Introduction

All families have their secrets (Knauth, 2003; Tracy, 2015). As a normative phenomenon, secrets do not automatically refer to something pathological. Keeping a secret might be indicative of a collective denial that manifests itself in the family as functional. In the family therapy tradition, family secrets refer to topics charged with intense fear, shame and guilt. If the secret becomes taboo, inhibiting dialogue and distorting the adaptability and development of the family system, it becomes problematic (Simon et al., 1985), affecting the dynamics of the family unit as an emotional and relationship system (Bowen, 1978; Vangelisti and Caughlin, 1997) and challenging the task of family therapists (Deslypere and Rober, 2018). In family secrets, the information that is withheld is considered to be critical to the ones from whom the information is concealed, because it has an effect on his or her life (Berger and Paul, 2008; Vangelisti and Caughlin, 1997). Qualitative research is needed to increase an understanding of the complexity of the phenomenon of a family secret and its systemic and multi-directional effects on family members. In this study, our interest was in how children position themselves in relation to the topics kept secret and how they cope in these demanding situations.

Secrets define boundaries telling us who is in and who is out (Imber-Black, 1993). From a systemic perspective, secrets affect all the participants involved in the therapy process. Secrets lead to collusion, psychological distancing, reduced trust, compromised communication and dissatisfaction and to unbalanced family loyalties (Dreman, 1977; Imber-Black, 1998; Vangelisti and Caughlin, 1997). The family as an emotional and relationship unit functions in ways that reflect each family member’s thoughts, feelings and behaviour. As all parts of the system are interconnected, no individual functions in a vacuum; instead, each individual responds to the other individuals and contributes to the integrity of the system (Bowen, 1978; Kerr and Bowen, 1988). Secrets in families can become multigenerational phenomena, transferred as rules of communication, delegations or legacies, which can carry and mediate complicated loyalty bond structures. Some stories can, for example, run in families as forbidden topics, or a family member can be determined to fulfil some predetermined duty or task (Stierlin, 1977a, b).

On the individual level, secret-holders experience tension, loneliness and stress-related physical health problems (Kelly, 2002). Maintaining secrecy binds psychic energy, causing holders confusion and anxiety, and affects communication within the family, leading eventually to family dysfunction (Imber-Black, 1998; Karpel, 1980; Vangelisti and Caughlin, 1997). Family secrets may hinder the natural growth of a child’s individuation process. Even secrets kept with the best intentions (protection) can negatively affect a family’s interactional patterns (Bowen, 1978; Imber-Black, 1998; Stierlin, 1977b). Those kept unaware of a secret try to deal with distorted communication practices, and may develop self-doubt, suspicion, fear and anxiety, eating disorders, and negative psychological functioning later in life (Imber-Black, 1998). The typical mechanism used to maintain secrets is topic avoidance. Berger and Paul (2008) showed that there is an inverse relationship between topic avoidance and family functioning. They found that, especially among mothers, general topic avoidance was the strongest predictor of family functioning, whereas parental joint disclosure predicted the highest level of functioning. Three distinct motivation categories relating to topic avoidance have been identified: relationship-based, individual-based and information-based (Afifi and Guerrero, 2000; Berger and Paul, 2008; Golish and Caughlin, 2002). The first refers to the need to maintain a close relationship and protect it from, for example, conflict and anger; the second focuses on self-protection; and the third is motivated by the desire to convey information in a clear and relevant way.

Family secrets include a wide range of topics in family life. Negative past experiences, adoption and infertility (Berger and Paul, 2008), alcoholism, extramarital affairs, and traumas such as suicide, physical and mental illness and death are typically veiled in secrecy (Imber-Black, 1993). Protecting children from sensitive and ‘toxic’ secrets (Imber-Black, 1998), for example in cases of violence taking place inside the family, is understandable. However, it is known that children, as the barometers of the family climate, are especially vulnerable when faced with an aura of secrecy as their self-regulation skills are still evolving. Children also differ in their reactions. Internalising behaviour may manifest as depressive symptoms and externalising behaviour as problem behaviour (Bowen, 1978). Dreman (1977) and Baird (1974) found that a child may become the scapegoat and symptom bearer of a secretive family communication system in which the secretive communication is intertwined with an aggression problem resulting from an inability to deal effectively with anger.

The concept of family secret focuses ‘one-sidedly’ on its negative effects and thus fails to capture the complex nature of secrecy. The concept of selective disclosure offers an alternative approach to this complexity, pointing to the dialectic tension between what is said and not said, between keeping the secret and sharing information (Rober et al., 2012). On the assumption that dialogue is a precondition for positive change in any form of therapy (Seikkula and Trimble, 2005), selective disclosure as a dialogical concept has earned its place in family therapy practice. The aim of a dialogical approach is not to induce or pressure open disclosure but rather to invite reflection on the meanings family members attribute to their hesitation and silences (Rober, 2002). From focusing only on the promotion of ‘openness’, this approach has shifted the focus towards highlighting the complexity of the dialectical tension between openness and closedness (Baxter, 2011). In the therapeutic conversation, clients are constantly selecting what to tell and what to keep silent about. Rather than focusing on the content of the unspoken story, the therapist should invite family members to talk about the good reasons behind their decision.

Some stories that might be relevant in the therapeutic dialogue are too difficult to tell (Rober, 2002). The decision to tell a sensitive story needs to be weighed against the emotional impact it may have on vulnerable family members (Rober and Rosenblatt, 2013.) Some stories remain untold because the context of the conversation is judged to be unsafe (Rober, 2002). The client’s silences and hesitations are important information to a therapist and become a therapist’s main tools to work within systemic therapy. It is also important to keep in mind that secrets in families are not necessarily toxic; sometimes they serve to create a story that family members can live with (Rober and Rosenblatt, 2013).

The therapist’s task is to listen to the client’s stories and help to open up a space for the not-yet-said (Anderson and Goolishian, 1988). In the case of family secrets, the task is demanding, given that secrets evoke powerlessness, uncertainty, and even anger. Moving too fast often results in clients closing up and recanting their story or breaking off the therapy (Deslypere and Rober, 2018). A genuinely respectful dialogical approach creates a context in which clients feel that it is safe to tell their sensitive stories (Rober, 2002). This calls for therapists to tolerate uncertainty in a way that can help provide the safety that enables family members also to tolerate uncertainty (Seikkula and Olson, 2003). Tolerating situations in which no ready-made responses exist and taking a not-knowing stance challenges the therapist’s role as an expert (Anderson, 1997). In a state of not-knowing, therapists stay in touch with the complexity, uncertainty and unfinalisability of the situation and thus expose themselves to a multiplicity of voices in their inner conversations (Rober, 2002).

Language (spoken and unspoken communication) acquires its meanings through careful attention to how it is uttered. Aristotle in Peri Hermeneias (De Interpretatione and Categories, 1975) formulated his idea that outer and inner words are not identical, stating that every sentence is only an interpretation of one’s thought. In practice, to understand ‘you’, it is not enough to understand ‘your’ words. It is also crucial to grasp meaning, thought and motivation (Vygotsky, 1971, p. 151). The only way to do this is to listen to what the other has to say. Harlene Anderson (1988, 2001, 2012) described family therapy as a meaning-generating system where people participate in an ‘in-there-together process’. Meanings are generated in an inter-relational context, through the fluid process of give and take, which by its nature is dialogic. In dialogue, meanings and understandings are jointly constructed. The listener’s active presence is what distinguishes dialogue from monologue (Bakhtin, 1986). In dialogue, every utterance needs to be answered. Answering does not mean giving an explanation or interpretation, but rather demonstrating in one’s response that one has taken note of what has been said. Hearing is always demonstrated in our answering words (Seikkula et al., 2012). According to Bakhtin, ‘For the word there is nothing more terrible than a lack of response’ (Bakhtin, 1975, p. 127). Although a key principle in family therapy is that children’s perspectives are heard (Strickland-Clark et al., 2000), it is obvious that sessions are typically constructed by adult-led talk and conversation. To hear children’s voices means engaging them as full members of the therapeutic dialogue, as participants who have important things to say. The process of engaging children has been found to be challenging. Willis, Walters and Crane (2014) showed that typically children were passive participants and excluded from much of the therapy dialogue. Hutcby and O’Reilly (2010) and Parker and O’Reilly (2012) found that children tend to occupy an unequal position, described as ‘half-membership status’, in adult interactions. Half-membership status refers, for instance, to the position of the child as the talked-about other (Parker and O’Reilly, 2012) and as being interrupted (O’Reilly, 2008). Positioning refers to the question ‘from where is the person speaking?’ (Hermans, 2006; Seikkula et al., 2012).

The aims of the study

The objective was to study how children participate and position themselves in episodes concerning secretive topics in family therapy sessions and how they cope in these situations. We also investigated how therapists and parents responded to children’s initiatives in talking about sensitive or forbidden topics. Qualitative research on family secrets in family therapy is scarce. This small-scale study contributes to answering this need.

Data

The research data consisted of video-taped family therapy sessions held at Kuopio University Hospital Child Psychiatry Clinic. The research material forms part of a larger family therapy research project on the fourteen families of children aged 6 to 12 years diagnosed with oppositional defiant or conduct disorder. All participants gave their informed consent to take part in the study and the research plan was approved by ethical committee of Northern-Savo Health Care District. One of the therapeutic processes was selected for further study owing to its distinctive feature of family secrets concerning multigenerational traumatic losses. This family therapy process comprised fifteen sessions, each varying in duration from 55 minutes to 1 hour 47 minutes, conducted over a one-year period. For a closer study, the first author selected three distinct types of family therapy session: (1) an at-home implemented genogram workshop (4th), duration 1 hr 37 mins; (2) a network meeting at the child psychiatry clinic (11th), duration 1 hr 43 mins; and (3) an at-home implemented session (13th), duration 60 mins.

The family consisted of (pseudonyms) Jane (mother), Brian (father) and 9-year-old Mark and his younger sister, 8-year-old Clara. They are identified in the excerpts by the abbreviations J, B, M, C. The sessions were conducted by two family therapists, T1 and T2. The therapeutic approach was systemic with elements of structured games and interactive tasks.

Methods and procedure

This study applied the multi-actor Dialogical Methods for Investigations of Happening of Change (DIHC) (Seikkula et al., 2012). Before the analysis, three videotaped sessions dealing with the theme of secrecy in the family were transcribed in full by the first author. Non-verbal information was also taken into account. The accuracy of transcription was planned to meet the needs of DIHC with an emphasis on the verbal content, without prosody. The analysis was made in Finnish, the participants’ native language, in order to capture all the nuances of speech. The translation process into English was done by the native English speaker, who has lived in Finland for a long time. The meanings of translations were, however, negotiated together with the first author. The analysis was carried out by the first author and the second and third authors acted as supervisors, and as the auditors of the analysis. After the authors’ careful reading, the research proceeded in the following steps. (1) Episodes defined as topical were explored. A change of topic was considered a new episode. The episodes concerning family secrets were chosen for microanalysis. (2) The responses to each utterance were noted to gain a picture of how each interlocutor participated in the construction of the joint conversation. In this study, the concepts used to analyse response categories were semantic dominance, referring to who introduces new themes or new words at a certain moment in the conversation, and interactional dominance, referring to the dominant influence of one participant over the communicative interaction. (3) In this step, the narrative process coding system was followed (Angus et al., 1999; Laitila, 2016; Laitila et al., 2001). The analytical tools used were concepts such as external process mode, referring to descriptions of things that have happened, internal process mode, referring to participants’ descriptions of their own experiences of the events they describe, and reflexive process mode, referring to participants’ efforts to understand the connection between the events in question and their personal experiences. (4) After analysis of the response categories, the focus shifted to the interlocutors’ voices, addressees and positioning. Voices refers to the speaking consciousness (Bakhtin, 1984) that becomes visible in exchanges between interlocutors in the context of the storytelling currently taking place. Positioning links a voice with a participant’s point of view. Addressees are the persons to whom an utterance is addressed. In analysing multi-actor dialogues, addressees are not always easy to identify. Speech can be also addressed to someone in the inner dialogue (Seikkula et al., 2012). The analysis and results were discussed and reflected on together by the authors and relevant literature was consulted, including research on family therapy. The results are presented partly in narrative form, following the chronology of therapy sessions (Braun and Clarke, 2006).

Analysis and results

The results of the analysis presented in this paper focus on two topical episodes concerning the family’s secrets, one relating to the past and the other to the present. The results concerning the secret of an uncle’s suicide (past) is presented first, but only in analytic narrative form. The second analysis concerns the mother’s health (present) and is presented in detail and in full in Table 1. The transcriptions in the tables are presented according to the following principle: first the original Finnish data is presented, then follows the English translation in italics and in parentheses.

Case history

Mark’s family had been referred to the child psychiatric clinic due to Mark’s aggressive behaviour and he had been diagnosed with a conduct disorder. He had spoken of having thoughts of suicide and this also occurred in the process of therapy. Mark’s younger sister Clara suffered from internalising symptoms, was problematically dependent on her mother and had fears and sleeping problems. In recent years, the family had experienced multigenerational traumatic losses (the suicides of the children's uncle and grandmother) that had remained unspoken due to their sensitive nature.

The secret of the uncle’s suicide

In the fourth session, the therapists suggested to the family that they attend a genogram workshop in order to study the family histories of both the parents over the period of three generations. This proved effective in getting the children to examine their complex family patterns, relational resources, significant events, and losses. The genogram offered them the possibility to approach hidden, unspoken themes. Both children positioned themselves as active on the topic of their uncle’s death. Clara took the initiative by informing the therapists that her mother’s brother had died. Mark, who posed several questions, wanted to know how it had happened. The therapists’ role was to balance the needs of the children and those of their mother. Using non-verbal body language (gestures), the mother indicated the difficulty she had in talking about the topic and answering Mark’s questions. T2 assumed the role of negotiator. She tried to encourage the mother to disclose something, however small. The mother’s reply was ambiguous, simultaneously opening and closing the topic. It was open in that she stated that the theme was a difficult one but closing in that she stated that answering ‘would have serious consequences’. The mother’s good reason for remaining secretive can be viewed understandably as protective; however, from a dialogical perspective it tied the hands of the therapists, categorised the topic as dangerous, as taboo, and thus reconstructed the secretive atmosphere around it.

The secretive atmosphere surrounding the mother’s wellbeing

The thirteenth session started in the family’s kitchen in an aura of secrecy. Mark and Clara were lying at the fireside. As a result of therapists routinely asking family members to complete in-session feedback forms at the beginning of the session, with the aim of tracking and focusing the intervention, T1 had noticed that the mother’s self- evaluated wellbeing scores were exceptionally low. As is usual in therapeutic conversational contexts where multi-actors are present, several themes were competing for selection and attention. These included Clara’s question to T1 and T2 about when the family could visit the child psychiatric clinic again, Mark’s defiance about attending school that day and the alarming observation concerning the deterioration in the mother’s wellbeing. The therapists decided to focus on the last of these. T1’s ‘let’s listen to mother’ was the starting point for the microanalysis of the topical episode.

Mark’s and Clara’s self-positioning

Mark and Clara reacted differently to the secretive atmosphere. Mark positioned himself in accordance with his diagnosis, as the following excerpt illustrates. The mother had just said that ‘it has been a difficult week and troubled times’ and the therapists were interested to learn more about those things.

Mark’s ‘I don’t need any health’ is significant in the conversational context in which the therapist’s ‘difficult week, in what sense?’ had just invited Mark’s mother to explain her response. Mark’s intervention can be interpreted as a rescue operation. Mark shifts attention, even negatively, to himself and away from the sensitive issue of his mother’s health. To protect his mother from having to talk about a sensitive issue, he assumes the role of a defiant child, one that he and his family are used to. By acting in this way, Mark reconstructs both the secretive atmosphere and his role as a defiant child. His utterance can also be understood from the perspective of his inner voice as challenging the multigenerational delegation. Were the real addressees his mother’s no longer present multigenerational relatives? What Mark was really saying was not taken up.

The mother had had a sudden seizure at home just a few days ago. Clara assumed an active and initiating role as a key informant concerning her mother’s seizure. Clara had witnessed this frightening situation and at her mother’s request had obtained help from her father. Clara’s positioning in the conversation was ambivalent. She asked her mother for permission to tell what she knew. In telling her story, Clara observed her mother’s reactions and sought to balance between her need to tell and her loyalty to her mother’s reluctance to embark on the topic. The voices in Clara’s storytelling can be interpreted as contradictory in both what she said (content) and how she said it (form), as in the following excerpt:

Clara’s initiative can be interpreted as multidimensional. She shows courage in broaching a sensitive theme but simultaneously fear of rupturing the multigenerational legacy of loyalty structures. While it remains unclear how permitted it has been in general in this family’s history for its members to talk about difficult themes and negative emotions, it is evident that for Clara it has been difficult.

The mother

The mother’s seizure had occurred a few days before the session took place. In discussing the theme, the mother positioned herself as unsure what to say in the presence of the children. When positioned by the therapists to give an account of what she meant by a ‘difficult week’ her response ‘family, work, health’ seemed to offer big themes for discussion. However, the words both opened and closed off any potential discussion. The distancing words, addressed to the therapists, indicated her reluctance to talk about it anymore. Simultaneously, echoes of loyalty to her multigenerational relatives (speaking about difficult topics around the kids is forbidden) can be heard in her inner voices. The therapists nevertheless tried to make more room for the mother’s suggestive and secretive topics and encouraged her to talk, as illustrated below:

The mother’s good reason for being taciturn was protecting her children as representative of the family past and present. Her hesitant and allusive response ‘otherwise they will have more…’ refers to her fear and difficulty ‘to tell the truth’ which she had talked about earlier in her private discussion with T1 at the clinic. In that discussion she made clear that she was not yet ready to tell the facts of her relatives’ deaths to the children because the suicides had provoked such a strong outburst of rage and guilt in her. T1 had encouraged the mother to talk about the deaths with the children in an age-appropriate manner, suggesting that unspoken themes can cause invisible anxiety. The mother admitted that this had been the case in her family. The mother’s health was also a sensitive issue, as the mother had also told T1 that Clara had spoken of being afraid of losing her mother and asking every now and then in the mornings ‘are you going to die today?’ Despite the mother’s good intentions here, her suggestive words made room for further imaginary fears and interpretations, and thus reconstructed an unsafe climate.

T1 and T2

The secretive atmosphere, with its ambivalent and contradictory voices, was inimical to the therapists’ task of opening up a space for the not-yet spoken. The therapists positioned themselves as listening and not knowing. They encouraged the mother to generate local meanings (Anderson and Goolishian, 1992) in order to construct an understanding of her response of ‘family, work, health’. They created a space for dialogue between the mother and Clara and tried to stabilise the unclear and emotionally demanding situation. However, the secretive atmosphere also aroused voices of ambivalence in the therapists, voices of confusion and hesitation in the competing dialectics of whether to talk or not to talk. The mother’s suggestive words were effective: at the point where the mother later appealed implicitly to the children to leave her in charge of her own health with the words ‘mother is not allowed to get upset’, T1 shifted the focus of the conversation to the arrangement of a next meeting, where the children would not be present.

Discussion

This study focused on how family secrets as a systemic phenomenon affect children’s positioning in the family therapy and how they cope in these challenging situations. In the present case we noticed, first, that both children were active co-participants in the complex dynamics of the secretive atmosphere in the family. Second, they involved themselves in the paradoxical processes of reconstructing and deconstructing this secretive atmosphere. The children participated actively in the topics concerning the family’s secrets. Although children’s self-positioning in family therapy is typically passive (Willis et al., 2014), the present results show that children may also engage actively in discussions dealing with sensitive and concealed issues. Both children took initiating roles in their approaches to a sensitive topic. They asked relevant questions and acted as informants.

Paradoxically, and simultaneously, in their ways of deconstructing the secretive atmosphere they also positioned themselves as reconstructing the secretive atmosphere. Mark’s symptomatic behaviour, manifested in his speech about committing suicide, offered the opportunity for forbidden themes to be discussed. At the same time, however, he paradoxically kept the attention on himself, thereby implicitly protecting the sensitive topics from becoming a therapeutically relevant topic of discussion. Mark’s threats to kill himself kept the suicide secret present, while simultaneously his provocative behaviour, his infantile protest, kept the focus on him instead of on the secret. In this context, deconstruction refers to Derrida’s idea that every utterance simultaneously contains contradictory aspects and escapes absolute determination; in other words the ‘meaning’ of a ‘thing’ comes into existence through and in relation to what the ‘thing’ is not (Derrida et al., 2003).

Mark’s and Clara’s coping mechanisms in the family’s emotional and relationship system showed differences. Whereas Mark’s way of coping was to react externally, Clara, who was problematically attached to her mother, assumed the role of an emotional regulator after she had risked putting her mother in touch with her own vulnerability. Clara had witnessed and even assisted her mother in the chaotic situation surrounding the latter’s seizure, which positioned her as having semantic dominance in that conversation topic. However, she found herself in an ambivalent position: on the one hand she wanted to talk, to tell what she knew, while on the other hand she sought to protect her mother from this difficult theme. Clara’s insecurity was masked by her cheerful appearance, which was inconsistent with her story, indicating the presence of at least two distinct voices. In the analysis of storytelling, it is important to note if ‘there is congruence between the story told and story lived’ (Rober et al., 2010, p. 36).

The present findings support previous reports on the negative impact of secrets on family communication, as discussed in the introduction. First, we noticed that a secretive communicative style produced a tense and psychologically distancing climate, producing voices of ambivalence, hesitation, and confusion. The concept of selective disclosure (Rober et al., 2012) enabled a deeper understanding of the mother’s good reasons for her reluctance to talk. Her reasons were intended to protect not only her children and her deceased multigenerational relatives (relationship-based) but also herself (individual-based). Taking the mother’s own words seriously, her personal grieving process over her mother’s and brother’s suicides had been blocked by feelings of anger that had kept her a prisoner of aggression for several years. The mother possibly saw Mark’s suicidal speeches as potentially dangerous and as a self-fulfilling prophecy that triggered intense fear in her. In line with the findings of Baird (1974) and Dreman (1977), the mother’s mishandled and uncompleted grieving process and anger might have led to secretiveness. The mother’s suggestive utterance ‘if the topic is discussed, the consequences will be harmful’ indicates that joint discussion of the secret would be dangerous. According to Imber-Black (1998), dangerous secrets poison relationships, creating barriers and reducing trust. Utterances intended as protection paradoxically have the opposite effect, increasing the emotional demands of the situation and the insecurity of the dialogical climate. A suggestive communication style tends to make room for imaginaries and children’s fantasies are often worse than reality (Fine, 1973). In the present instance, suggestive communication succeeded in influencing the emotional climate of the therapeutic system, leading to dysfunction, manifested by the exclusion of the children from the therapeutic discussion on the sensitive topic.

Secrecy had an impact on the therapists’ decisions. First, the mother’s decisions ultimately determined what could be talked about in the presence of the children. Second, the therapists, who were to become shareholders in the secrets, found their hands tied. They used their mandate in attempting to persuade the mother to say at least something to the children. It can be asked, what more could they have done without losing the mother’s confidence? Their task of balancing the needs of the children to talk about sensitive topics and taking the mother’s words seriously was challenging. In this case, the therapists saw Mark’s visible aggression problem as in some way connected to the invisible constructions of family secrets.

Utilising the genogram, they promoted discussion around past losses. In their attempts at negotiating they vainly endeavoured to motivate the mother to talk about painful issues that would have promoted the shared grieving process. The mother’s therapeutic goal was to get help for Mark’s aggression problem rather than to talk about past losses. There is no royal road to knowing for certain whether Mark’s suicidal talk and aggressive behaviour was connected to the hidden themes of his relatives’ suicidal deaths. However, it has been noticed that a blocked grieving process (Bowen, 1978), secretive communication and mishandled anger (Baird, 1974; Dreman, 1977; Fine, 1973) may unwittingly scapegoat the child.

In this family the mother found that the family secrets concerning the relatives’ suicides were topics that were too threatening to be jointly discussed and shared. However, her decision to refuse to talk about the relatives’ suicides with children was her conscious, and articulated choice. The family members effectively kept the attention on their visible symptoms, preventing invisible and sensitive topics being effectively and explicitly brought into therapeutic focus. Knowing that keeping secret binds psychic energy, causing stress, loneliness and tension, it was not surprising that the mother’s seizure appeared to have been related to stress-related symptoms, symptoms indicative of a keeper of secrets (Kelly, 2002).

Mark’s defiant behaviour can be interpreted as a ‘cover story’ concerning his vulnerability. One can only guess at the role Mark’s defiance plays in his family’s multigenerational pattern of facing difficult feelings, such as anger. Mark had told the therapists about his need to receive more attention from his parents and had manifested implicit irritation with his mother. Mark’s utterances ‘Mum doesn’t know me’ and ‘I don’t need any health’ can also be interpreted as voicing isolation and loneliness. However, provocative utterances by an individual positioned as defiant typically make hearing a demanding task. Mark’s utterances were interpreted by the adults in accordance with his symptomatic behaviour. An interesting question remains: what role did the father’s absence play in the sessions where the family’s secrets were offered for joint discussion?

Conclusions

The findings have clinical implications. Granting that family patterns tend to repeat themselves (Bowen, 1978; Kerr and Bowen, 1988), we suggest that the topic of family secrets should be taken seriously in the family therapeutic context. It is recommended that family secrets are asked about in the pre-therapy assessment and diagnostic interviews where all the family members are present. At its best, the genogram as a therapeutic tool can enrich therapeutic processes, enabling open exploration of multigenerational family patterns and functions that might be influenced by family secrets (McGoldrick et al., 2008). By normalising the phenomenon of family secrets, therapists could make room for joint discussions on these and encourage family members to talk about their good reasons not to talk (Rober, 2002). According to Tracy (2015), ‘family secrets can be a driving force, whether explicitly or implicitly, for many seeking therapy’.

A limitation of this case study concerns the generalisability of its results. Because they remained hidden from the children, the effects of the family’s secrets on its functioning remain obscure. While conceding that the conclusions drawn in this study are tentative, as they tend to be in studies of this kind, we believe that the study enriches understanding of the multifaceted and systemic nature of family secrets and the self-positioning of children in them. Furthermore, this study offers new insight on the utilisation of the multi-actor DIHC method when children are present. Children’s conduct disorders in the context of family secrets merit further research. In child psychiatric care there might be many ‘cover stories’ behind such diagnoses. The meanings embedded in these stories cannot be approached and worked through without safe disclosure. Family therapy can be a forum to investigate them seriously and with respectful curiosity.