a total of 374 pages of transcribed data.

Our approach to the data was inductive, meaning that we allowed relevant themes to emerge freely from the data. In analyzing the data, we followed the guidelines for inductive content analysis as described by Kyngäs (2020). Hence, the analysis was executed according to the following steps: data reduction, data grouping, and formation of concepts, i.e., data abstraction. Since SI emerged as a ubiquitous theme, this study set out to explore it. In the data reduction phase, MK extracted all the parts of the transcribed group session data pertaining to SI and compiled them into a single text. In the data grouping phase, all utterances expressing a single idea related to SI were assigned an open code. For instance,“if someone needs me, I never say ‘no’ but reschedule or cancel my own engagements, otherwise I feel really bad” was initially coded “self-erasure.” In the data abstraction phase, the similarities and differences in the content of these open codes were compared to determine which codes could be grouped together to form larger sub-concepts, such as “self-erasing, compliant behavior.” Based on similarities and differences in the content of the sub-concepts, the data abstraction phase continued until no (or not enough) shared meaning between the sub-concepts was found and the core categories could be constructed (Kyngäs, 2020).

MK and TM negotiated the clustering decisions made by MK. SL read the transcribed data and, in negotiation with MK and TM, supervised the clustering. No other validation strategies were applied. Finally, the data were quantified. MK counted the number of utterances related to each core and subcategory.

Reflexivity

In qualitative research, researchers are interpreters of basically ambiguous human experience and hence their preconceptions will inevitably affect the findings (Binder et al., 2012). More specifically, what is oriented towards and noticed as well as what is left out of awareness, is unlikely to be random but rather selected and affected by the researchers’ preconceptions. Here, we attempted to deal with our biases using reflexivity (Morrow, 2005) as a tool and, perhaps most importantly, noticed that the relevance of SI to the psychopathology of BPD was indeed coloring our preunderstanding. Since we took care not to select themes in line with our own preunderstanding, we questioned the choice of studying SI in dialogical interchange. However, since SI was a recurring and apparently poignant phenomenon that manifested in 39 out of 40 sessions, we selected it as the theme for this study. Furthermore, our primary intention was to describe the data with minimal interpretation. In all instances, we back our interpretations with excerpts from the data in the hope that they will increase transparency and help the reader follow our reasoning.

Ethics

The study was approved by the ethics committee of the Central Finland Health Care District on 9 May 2017 (No. 10U/2017). All participants signed a written informed consent after receiving a full description of the study procedure.

Results

Descriptions of how invalidation had been adopted as the patient’s stance towards emotions and the self were abundant in the data. Specifically, the content analysis yielded three core categories of SI: (1) a self-critical and harsh attitude towards self, (2) a deficient sense of normalcy, and self-doubt, and (3) self-stigma. Online Supplement 2 depicts these core categories and their relative proportions.

1. A Self-critical and Harsh Attitude Towards the Self

All eight participants exhibited critical, self-disparaging processing that made them feel vulnerable, insecure, blunted, dysfunctional, worthless, not deserving of good things, or suicidal. Of the total of 534 self-invalidating expressions found, 440 (82.4%) were in this core category. Four subcategories were identified: (a) self-critical and harsh self-observation and self-talk, (b) self-erasing, compliant behavior, (c) initial self-validation followed by subsequent self-erasure, and (d) deliberate display of counterfeit reactions as a coping strategy. The relative proportions of these subcategories are presented in Online Supplement 3.

A. Self-critical and Harsh Self-observation and Self-talk

Sarah: It is as if I had a voice in my head telling me … “you never do anything right”. Somehow I feel this voice isn’t even my own.

Mary: Yeah, I feel like you. In these moments, I feel the urge to injure myself physically, to hit myself, for instance. And if I happen to look at myself at the mirror, I feel like breaking the mirror.

The self-disparaging attitude was ubiquitous, affecting different phases and areas of processing, from observation to overt thinking. All eight participants exhibited self-observation or self-talk of this kind. With a total of 287 expressions, this subcategory accounted for 65.2% of the core category “self-critical and harsh attitude towards the self.”

After being acquainted with the concept of the “observer self,” one patient kept asking whether this type of observation includes emotions. Further exploration of her question revealed that she habitually observed her experiences from an extremely critical perspective that communicated a disallowing of her own emotions and thoughts:

It’s interesting how easy it is … or at least how easy it is for me to confuse the punitive parent with the observer self. The punitive parent has become or pretends to be my observer self … It tells me I’m not allowed to experience my feelings and thoughts.

Another patient noticed how she experienced a critical, punitive internalization as well as her emotional reaction to this internalization at the level of a child:

I’ve noticed that I still experience the punitive voices at the level of a child. They are exactly same voices I was taught when I was young. I haven’t had the chance to outgrow them.

Participants experienced bewilderment at noticing how the harsh processing occurred at an automatic level, escaping meta-awareness and therefore being difficult to discern:

Amanda: What is this … if I’m, say, in a cheerful mood … Like life is wonderful, and here I am happily driving my car and feeling fine. Then, all of a sudden … you just crash … There comes the experience “you failed in this” or “you said this or that to somebody”. Those thoughts just pop up, hit me, and place a damper on my mood. Is it that I don’t deserve to feel good? Or what is it? And why does it come like that?

Pauline: That’s exactly what I’ve been wondering. Is it that I don’t deserve it?

Amanda: This occurs particularly when I’ve just achieved something. When I’m feeling happy because of succeeding at something … Then, abruptly, like [makes a gesture that represents shooting self in the head] … Then I feel ashamed … like why did I undertake this task at all? … And I feel like crying and would like to drive under a truck even though I was happy smiling just two minutes ago.

Sarah: Exactly. It’s particularly after those happy moments that the punitive parent’s voice drags you down.

According to the present data, typical triggers for the activation of the critical, self-disparaging type of processing were performance (including success), interpersonal interaction, and the activation of emotional reactions. As the above quotation indicates, downplaying of the self can blunt positive experiences without the individual necessarily being aware of the process. Focusing on or rejoicing in success triggered thoughts such as “don’t think you amount to anything,” and “it’s bad to let your happiness show.” This seriously hindered the reception of positive experiences: “the punitive voice wants me to stay stuck.”

Overall, performance was an area in which the self-critical and harsh processing was particularly active. The participants anticipated harsh or arbitrary evaluation and believed that mastery should be achieved immediately and without practice. This left them with a constant fear of failure that often resulted in freezing, resignment, or inertia, thereby paralyzing the capacity for healthy action:

I’ve always felt I can’t affect my life. In fact, I haven’t even tried to … I’m so afraid that it will turn out that I’m not capable of doing anything right.

Agentic pursuit of own goals was also hindered because participants felt they were not entitled to it:

Pauline: It’s interesting that these exercises almost always end up with a conclusion like “how would you like to think or act if you were able to act as you want?” ‘Cos often, I’m not able to do this because I feel I don’t deserve it. I’m not allowed to even think about what I would like … 

Sarah: Yeah, it’s difficult to even know what you would want for yourself.

Pauline: That [awareness of wants] is something that’s not allowed to exist.

Sarah: Exactly. Then you just freeze and feel like “I don’t wanna do anything”.

Interpersonal interaction was another common trigger for self-doubt and related feelings of uncertainty. Expressing one’s opinions or revealing felt emotions induced fear of having expressed something bad. Occasionally, self-disparaging processing manifested itself in the form of explicit self-talk that the participants were aware of. At other moments, however, the participants only noticed the emotional consequences (e.g., panic) while remaining unaware of the cognition part of the experience:

Such a wicked problem … Sometimes, I’m unable to notice any thoughts. Then it’s impossible for me to pinpoint the problem. I only sense that the child [part] panics because she feels she did something wrong.

Not infrequently, the critical processing triggered by interpersonal interaction included a guilt-inducing tone: “because of you, others are now feeling bad.” Participants expected that expressing themselves in their ordinary way would potentially hurt others and feared the loss of connection. When the interpersonal situation was over, they would scrutinize their actions. This post-event processing had the potential to induce emotions so overwhelming that they even wished to skip future group sessions.

Sarah: … Even though I’d said something positive, and even though the reception was ok, and everything was supposed be all right, afterwards, I’d think anyway: … “Why didn’t I think it through!”, and “I shouldn’t have said it.” A kind of fear, or a strange panic … 

Pauline: If everything you did was always wrong, you just get used to it.

Finally, activation of one’s own emotional reactions often triggered self-critical processing, encapsulated in thoughts such as “this shouldn’t be a problem for anyone.” During the check-in round, Suzanne recounted an incident that was emotion-laden to her. Shortly thereafter, she invalidated her experience: “this sounds so slight … Anyway, I shouldn’t feel this way since nothing is wrong.” The participants’ narratives revealed the internalization of an oversimplifying attitude towards their inner reactions. This internalization communicated to them that it did not matter what they were experiencing, that negative emotions should not be expressed, and that they should behave well irrespective of their emotions. These participants anticipated that their reactions would be judged by others and were, therefore, prepared for guilt-inducing responses, punishment, and humiliation, and thus focused on self-protection instead of free exploration or sharing.

Processing of the hypothesized origins of the cruel attitude towards the self appeared helpful. All eight patients contributed to these discussions. Of the previously mentioned 287 expressions in the present subcategory, 87 (30.3%) were related to the developmental history of the harsh tone of self-observation and self-talk, as exemplified below:

Pauline: … it didn’t even matter what I did. If I blinked, it was a bad thing. If I left, it was bad. But if I stayed, that was bad, too … Well, I understand her [mother] because I know her background. But in any case, it’s wrong … And I even feel that’s the main reason for my being here in this group.

Several group members were able to relate to this and shared their own experiences.

Pauline: The common thread here is that we didn’t do anything wrong, but our treatment was demeaning, anyway. That’s the point, isn’t it?

Facilitator: Yes, it was very unpredictable. As you said: you blink or don’t blink but have no chance to predict the consequences of your actions.

Amanda: Or: “Because of you, daddy and I had to stay home … ”

Facilitator: What emotion does that induce?

Amanda: Guilt.

Facilitator: Like “this is my fault. I should do better somehow.”

Sarah: That’s exactly how I relate to myself nowadays, as an adult.

Facilitator: And how about shame?

Pauline: … You feel ashamed, too, because every imaginable thing is your fault. But you have no opportunity to compensate for it because it actually wasn’t your fault … Then you’re left with endless shame, since you can’t make amends or treat guilt that isn’t even real.

The harsh attitude towards the self was not without consequences. Participants recognized how it affected their processing of emotions. It engendered various secondary emotions, including guilt, shame, anxiety, anger, and resentment, as well as blocking of emotions:

Pauline [to Suzanne]: What you mentioned during the check-in round, I feel you. Like “it’s no big deal, so why can’t you just stop dwelling on it?” You kind of push the emotion away, even though you know it isn’t the right thing to do.

The critical, self-disparaging attitude towards the self affected the whole self. Participants even questioned their right to exist.

Pauline: As a child, I had to behave as if I was non-existent … 

Sarah: … I also feel I don’t have the right to exist, to be visible or to affect other people in any way.

This was followed by a discussion about how the protagonist in the Moomin tale “The Invisible Child,” became visible again; that is, through experiencing and expressing the whole spectrum of emotions, including the “negative” ones.

B. Self-erasing, Compliant Behavior

This subcategory describes behavioral responses to the harsh attitude taken towards the self. Seven of the eight participants talked about self-erasure. This subcategory contained 111 expressions and was the second largest subcategory of the core category “a self-critical and harsh attitude towards the self” (Online Supplement 3).

Earlier, I had a strong feeling or idea that I should kind of fade away … I shouldn’t make a fuss, but just silently evaporate, like vanish in order not to burden anybody.

Such self-abandoning behaviors included subservience and inadequate self-care, that is, denying oneself pleasurable or beneficial activities. A central example of the latter was making detrimental health choices. Specifically, participants reported denying themselves meals or adequate rest or sleep, as these were perceived as undeserved:

I find it hard to go to bed. As far as I understand it, it might have something to do with the fact that sleep is so important. It feels good, and I know I feel really bad if I don’t get enough sleep. But then, I’m not allowed to go to bed because I don’t deserve to feel good … My mind wants me to feel bad.

As opposed to deliberate, active self-erasing behavior, some participants’ self-erasing behavior was primarily characterized by indifference and lack of orientation towards the self:

It’s not about holding to my needs, but more about noticing them. That might be the point there: do I even notice them.

Disconnected from their own needs, participants silently focused on others’ needs:

I was so anxious. I just couldn’t decide. I wanted somebody else to tell me what to do. But then the healthy adult took the reins, and I messaged my friend “sorry, I won’t be able to drop by”. And I felt relieved right away. But at first, I felt I’d never be able to cancel it. Then a thought crossed my mind: “Perhaps I sometimes have the right to express that something is inconvenient for me.”

Self-erasing, other-oriented behavior often seemed to originate in the internal dynamics of the individual’s punitive-submissive counterparts, where the self was experienced as worthless and fear of the interpersonal consequences of noncompliance dominated the patient’s mental world. Thus, compliant behavior could serve to appease the harsh and critical voice:

When I’m not angry, I comply with everything ‘cos I fear the consequences of not complying … I never say ‘no’. That will only backfire … I go along, like “suits me fine” … Were I to negotiate with someone, like “you want it like that, and I want it like this”, I would always give up … I was explicitly taught that I don’t matter … I’m not allowed to think of myself or be assertive if there are other people around, to the point that I don’t even know what I want.

As evident from the excerpt above, self-erasing behavior constituted part of a vicious cycle that had detrimental effects on the self. However, the present data also included numerous examples of change-talk:

It just seems so silly that it’s so important to me not to ever hurt anybody, or that nobody should ever get annoyed because of something I said. I mean, people will inevitably have emotional reactions all the time. They’ll react to aspects I might not even be aware of.

Occasionally, self-sacrificing behavior was driven by the projection of one’s own, deprived needs onto others. Since the participants attempted to avoid inducing similar, presumably intolerable experiences in others that they themselves had endured as children, they tended to give excessively to others without considering their own boundaries:

It’s difficult for me to deny somebody something because it activates my own childhood experiences … Because you … did not feel quite understood, or people just said “no” to you without explaining it … Then, as a grown-up … you are careful not to repeat the same mistake. So, you don’t want to say just “no” but rather want to explain the reasons for that “no” so that the other person understands and won’t feel so bad about it.

Participants also recognized how self-erasure could function as a reenactment of the victim position. They became aware of how, in their inner dynamics, they continued to repeat the traumatic invalidation they had been exposed to during their personal histories. Paradoxically, the bittersweet victim position could even be a source of pride:

Like taking the victim role, even though at that moment, you don’t regard yourself as a victim … Kind of a Jesus spirit. I don’t mean in a religious sense, but anyway like: I surrender to be crucified for the sake of others, ha ha ha.

Compliant, self-erasing behavior was state-dependent as opposed to static. Subservient behavior might be followed by different mental states. In the subsequent angry, impulsive, and/or rebellious states, the compliant behavior temporarily subsided:

You volunteer to do it all by yourself, but then complain: “Fuck it! Why do I always have to serve you!” Ha ha ha.

Their collaboration in processing these experiences facilitated the participants’ awareness of their complex and often conflicting emotions. Further, they came to realize how they themselves contributed to the problem in the present, as opposed to their habitual enactment of the old victim role.

C. Initial Self-validation Followed by Subsequent Self-erasure

This subcategory refers to temporal fluctuation, that is, experiencing self-doubt and self-erasure following the initial validation one’s emotions, thoughts, or actions. The behavior itself (self-erasure) is analogous to that in the previous subcategory. However, since this subcategory is characterized by temporal fluctuation in the experience, we consider it merits existence as a distinct entity. Of eight patients, five described initial self-validation followed by subsequent self-erasure. A total of 15 such expressions accounted for 3.4% of the core category “a self-critical and harsh attitude towards the self” (Online Supplement 3).

Sometimes I’m in fact able to say “not now” but after a while, I feel so bad I go back on it. For instance, if I’d negotiated with my boyfriend about who was going to do the shopping … If we’d managed to settle it so that he’d be doing it, and he’d been ok with that and not even annoyed, I’d still get the feeling that it’s no big deal for me, like what’s the problem here for me, why can’t I just do it … Very often, I end up eating my words.

Self-validation followed by self-erasure was also found in group interaction. Neutral responses, perhaps experienced as lack of validation, sufficed to induce this, and the participants recanted, apologized, readjusted, or critically reflected on their personal contribution: “Am I being too talkative?” or “I’m talking way too much.” We observed numerous retreats from interaction with the group, as manifested in expressions such as “never mind” and in speech turns that commenced but soon after were aborted. Withdrawals into long silences also occurred. These sequences could be indications of initial self-validation followed by self-erasure. However, these data preclude the drawing of conclusions on the motives of silent behaviors, as alternative explanations also exist (e.g., angry retreats due to feeling misunderstood). Ambiguous utterances or behaviors of this kind were not included when the data were quantified.

D. Deliberate Display of Counterfeit Reactions as a Coping Strategy

Everything I say, I say with such uncertainty that it’s easy for me to recant anytime … It’s often not that I even wanted to agree in the first place. It’s more that I’ve learned that my opinion is automatically seen as irrelevant, or perfectly stupid, like: “You idiot, how can you think like that” … I’ve noticed I hardly ever act according to how I feel but rather opposite to my emotions … I try to appear as calm, composed and brave as I can … I can act sociable … I may appear happy … but that doesn’t feel good to me … I’m super good at faking positive emotions … When I’m angry, I might smile even more … But then, without anybody being able to anticipate it, I suddenly fall apart as I’m no longer able to act composed. I wonder how I could allow myself to show some [emotional reactions].

In this subcategory, as in the previous one, trust is initially experienced in one’s own reactions. However, unlike in the previous subcategory, this self-validation is maintained, with the individual deliberately displaying other, feigned reactions including inauthentic positive emotions, counterfeit opinions, or feigned self-invalidation. Three of the eight patients described behaviors in this subcategory. The 27 expressions in this subcategory accounted for 6.1% of the core category “a self-critical and harsh attitude towards the self” (Online Supplement 3).

Clearly, displaying inauthentic reactions while concealing true ones was a survival strategy. Pretended self-invalidation, for example, functioned to protect patients from the pain of social invalidation and related feelings of disappointment, shame, or humiliation. More specifically, since they had learned that others could be unresponsive to their feelings, ridicule their dreams and fail to share their joy or pride in their accomplishments, or even turn aggressive or destructive, they hid these experiences while presenting a self-downplaying façade supposedly acceptable to the invalidating other. Thus, besides protection of the self, the purpose of this strategy was to maintain contact with others.

I was taught to @be positive! Think positively!@ And so on. So, my attitude towards my own experiences has been like … if someone, say, sawed my leg off, I’d think about it like @fortunately I still have the other one left.@

Text inside @-marks is uttered in a phony voice.

Importantly, patients also applied this interpersonal strategy in their treatment:

Previously, I only talked to my therapist about issues I assumed she wanted to hear. For instance, I told her that everything is fine and that I’m busy studying … But then I noticed it didn’t help me, and somehow, nowadays, I’m able to voice all the thoughts that go through my mind.

Participants recognized the harmful consequences of this coping strategy; displaying counterfeit behavior entailed consequences for both oneself and one’s relationships. It led to a vicious circle where one was misperceived – and thereby invalidated – by others:

Sarah: I’ve noticed that my actions have no correlation with who I really am. I mean, I feel my behavior reflects a completely different person. One can’t infer anything from my behavior because I play a role without ever revealing my true self. No wonder people are unable to know me; no wonder I’m given gifts I don’t even like … Gee, I realized how I was taught to deal with emotions when I was a child. It’s as if they inserted a script in my brain with a message telling me “don’t pay attention to how you feel, but just behave”.

Pauline: And those gifts, they are for the self you present outwardly.

Hidden underlying needs were unnoticed by others and therefore not responded to:

Sarah: I get a lot of positive feedback but it’s not the kind of feedback I’d need. It feels too superficial.

Facilitator: The underlying need is not met.

Sarah: No, it’s not.

Pauline: The feedback is not authentic because others are unable see the real me.

This discordance between felt, inner states, and those presented to others had an alienating effect on participants’ relationships.

2. A Deficient Sense of Normalcy, and Self-doubt

Don’t make a scene about this, since it might just be your BPD kicking in.

Content analysis yielded a second core category of SI that was characterized by doubting one’s perceptions. Participants experienced global uncertainty about “what is normal.” Being diagnosed with BPD added another layer to self-mistrust, since one’s perceptions could potentially reflect personality pathology. Six out of eight patients described experiences related to this core category. The 79 expressions observed in this core category accounted for 14.8% of the total of 534 self-invalidating expressions (Online Supplement 2).

Participants experienced confusion in attempting to figure out to what extent they could trust their reactions: “I constantly monitor my emotions and thoughts, but I’m uncertain whether I can trust them or not. How can I tell a misinterpretation due to BPD from a valid reaction? How do I know what is justified?” Adjustment of one’s emotions and emotional needs and how to express them were issues the participants were particularly eager to work on:

Amanda: How can you tell what are realistic limits? I mean, for instance, in relationship with a partner. I’ve been reflecting on whether these are just my own issues, or this BPD again. Like am I asking too much from him. Like do I dare to mention my needs out loud or does he think “you’re just crazy”.

Sarah: That’s exactly what I fear, too. Always.

Amanda: Then you give up expressing your needs.

Suzanne: I get so hurt so easily that I feel it just can’t be normal.

Sarah: The problem is that you’re unable to discriminate between when it’s ok to get hurt, like when not to take offense is too much to expect from you, and when it’s not ok.

Pauline: It’s annoying to always have to distinguish between what is due to BPD and what is normal, I mean doubting whether my emotions are based on reality.

Sarah: I could easily attribute everything to BPD. I refrain from expressing nearly any hurt feelings since my automatic appraisal is that “this probably isn’t a real issue”. Then I analyze the situation in my mind and might consult my best friend … But I have the tendency to say to myself “you’re just exaggerating”. Then I feel like “you shouldn’t feel hurt because of that”.

Pauline recognized the devastating effects of SI on the processing of emotions and noticed that by self-invalidating, she continued to repeat the previous traumatic invalidation in the present:

Earlier, I did that a lot, too, and still do … But I feel it’s very dangerous to always bypass your experience, judging your reactions as wrong … and not fitting the situation. Then you say to yourself … “how foolish” … or “that’s irrelevant”. Doing that, I’m not open to my emotions, I avoid facing them … As you said [to the facilitator, referring to what had been read out loud a moment earlier]: “had no permission to express his or her emotions”, I continue to repeat it myself.

“Normalcy” was an intriguing but elusive concept for the participants. In addition to a wish to learn to adjust their reactions, one motivation behind the frequently posed question “What is normal?” was participants’ limited sense of normalcy. Looking back on their developmental histories, participants felt they had been left without templates for normative reactions and thus deprived of the opportunity to learn to gauge their reactions and how these affected other people. Moreover, participants’ childhood experiences that had disrupted the development of an understanding of “what is normal” included deprivation of validation and congruent feedback:

Facilitator: … of course you felt bad.

Pauline: … You said “it was an app-rop-riate reaction”. That’s exactly what’s missing in my script. That doesn’t even exist in my world. I don’t have such a word. The point is: since my feelings were never validated … I’m unable to tell whether it’s ok to feel something.

Participants were prone to experience themselves as highly abnormal and deviant. Reflecting the stark contrast they perceived between themselves and others, they frequently spoke about “‘normal persons,” or “integrated persons” when referring to others. This seemed to drive a wedge between themselves and others and fuel additional judgmental labeling of their own reactions.

3. Self-stigma

From an outsider perspective, my life looks like a soap opera … I’m a drama queen … I’m so ashamed of being this kind of a person.

The third core category concerns the stigma and self-stigma associated with being diagnosed with BPD. All eight patients talked about it. The 15 expressions on this theme accounted for 2.8% of the total of 534 expressions of SI in this data set (Online Supplement 2). The present participants were painfully aware of the negative stereotypes and prejudices attached to BPD. They could apply these labels to themselves, thereby inducing additional feelings of worthlessness and shame: “What’s the logic behind this I-am-good-enough talk? After all, diagnostically, we are mentally ill. Try and think ‘I’m adequate just as I am,’ then.”

During one group session, Emma recounted an occasion when she had expected that disclosing her BPD diagnosis in a medical encounter would have ameliorated its psychological-interactional impact. To her disappointment, her disclosure failed to achieve this objective. Her fellow patients who had experienced stigma in medical encounters had no illusions about the consequences of disclosure:

Judy: I usually avoid mentioning I have BPD … I prefer to keep my cards close to my chest because it easily gets misconstrued. I’m borderline means … I’m an ugly person - I mean, psychologically.

Sarah: Yeah, I don’t like to disclose it either.

Judy: Really, I mean [if I disclose it], the interaction always turns like I am the source of all problems. From then on, they interact with me with kind of a psycho attitude, you know, in a way you interact with a nutcase … The diagnosis is more like a burden … you’re branded on your forehead, and you’re treated accordingly. And the mental diagnosis will then be emphasized in all sorts of irrelevant contexts.

Sarah: Persons with BPD have a bad reputation … I’ve noticed that if they know, then the treatment turns extra lousy.

Judy: … and when the staff report, they say “that one’s borderline; no wonder she might be a bit spiky” … That way, I’m denied the opportunity to evaluate my treatment anymore: I’m spiky because I’m borderline.

Occasions were observed where participants were extremely quick to latch on to expressions which they then attached to themselves. For instance, when the topic of mindful observing was presented in the group, the facilitator acquainted the participants with the idea of observing the drama of your life. The word “drama,” unfortunately, has a strong pejorative connotation in Finnish. This direct translation from English to Finnish may inadvertently have led one participant, who was speaking just before the facilitator took the floor, to call herself a “drama queen.” Moreover, although the present data preclude causal conclusions, we suspect that some of the wording (e.g., “borderline patients often are … ”) in the psychoeducational material might have invoked self-stigma talk.

Discussion

In this study, patients with BPD strongly experienced SI, as manifested by a self-critical and harsh stance towards the self, a deficient sense of normalcy and self-doubt, and self-stigma. A pernicious dialogue between different self-aspects (modes) was observed. In this dialogue, a harsh, punitive mode attacked other modes that reacted either with fear, concealment of authentic reactions, and outward compliance, or with anger, resentment, and rebellious behavior. Our findings show a striking number of similarities with those of Jack (1991), who reported that women with depression also shifted between these modes.

Sometimes the harsh, critical mode spoke to the patients using the words of their significant others. Hence, patients could experience this voice as ego-alien (“I feel the voice isn’t even my own”), possibly reflecting an introject not truly processed, but rather “swallowed whole” (Perls et al., 1994). Consistent with our findings, Zinker (1994) notes how such introjects are often force-fed by a significant other. This phenomenon, reflecting introjecting with minimal effort (Zinker, 1994), may resemble the MBT concept of the development of an “alien self” (Bateman & Fonagy, 2004).

In the psychotherapy literature, SI often refers to doubting one’s own perceptions (Linehan, 1993; Livesley, 2017). As proposed by Linehan (1993), individuals with BPD may also have very little sense of what is normal. Evidence from the present study supports these clinical observations. Moreover, we found empirical support for Linehan’s observations on the fluctuating nature of self-validation and SI (Linehan, 1993), that is, how individuals with BPD may initially observe themselves accurately but soon after discount their perceptions owing to self-mistrust.

The present participants described how SI obstructed their agentic behavior. These observations accord with those reported by Shahar et al. (2006) and Shulman et al. (2009), who found that a partly overlapping phenomenon, self-criticism, adversely affected young adults’ goal construal, predicting low levels of autonomous motivation and positive life-events. The present findings also provide evidence for the connection reported by these researchers between self-criticism and avoidance, that is, how self-critical individuals shy away from doing things they really want to do, and experimenting (Shahar, 2015).

Our findings on stigma and self-stigma support existing findings indicating that BPD is highly stigmatized (Grambal et al., 2016; Quenneville et al., 2020; Rüsch et al., 2006; Sheehan et al., 2016). Our findings suggest that to reduce self-stigma, the wording used in psychoeducation may be important. Other researchers have also emphasized the power of language and assumed that the way it is used may influence stigma construction (Aviram et al., 2006; Masland & Null, 2021).

SI is an important treatment target in many contemporary psychotherapies for BPD. Treatment for SI includes increasing patients’ awareness of the numerous ways in which they engage in SI (Livesley, 2017), both during sessions and in their lives outside therapy, for example, by asking them, “Do you notice how you said/felt X but then took it back? What happened?” Moreover, therapy strives to convey the idea that the harsh, punitive mode is merely one mode within the patient (Arntz & van Genderen, 2021; Shahar, 2015). Learning to observe mental events mindfully may aid in putting this mode in its place. Targeting SI also requires that the therapist constantly searches for the contrary, that is, the valid in patients’ responses and communicates this to them (Fruzzetti & Ruork, 2019; Linehan, 1993). Heightening awareness of the early causes of SI is also a necessary component of therapy. In ST, traumas contributing to the development of SI are treated using experiential techniques, including imagery rescripting (Arntz & van Genderen, 2021).

A strength of the present study is its large data size. The use of investigator triangulation, that is, the involvement of multiple observers and interpreters, may increase the credibility of the results. A limitation is that the content analysis of group session data can only describe SI as it manifests in explicit verbal expressions or overt behaviors, and thus neglects more subtle, less conscious, or unverbalized aspects of SI. Caution must be exercised in interpreting the results of the quantified data; the participant who talks least may be the one who suffers the most seriously from SI. Because the present patients were acquainted with the concept of the so-called mode of the punitive parent (Online Supplement 1) during the group intervention, it is possible that expressions related to self-criticism and harshness may be over-represented in our data at the cost of other aspects of SI. However, the patients had not been familiarized with the concept of SI in their treatment context.

Questions for future research include: What exactly is SI? What are its boundaries with neighboring concepts, such as self-criticism and self-silencing? What is the role of biological factors, for example, neuroticism, in SI? Do gender or cultural aspects in the upbringing of female children play a role in the development of SI? Does SI constitute part of a larger, overarching self-concept pathology (see for example, Shahar, 2015)? These questions should be explored using a diversity of methodologies. Experimental designs focusing on implicit in addition to explicit information processing are needed to further our understanding of how individuals with BPD trust or doubt their perceptions. Self-report instruments tapping a closely related phenomenon, that is, self-silencing in intimate relationships, already exist (Jack, 1991, 2017, 1992), and could be developed to explore SI. Qualitative research could also explore SI using in-depth interviews.

To conclude, SI may be a devastating vulnerability relevant to BPD. It may engender serious consequences, including difficulty in the observing of emotions, thwarted exploratory behavior, poor self-care, and suicidal urges. It is imperative that efforts are made to reduce self-stigma, one pernicious aspect of SI.

Acknowledgements

The authors are grateful to the participating patients and therapists (Merja Saarela and Nina Forsman) at the community outpatient mental health care services in the Central Finland Health Care District, without whose cooperation the study would not have been possible.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Additional information

Funding

This work was supported by the Finnish Association for Cognitive and Behaviour Therapy: [Grant Number Not available].