Changing patterns of self-presentation by depressed clients: from shame to self-respect

Depression, self and gender

The depression looked at in this study is not the severe, clinical depression that makes any social functioning nearly impossible. Only a milder yet chronic depression is meant here, which considerably burdens these people’s lives, but still allows them to work and study, albeit on a lower key: dysthymia, to use the terminology of the current DSM-IV. Dysthymia is rather pliable in daily life, as people are able to mask their problems for a long time as burn-out or exhaustion. Although current knowledge cannot say anything with certainty about the causes of depression, there is growing consensus that biological, psychological and social factors all play a role (Beck & Alford 2009). In this paper, the focus lies mainly on the social element, more precisely on the communication processes about self and depression.

By self (or identity, character) we refer to the awareness of one’s own place in the social world. This self is closely connected with what others think of it (Goffman 1990; Holstein and Gubrium 2000). People have a reflective capacity to think about themselves as subject and as object, a vision that goes back to Mead’s originally theory of mind, self and society published in 1934 (Mead, 1974). Depressed persons experience an incapacity of their will, thinking and feeling – in sum, of their self. They feel rejected and worthy of rejection, even though this self-image seems to be subject to constant revisions (Karp 1996; Metzl 2003; Westerbeek & Mutsaers 2008). According to the DSM-IV, low self-esteem is one of the core symptoms of dysthymia and of major depressive episodes.

The way in which people express emotions is one of the most important modes to display their female or male identity (Sheilds 2002). Given the fact that depression takes place largely in the domain of feelings (like gloominess, sorrow, taking no pleasure in anything anymore, hopelessness, despair, loathing and often fear too), this makes it a strongly gender-loaded subject. Western culture has no unambiguous gender scripts for expressing emotions. Fisher (2000) and Sheilds (2002) find that women are more allowed to express their emotions than men, and men must be more self-constrained. Women are ascribed the more positive feelings of being caring, whereas men can express aggression more openly. With regard to sadness, it is found that among the general population in the United States women mention this more frequently than men (Simon & Nath 2004), a finding consistent with the vast literature on gender and depression (Boughton & Street 2007; Piccinelly & Wilkinson 2000). Men are still expected to curb their despair and sorrow (Addis 2008; Addis & Mahalik 2003), although this is less the case among the highly educated cultural elite than in circles where a macho culture reigns, like the soccer world (Westerbeek & Mutsaers 2008; Wolpert 2001). However, most authors emphasise it is the specific social context which determines if and how one measures up to these gender script standards. Meta-analytical research also confirms that many gender differences can indeed be manipulated in terms of size and even direction by altering the context of the experiment (Hyde 2005).

In this study, depression, self and gender are considered as interrelated social constructs that are constituted in the context of countless small interactional situations. One of those microsituations, the psychoanalytic clinical assessment interview, is the subject of analysis here.

No elaborated theoretical framework is available for this research. From the symbolic interactionist vision on strategic impression management (Collins 2004; Goffman, 1990; Scheff 1994) we borrowed key concepts like self-presentation and the definition of a situation. Since more modern interpretations of Goffman’s ideas stress that they are not as voluntaristic as initially seen, and that in his perspective the system is always part of the microsituation (Collins 2004; Lehmert & Branaman 2008; Scheff 2006), we can embed this symbolic interactional perspective in the more general figurational sociology of Elias (1978; 2000) and, among others, Wouters (2007). Where Elias described a gradual process from the thirteenth century to 1850 in Western Europe in which a more equitable and encompassing self-control became disseminated in the social habitus of all social classes, Wouters points to informalising tendencies in the twentieth century, especially in the 1960s. Informalisation, also described as a ‘controlled decontrolling of emotional controls’ (Wouters 2007: 93) cannot be seen as a contradiction of Elias’s theory of the civilising process. On the contrary, informalisation requires, ‘a self-regulation that is more highly controlled as well as more reflexive and flexible’, and comes to the fore in periods of democratisation and growing social integration, which presses both parties to a subtler navigation in social relations (Wouters 2007: 93).

In figurational sociology, human actors are seen as thoroughly social, social situations as ever-changing formations of mutually interdependent human beings, and emotions as emanating from the way people are bound to live with each other. It is no coincidence that Elias, Wouters as well as Goffman, Scheff and Collins stress the fundamental importance of social emotions like shame and embarrassment for sustaining social reality. Above all, figurational sociology commits us to add a processual dimension to the snapshot approach of the symbolic interactionists. The figuration studied here is the psychoanalytic assessment interview during the second half of the twentieth century in the Netherlands.

Client-therapist interviews

Not much empirical research has been conducted into self-presentation in a therapeutic setting, especially taking possible gender influences into consideration. One of the first to do this is Wodak. Based on transcriptions of group therapy sessions, she concludes that men report their problems more briefly and in a business-like fashion, while women place their problems more in the context of their life history: ‘Women relate, men report’ (Wodak 1981).

In a retrospective article, Scheff (2001) looks back upon his earlier observations of 1965 during intake interviews with seventy elderly working-class men diagnosed with depression in a mental hospital near London. He had noticed these men put all the blame on themselves and spoke in low voices, stammering, without making eye contact. Only momentarily did they show more aliveness when asked for their activities during World War II, a period in which these old men had felt they belonged to a community. Years later, Scheff came to relate this observation to these men’s daily lives, within which they lacked a secure bond, and to the social distance with the psychiatrist they spoke to. Scheff now supposes these men were deeply ashamed for most of the interview. Although his observations exclusively regarded men, Scheff does not elaborate on a possible gender issue here and seems to interpret his findings more in terms of class inequalities.

A more theoretical contribution comes from Friedlander and Schwartz (1985), social psychologists inspired by Goffman. They developed a taxonomy of self-presentation strategies in the context of a counselling situation, albeit without paying attention to gender. They distinguish five self-presentation strategies of clients: a defensive or protective one, in which clients try to avoid the disapproval of the therapist (for instance, by giving apologies or justifications, by presenting themselves as inefficient in the hope of gaining sympathy); and four assertive ones, along gradually increasing lines from engaging to self-promotion and even intimidation.

Based on this taxonomy, Schütz et al. (1997) and Kelly (1996; 2000) describe self-presentation strategies during therapy and intake respectively. In their observations, the ambivalence of clients’ presentation strategies comes to the fore. Clients do not know exactly how they will be evaluated, and navigate between fear of being seen as a poseur or as a hopeless case, switching between protective and assertive strategies. Further, moderately depressed clients present themselves as more depressed than they deem typical of themselves (Kelly 2000: 477, also found earlier by Paykel & Norton 1986). However, clients are also anxious about evoking too negative an image about themselves, as labels like narcissistic, borderline and psychotic are feared. Kelly concludes that clients want to be seen ‘as good people who are bravely struggling with some problems’ (Kelly 2000: 476). As mentioned already, women seem to have a different presentation style than men in group psychotherapy (Wodak 1981), which in the context of medical interaction has been confirmed over and over again (Meeuwesen, Schaap & Van der Staak 1991; Roter & Hall 2006).

In conclusion, there are solid grounds to presume that clients manipulate the impression they make on their psychotherapist, consciously or not, even during intake interviews. They try to present themselves as persons who need help, with an outpatient clinic as the right setting. Two additional aspects need to be given attention: the cross-sectional nature of most studies so far, due to which they lack a perspective on any changes in self-presentation strategies over time; and the gender issue, which has hardly been explored in relation to self-presentation in a therapeutic context thus far.

Selection sample

We had the opportunity to study parts of the clinical records from the archives of the Netherlands Psychoanalytic Institute (NPI), which date from 1946 to the present. To assure a comparable and homogeneous selection, only those files were requested in which clients presented one or more depressive symptoms at the first interview. These files had to be evenly spread over the examined fifty years, and by the sex of both the client and the analyst conducting the talks. These files had to contain reports of the first intake interview as well as of the next three or four clinical assessment conversations, so we only investigated files from clients accepted for further psychoanalytic assessment. Out of their original 24,000 records, the NPI kept roughly 3600 anonymous files for purposes of scientific research. NPI staff collected 118 records from this corpus (1954 to 2000; 59 male and 59 female clients, quota sample), of which 108 (53 males and 55 females) were analysed for this article. The remaining ten files (six males, four females) contained no information about clients’ self-image – either it wasn’t mentioned at all or the client could not answer the questions about this subject. The assessment reports were intended to find out whether clients were suitable for psychoanalytic treatment and should result in a personality diagnosis as well as a treatment indication. To this end, mostly three or four conversations and a psychological assessment were necessary, as well as a staff meeting. The reports of these intake and assessment consultations contained an elaborate description of the symptoms, life history, current functioning and self-image of the client, supplemented with more subjective impressions by the psychoanalyst. For ethical reasons all further personal identifiers in the quotations in this article have been removed or disguised, so the clients described are not recognisable.

The clients whose documents we could study were on average 29 years old, half of them were single (men as well as women), and had finished higher education (or were qualifying for it), a level which increased over time. About 30 per cent of them were students at higher vocational schools or universities, and over time an increasing percentage had a job at academic level. Their depressive symptoms could be typified predominantly as milder but chronic depression – dysthymic disorder, as it has been called since 1990 when the NPI was obliged to adopt the DSM-III and subsequent versions. About 70 per cent of the clients from the 1990s had had psychotherapy before, compared to 36 per cent in the 1950s. In the 1980s and 1990s about 20 per cent of the clients themselves were psychotherapists, clinical psychologists or psychiatrists-in-training.

The setting

The assessment interviews took place in the consultation rooms of analysts associated with the NPI, at home or at the institute. Clients and therapists sat facing each other, and each interview lasted about 45 minutes. This setting as well as the subjects discussed remained unchanged for the fifty years over which information was collected.

The NPI, with locations in two larger Dutch towns, found its clientele among higher educated (upper) middle-class men and women in their twenties to forties. From 1965 to 1980 the Institute grew rapidly thanks to collective benefits. Since then it has become increasingly restricted by government and health care demands. Not much is known about the clients that were assessed here for their suitability for psychoanalytic treatment, but this is a relatively verbally apt and introspective selection of their general clientele. The psychoanalysts at this institute gradually became more trained as clinical psychologists and less as ‘nerve doctors’, and more of them were females. Compared to other Western countries, they left behind the traditional Freudian drive theory rather late, but from the 1970s onwards, a more relational vision of the early psychological development predominated. The assessment procedure became more formally reported, though its purpose remained unchanged during those fifty years (Westerbeek 2010).

In the eyes of analysts: validity

The data consisted of reports written afterwards by one of the two parties in those interviews: the assessing psychoanalysts. We did not have the opportunity to study the actual course of the conversation and have had to refer to the interpretations of the analysts. When preparing their report, psychoanalysts were bound to the logic of their therapeutic tradition and to the aim of the assessment procedure. In addition, by means of their report they wanted to present themselves to their colleagues as good observers and capable interpreters.

What psychoanalysts reported about their clients will not be untrue, yet it is certainly a selection of what they deemed worth mentioning. Thus, this research cannot study clients’ self-image or the image they have of their analysts. It is, however, possible to trace the specific self-image clients displayed when they were scrutinised for psychoanalytic treatment, as written down by the analysts that assessed them.

Analysis

The hard copies of the clinical records were converted into digital Word documents and imported into software for qualitative content analysis (MaxQDA). Based on grounded theory (Glaser & Strauss 1967) as well as on critical discourse analysis (Titscher, Meyer, Wodak & Vetter 2002), the reasoning method was inductive as well as deductive. Spiralling back and forth between empirical observations and existing scientific knowledge, our method could best be identified as abductive (Scheff 1994). To support the qualitative findings a descriptive quantitative analysis was conducted. Effects of time and gender were checked by Chi-quadrants.

During the qualitative coding process, all self-images were labelled and categorised according to the mentioning of negative and/or positive character qualities. And this resulted in a three-part typology of self-presentation strategies. To find out whether the coding meets the reliability requirements, inter-rater reliability was assessed. Of all the files, 6 per cent (8 reports) were coded by two researchers independently; the percentage of agreement was 75 per cent, which can be considered as satisfactory to good.

Operationalisation of self-presentation

At the beginning of the assessment interviews all clients were presented with the following, identical questions: ‘How would you describe your character? How would an imaginary good friend do this, and how would an imaginary enemy?’ The answers to these questions provided the data for this analysis of self-presentation. The relatively short text fragments enclosing the answer to these questions are always clearly identified in the records with the heading ‘Self-image’ and contain mostly a succinct answer of several sentences, up to half a page of text. In contrast to the rest of the report, the answer was recorded as a literal quotation remarkably often. Despite being co-constructions, these text fragments could be seen roughly as a selective yet fairly good representation of what clients answered.

When answering these questions, clients almost always described their own character in more than one or two qualities – sometimes five or six. These qualities of character have been organised into 22 main categories, which are shown in Table 1.

General image

The passages about self-image were fairly short, but the average length decreased over the course of time, from 160 to 110 words. Male analysts provided shorter material than their female colleagues (average 112 words compared to 176), especially when describing male clients (F = 4.656 with p ≤ 0.01): men produce conciseness with each other, a fact that is also known from medical consultations (Meeuwesen et al. 1991). For male and female clients, largely the same character qualities were reported, with depressed and anxious forming a stable core and being mentioned most. Given the selection of the documents, this of course could be expected, but it also corresponds with the fact that anxiety and depression are often co-morbidities (Moffitt et al. 2007; Sanderson, Beck & Beck 1990).

The other traits were, in order of frequency: aggressive, social, talented, closed, professional, virtuous and oversensitive. Smaller gender differences were found though: women mentioned more traits and over time changed more in their self-portrayal than men. The oversensitivity traditionally ascribed to women dropped visibly in the course of time (Chi² =13.691, p ≤ 0.01; Cramèr’s V = 0.48), while women’s newly acquired possibility of professional qualities has come to the fore more strongly since the 1970s (Chi² = 12.411, p ≤ 0,05; Cramèr’s V = 0.50). Men changed less and had a more diplomatic and restricted self-presentation: they almost always combined favourable and unfavourable qualities, and mentioned fewer of them. In the context of a psychoanalytic assessment interview, women appeared to be able to list more aspects of themselves, and in subsequent decades showed themselves from a more strongly positioned positive side than men.

Typology of self-presentation strategies

The way these clients struggling with gloominess presented themselves during assessment could be described in a typology. Three different and mutually exclusive types of self-presentation strategy became apparent that cover all character descriptions: a self-blaming, a self-relativising and a self-respecting strategy. In a self-blaming strategy clients – at least as represented by the analysts – gave a completely negative picture of their own character. Even from the perspective of a good friend, they were unable to bring any positive qualities to the fore. The two other strategies contained positive as well as negative qualities, but in the self-relativising strategy this self-image was presented as a highly pliable one, depending on either the social situation or the mental condition the clients were in. It happens that because of their depression they were confronted with changes in their self-image and feelings about self, and said they no longer know exactly who they ‘really’ were. In the self-respecting strategy, by contrast, clients described themselves as having a stable character in which both favourable and unfavourable qualities had a place.

The three strategies have been present in every decade, but were not distributed evenly over time. The self-blaming strategy was present mainly in the 1950s and then dropped down and stayed low. The self-relativising strategy gradually increased until the 1980s, before dropping sharply. The self-respecting strategy was the most common one and remained constant for a long time, but its use strongly increased in the 1990s. The trend of the overall changes was nearly significant (p = 0.09), but weak (Cramèr’s V =0.25).

Figure 1. Self-presentation strategies over a fifty-year period (N = 108)

Chi² = 13.871; p = 0.09; Cramèr’s V = 0.25

In these types, the gender ratios were divergent: the self-blaming strategy occurred more in women, the self-relativising about the same in both sexes, and the self-respecting was used slightly more by men. Nevertheless, these differences were not significant (Chi² = 2.432; p = 0.30; Cramèr’s V = 0.15). If we look at the changes by time period for each sex separately, the differences are noticeable. The changes women showed in their self-presentation strategies are significant and moderately strong (Chi² = 19.389 with p = 0.01; Cramèr’s V = 0.42). They left behind the self-blaming strategy and came to present themselves gradually in a self-respecting way. Men did not vary that much in the way they showed themselves in this context (Chi² = 3.696, p = 0.88; Cramèr’s V = 0.19); they almost never displayed themselves as wholly negative and always preferred a more self-appreciating posture.

Figure 2. Self-presentation strategies: Females over a fifty-year period (N = 55)

Chi² = 19.389; p = 0.01; Cramèr's V = 0.42

Figure 3. Self-presentation strategies: Males over a fifty-year period (N = 53)

Chi² = 3.696; p = 0.88; Cramèr's V = 0.19

The self-blaming strategy

People visiting a psychoanalysis clinic for help may have problems with those around them or with tasks they have to accomplish, but they mainly experience problems with themselves. For different problems they would have been referred to other professionals at an earlier stage already (De Swaan 1990b). This also appeared in the self-descriptions. No one gave a wholly positive picture of their own character, but in 21 per cent of the records an entirely negative image was given (N = 23; 15 women and 8 men). Such self-presentations were classified as self-blaming. For example:

This was primarily a strategy of women in the 1950s; afterwards, and among men in general, it became much less common. These people presented themselves as modest, if not inferior, but not as victims. In fact, they took all the blame upon themselves: it all was because of their grumpy, oversensitive, forced, egoistic, gloomy, doubting and introverted character. In the literature this is called a protective strategy, and is used mainly by persons with a low self-image. They do this out of fear that later in the contact with their therapist they will fall short (Friedlander & Schwartz,1985; Schütz 1988). It may be possible to qualify this attitude as a strategy prompted by shame (Scheff 2000; 2001). People who speak about themselves in this way feel inferior in their core, because they fear being rejected by someone whose judgment is important – in this study the analyst they are talking with and who is assessing them. For the most part, this shame was disavowed and unacknowledged, but manifested itself in things like stammering, a defeated look, blushing, sweating, feeling paralysed. Analysts described the postures of these particular clients elsewhere in the report as follows:

Or:

This is typical of a situation in which people feel handed over to the presumed evil glance of important others: shame is social pain (Goudsblom 2007). In the 1950s, women risked social disapproval when they asked for attention for their incapacity to meet the normal expectations of cheerful servitude to parents, husband and children. Their exhaustion, apathy, disappointment and crying fits were held against them by their parents and husbands, as well as by themselves. The emancipation movement of women, patients and other less privileged groups had yet to come. To make things worse, at that time in the Netherlands there was a major taboo about getting help for psychological problems. Very few people did that, one had less personal knowledge of who to go to with such problems and seldom learned about these things from others. Depression was a frightening diagnosis at the time, one that people did not face easily (Pieters et al. 2002). Psychoanalysts, for their part, did not go out of their way to make clients feel at ease during those assessment interviews, reflecting the common, more hierarchic doctor-patient relationship of those times. In such a situation it is not surprising that it is mainly women that were so ashamed of asking for help and embarrassed by the intimate subjects that came up; who were so afraid of a disapproving judgment, of being seen as a fake or ungrateful, that, out of precaution, they described their own character exclusively in negative terms.

The description of one’s own character tended not to be as one-sidedly negative as in the previous strategy. In the majority of the records studied (79 per cent; N = 85), a more nuanced or balanced picture comes to the fore. Aside from the self-blaming strategy, two other types were distinguished: the self-relativising (30 per cent) and the self-respecting strategy (49 per cent). The difference lies in the static or non-static picture that is given of one’s own character.

The self-relativising strategy

In the case of a changeable picture, this was typified as a self-relativising strategy: people put into perspective the impression they make on others, indicating that it is just appearances, can be strongly influenced, or does not correspond with how they ‘used’ to be. For example:

The self-relativising strategy was described in 30 per cent of the records, nearly as often by male as by female clients (N = 33; 16 men and 17 women). Those who presented themselves in this way indicated that they wrestled with questions about their character, and displayed an almost Goffmanian realisation of the social suggestibility of what they could show to others. This could contain a warning to the therapists who were examining them: ‘Watch out, this is just a momentary or circumstantial take’. At the same time, they gave a layered self-image: under the surface still hides a more vulnerable self that doesn’t reveal itself so easily or is entirely covered by the depressive self. This strategy, however, mostly showed the capacity of clients to think about themselves in the interview situation in a more reflexive way, without falling into moral disapproval as is the case with the self-blaming strategy. These clients were able to put themselves into perspective – a strategy not found in the literature.

It is not a coincidence that this strategy comes up mainly in the 1970s and 1980s, when a broad emancipation process had taken place that empowered, among others, women and mental patients, enabling them to present themselves in subtler tones (Wouters 2007). The attitude of psychoanalysts in the assessment interviews became more empathic (Gomperts 2002); the taboo against going to a psychotherapist waned under the influence of anti-psychiatry (Pieters et al. 2002), and the financial threshold was drastically lowered thanks to collective health care insurance (Brinkgreve 1984; Netherlands Institute for Social Research 1998). Clients did not have to feel as ashamed as before.

The self-respecting strategy

In this strategy self-images of a more constant identity were described, and next to problematic character traits, there was also space for positive ones. This was the most commonly used strategy (49 per cent, N = 52; 24 V and 28 M), which for a long time remained at a constant level, but greatly increased in the 1990s, especially among women. Men had always presented themselves more in this way and over time of the years changed less than women in this respect. For instance:

In the literature such a strategy is described as ingratiating – ‘trying to look good by presenting a favourable image’ (Schütz 1988: 614-615; Friedlander & Schwartz 1985). It is an active, yet non-aggressive attempt, to make a positive impression. In this way someone wants to be known as a motivated and ‘nice’ client. Clients strive for agreement with the therapist, for a situation of mutual attunement in which there is understanding of the patient’s inner world. These people do not manifest themselves as totally desperate and needy. They present a differentiated self-image that provides them with the necessary self-respect with which they can face the less glorious sides.

This strategy was the most used in the 1990s, in this decade even slightly more by women than by men. In addition to all the changes from the 1970s and 1980s which allowed for a more balanced self-presentation, in the 1990s other developments played a role. The power position of clients became considerably stronger, as evidenced by the legal right to inspect one’s records that came into effect in 1995. The social distance between client and therapist diminished: they were to an increasing degree part of the same social and even professional circles. This pressed psychoanalysts into a more cautious style of reporting, without any display of class superiority. At least as important is that the way people thought about depression underwent a change. From 1990 onwards, the lay public became familiar with modern antidepressants (SSRIs). The image of depression changed partly because of these pills, from an exceptionally severe condition bordering on madness that no one is allowed to know about, to a biochemical imbalance in the brain, treatable with medication. At the same time, the public image went from being something that happens to you (destiny, God’s will, bad luck) to a condition you are responsible for. This means that you yourself must make sure you get rid of it by living a healthy lifestyle, taking antidepressants, following training and going to therapy (Dehue 2008). Although this was (and still is) an ambivalent, partly emancipatory and partly oppressive discourse, it does imply that asking for help for depressive problems was experienced as less shameful. This applied to men as well as to women. Still, each sex faced a slightly different task. During the 1990s, women, based on their increased educational level, labour participation and income, could avail themselves of more social, cultural and financial capital than before (Netherlands Institute for Social Research 1998). When they were assessed by the NPI in the 1990s, they felt like capable professionals who could nonetheless use some help for troubles with depression. For men, their identity during the entire assessment period stayed centred in a self-evident way around their education and work, and this was expressed in the self-presentations, though accordingly less emphasised. Their gender script has changed less than that of women. When they came to the NPI for help, men remained – at least in the representation of the psychoanalysts – more reserved in their self-presentation, and in this respect little changed in the examined period.

Dr Jette Westerbeek

Faculty of Social and Behavioural Sciences, Utrecht University, The Netherlands

Jette Westerbeek is assistant professor at Utrecht University, Interdisciplinary Social Science, The Netherlands. She finished her thesis on the social construction of depression and publishes on qualitative research in the domain of mental health and gender.

Dr Ludwien Meeuwesen

Faculty of Social and Behavioural Sciences, Utrecht University, The Netherlands

Ludwien Meeuwesen is a psychologist and associate professor at Utrecht University, Interdisciplinary Social Science. She has published on communication in health care, intercultural communication, interpreter issues in health and social care, women studies, social support, social networks and sustainability.

Professor Dr Christien Brinkgreve

Faculty of Social and Behavioural Sciences, Utrecht University, The Netherlands

Christien Brinkgreve is professor at the University of Utrecht in the interdisciplinary study of family relations. She has published many books and articles about the developments in behaviour and emotions; her dissertation was about the development of psychoanalysis in The Netherlands.

Dr Wouter Gomperts

Faculty of Social and Behavioural Sciences, University of Amsterdam, The Netherlands; Netherlands Psychoanalytic Institute, Amsterdam, The Netherlands

Wouter Gomperts is assistant professor at the Department of Clinical Psychology at the University of Amsterdam and a psychoanalyst at the Netherlands Institute of Psychoanalysis.