In this chapter, we explore the potential of discursive research for studying the interactional processes through which subjectivity is constituted, negotiated, and reformulated within and through the psychotherapeutic dialogue. We argue for the usefulness of the concept of subject positioning as a means for studying how subjectivity is talked into being in therapeutic conversations, while recognizing that psychotherapeutic conversations can provide rich material for social constructionist analyses of contemporary ideologies regarding selfhood and their intersections with discourses around gender, family values, parenthood, mental health, and so on.

Initially, we present our perspective on psychotherapy as discursive practice, outline the main ways in which subjectivity has been conceptualized in social constructionist accounts, introduce the concept of subject positioning, and provide a selective review of discursive research on psychotherapy with a focus on the study of subjectivity. Next, we illustrate the potential of discursive research for understanding how subjectivity is transformed within therapy through an example of analysis of part of a session from a couple therapy. Finally, we discuss the theoretical, clinical, and research implications of this body of work.

Social Constructionism and Psychotherapy

In line with constructionist principles, we approach psychotherapy as a discourse and a set of practices, which rely upon the so-called psy-complex, the set of professional concepts and practices that promote a psychological understanding of human life (Pulido-Martinez, 2014), that has become increasingly pertinent, over the course of the twentieth century, in constructing personhood and in regulating social and emotional life in Western cultures (e.g. Parker, 1994). Within this framework, psychotherapy is conceptualized as a primarily semantic process, that is, a process of meaning construction, that simultaneously relies upon and promotes psychological understandings of life and human subjectivity. The intersections and cross-fertilizations between social constructionism and psychotherapy have a relatively long history; starting from the articulation of psychotherapy approaches explicitly committed to the turn to language and post-structuralism, the concepts of narrative, dialogue, and discourse have become key metaphors in the majority of traditional psychotherapy schools, while the focus on meaning co-construction is increasingly recognized as a potential unifying paradigm in psychotherapy (e.g. Wahlström, 2006). Although not a unitary field, discursively informed approaches to psychotherapy share an assumption that psychological distress is associated with limited and limiting self-narratives, in the sense that aspects of lived experience remain poorly discursively articulated or articulated in ways that maintain problem-saturated subject positions. Furthermore, it is assumed that problem-saturated constructions often have an ideological dimension, in the sense that they are shaped by, and accordingly sustain, culturally preferred discourses. Correspondingly, discursively informed psychotherapy aims to expand or reconstruct problem-focused constructions of the client’s life and subjectivity and facilitate the discursive articulation of inclusive, polyphonic, and empowering self-narratives.

Despite recent developments in discursive theory and practice, discursive research occupies a relatively small place—in both numbers and scope—within the psychotherapy research literature. This could be attributed to several different factors, including theoretical/epistemological (the positivist bias of psychotherapy research), political/economical (selective funding for outcome studies, aligned with the “drug metaphor”) theoretical/methodological (challenges of post-structuralist research), as well as the fact that many social constructionist researchers assume a critical position towards psychotherapy as an institution.

In our earlier work, we reviewed discursive studies on psychotherapy and suggested that discursive research can contribute to our understanding of how the meaning of the problem and clients’ subjectivity are negotiated and transformed within psychotherapy talk; we also argued that discursive research provides rich, detailed, and critical analyses of the process of psychotherapy, thus potentially promoting clinically relevant research (Avdi & Georgaca, 2007, 2009). Since that time, discursive research on psychotherapy has grown, as a small but increasing number of researchers utilize discourse and conversation analysis to study their own and others’ practice. In this chapter, we do not aim to review this literature comprehensively, but selectively discuss work that focuses on the manifestation and reformulation of subjectivity in therapy talk.

Social Constructionist Accounts of Personhood and the Concept of Subjectivity

Mainstream psychological accounts of selfhood tend to represent the self as an identifiable, stable, internally consistent, and self-contained entity. These psychological accounts reflect the ideal of “self-contained individualism,” which arguably constitutes the dominant version of personhood in contemporary Western cultures (Sampson, 2003), and which approaches the “self” as an independent, undivided, unitary subject, a center of motivation and agency that is clearly demarcated from its social context. Over the last decades, these conceptualizations have been powerfully critiqued from several different perspectives, as naïve realist, essentialist, individualizing, culturally specific, and ethically problematic (e.g. Henriques, Hollway, Urwin, Venn, & Walkerdine, 1998). In this chapter, we use the term “subjectivity” to denote subjective experience and one’s sense of self, and we assume that this is constituted in and through language and social interaction. From this perspective, subjectivity is situated, contextualized, variable, and shaped by ideology and power dynamics, yet also affectively charged, private, and intimately personal.

Currently, there is not one, unified discursive theory of subjectivity, as discursive accounts draw upon different traditions, including Althusser’s interpellation theory, psychoanalysis, post-structuralism, and feminism (e.g. Törrönen, 2001), and this theoretical multiplicity is reflected in the different ways in which subjectivity is conceptualized and studied in discursive studies of psychotherapy. In terms of theory, a shared element of discursive approaches is the focus on language as constructive, functional, and variable; as such, discursive studies tend to examine the processes through which reality, agency, and accountability are created and negotiated within interaction. Consequently, subjectivity in discursive accounts is approached as jointly constructed in situated interactions and shaped by culturally available systems of meaning (e.g. Avdi & Georgaca, 2009).

A key concept for examining subjectivity in discursive approaches is subject positioning (Davies & Harré, 1990; Harré & van Langenhove, 1998). Subject positioning can be broadly described in terms of two interrelated, yet distinct, levels. On the one hand, it underscores the relational and interactional nature of subjectivity and refers to the position one assumes in specific interactions. When we speak, we assume a specific position and each story we tell is also—more or less explicitly—a story about who we are. We tell a different story of our troubles, for example, when we assume the position of a concerned parent, a troubled adolescent, or a mentally ill patient. Furthermore, each utterance is always addressed to someone, who we thus “call into” a particular position; the other can in turn accept, resist, challenge, or change this position invitation, and through his or her response, he or she in turn positions us (Drewery, 2005) in an ongoing process that constitutes the dynamics of the interaction. These exchanges are context-dependent and imbued with power, in the sense that in some contexts some participants have more say in positioning others, while some positions may be harder to resist. In psychotherapy, for example, the institutional position of “therapist” has more say in both defining meaning and in regulating the sequence of interaction than that of “client.” From this perspective, our habitual positionings—which can be distressing, painful, or problem-saturated—are maintained through our interactions with a “community of subjection” (Guilfoyle, 2014), that is, significant others in our life who, often unwittingly, participate in our interpellation in problematic subject positions, enforcing norms associated with dominant discourses. Furthermore, when taking one position, we automatically do not take another; this selective positioning is relevant to psychotherapy, as clients can be seen to repetitively assuming problematic subject positions, while disavowing aspects of their potential multiplicity.

In terms of analysis, we can examine positioning in therapy whenever a participant speaks, is addressed to or is spoken about, through questions such as: Who speaks? In whose name do they speak? Who do they address? Who do they speak for? Different positions entail differing degrees of accountability and can have a variety of functions in the interaction, such as attributing blame or refuting an unwanted identity (Georgaca & Avdi, 2011). In sum, from a perspective that focuses on the performative and functional aspects of language use, positioning is a key process through which selves are performed, jointly constructed—and potentially reconstructed—through language and within interaction. Psychotherapy, in this framework, works through creating a particular type of conversation within which the problematic or distressing subject positions clients occupy are explored, challenged, or expanded.

The second level of conceptualizing subject positioning focuses on the intersection between ideology, power/knowledge, and subjectivity; this level concerns the location of the person in discourse and within a moral order (Harré & van Langenhove, 1998) and the focus is on the ways in which speakers are positioned through particular discourses. This perspective draws primarily upon Foucault’s work and assumes that discourses entail an array of subject positions that people take up when they talk; these positions influence the course of interactions, the actions available to people, as well as their sense of self (Parker, 1994). Furthermore, these processes often take place irrespective of speakers’ intentions and outside of awareness. For example, when a psychotherapist meets a client, the way each participant interacts and experiences him- or herself is influenced not only by their individual biographies, but is also powerfully shaped by their respective, institutionally sanctioned positions of “therapist” and “client,” each of which has specific claims to knowledge and authority. As mentioned above, subject positions are closely implicated with power/knowledge; power here is approached as a constitutive force, closely associated with specific sets of knowledge that maintain their status through disqualifying alternative knowledge and naturalizing dominant discourses. As such, when deploying a particular discourse and its associated subject positions, we are implicated in an—often implicit—power struggle over the meanings that are seen as normal, good, and true. In terms of analysis, subject positioning on this level can be explored through investigating the discourses that are implicated in clients’ and therapists’ talk and identifying the subject positions thus rendered available.

In sum, discursive accounts propose that identity has no stable essence but is constituted within interactions and consists of a multiplicity of—more or less transiently held—subject positions. Different subject positions are associated with rhetorical devices that place oneself and the other in different relations, for example, relations of power, competence, knowledge, moral standing, and so on, and these positions are created within the context of existing, culturally provided categories and storylines (Harré & van Langenhove, 1998).

Discursive accounts underscore the multiplicity, fluidity, and fragmentation of subjectivity, in contrast to most psychological accounts that focus on the integrity and coherence of self-identity. This tendency has been criticized for losing touch with the phenomenological and experiential realities of everyday living, whereby most of us experience our self as having a sense of continuity in time and across situations. Lived experience would suggest that subjects may be discursively decentered, that is, multiply positioned, as suggested by positioning theory, but phenomenologically centered, that is, experienced as defined by internal processes and with a limited range of positions available (Guilfoyle, 2014). Indeed, discursive theorizing cannot readily account for the observation that people tend to inhabit specific subject positions, often holding on to them rigidly and with great tenacity, such that any shift in positioning seems difficult to achieve. To address this observation, several authors combined psychoanalysis with discursive analyses to explain individuals’ unconscious investment in specific subject positions (e.g. Frosh, Phoenix, & Pattman, 2003; Georgaca, 2005; Hollway & Jefferson, 2000), while others examined identifications with specific subject positions in terms of ideology and power drawing upon Foucault and post-Foucauldian theories (e.g. Guilfoyle, 2014). Despite these attempts, the dynamics and processes through which individuals tend to repetitively deploy an—often limited—range of subject positions are not fully addressed within discursive theorizing.

Related to the above, another contentious issue in discursive theories of subjectivity concerns so-called discourse determinism, that is, the assumption that culturally available discourses define experience and identity. This has been criticized for producing a version of “blank subjectivity” (Parker, 1994) and failing to account for changes in positioning, agency, and resistance to the power of discourse. Several authors agree that subjects are constituted by discourse but not completely subjected to its power; this brings forth questions regarding the nature of that which lies outside discourse and resists its power that have not been fully addressed in constructionist theory.

Finally, the almost exclusive focus on language has been criticized for sidestepping the embodied aspects of human life as well as the role of the material environment in the processes of social construction; we discuss embodied aspects of positioning in more depth in the next section.

Extra-discursive Aspects of Positioning

In recent years, several authors have increasingly commented on the limits of positioning theory in describing the embodied and affective aspect of human life and argue that our sense of self, the experience of distress, and the processes implicated in social construction are always embodied (e.g. Cromby, 2012). In these accounts, subject positioning is considered a corporeal as well as a discursive accomplishment, as the subject is seen to be constituted through joint actions that take place in situated, embodied, and material contexts (e.g. Shotter, 1993). In this framework, embodied and affective processes are not conceptualized as independent of culture and discourse but as distinct, dynamic processes that are inscribed in discourse (e.g. Burkitt, 2014; Wetherell, 2015).

Research on communication suggests that face-to-face dialogue takes place through the intertwined cooperation of different modalities (speech, facial expression, gesture, gaze, body posture, prosody, and aspects of the material surroundings) that work together to create meaning (e.g. Goodwin, 2000). Discursive analyses can discern some aspects of non-verbal positioning through examining the form and organization of talk (e.g. shifts in footing, pauses, hesitations, etc.) but some aspects of the construction of subjectivity arguably take place outside the verbal realm; this is a point not yet adequately addressed in discursive research. Moreover, expanding our conceptualization of positioning to include extra-discursive aspects creates tensions on both theoretical and methodological levels that have only recently begun to be systematically addressed in the constructionist literature (e.g. Wetherell, 2015). Below, we briefly outline some literature that we consider could help expand our conceptualizations of subject positioning to include non-verbal and embodied aspects.

Although recognized as important, prosodic and other non-verbal aspects of talk are rarely analyzed systematically in discursive studies, with a few exceptions (e.g. Tomicic, Martinez, & Krause, 2014; Weiste & Peräkylä, 2014). These studies have shown that prosody plays an important role in creating meaning and in mutual positioning, independently from the content of talk. Even more difficult to incorporate in discursive analyses are visuo-spatial aspects of communication, for example, gesture, facial expression, orientation, and so on. Starting with Freud’s view of symptoms as symbolic expression of underlying unconscious conflict, several theories contend that un-narrated experiences are “told” through the body or through action. Recent knowledge about procedural memory and implicit relational knowledge has provided a basis for the assumption that some aspects of experience may be potentially unsayable (e.g. Cromby & Harper, 2009). A few discursive studies have started to include non-verbal aspects of communication in their analyses (e.g. Bavelas, McGee, Phillips, & Routledge, 2000) but none have done so in a systematic manner in the context of psychotherapy. Moreover, very few studies have attempted to examine silences or that which is not said in therapy (e.g. Itävuori et al., 2015).

Another interesting development includes the study of the embodied aspects of co-construction as reflected in psychophysiological reactions. There is some evidence that processes of self-construction, identity negotiation, and positioning are associated with increases in autonomic arousal (e.g. Lyons & Cromby, 2010). Similarly, some recent studies found evidence for psychophysiological synchrony in interacting dyads during narration of emotionally laden stories (Peräkylä et al., 2015; Voutiläinen et al., 2014). Furthermore, few recent studies have attempted to link embodied aspects of interaction with meaning construction and positioning in psychotherapy, but this literature is still in its infancy (e.g. Päivinen et al., 2016; Seikkula, Karvonen, Kykyri, Kaartinen, & Penttonen, 2015).

In summarizing the above, social constructionism has provided a powerful critique of traditional accounts of the self but not a complete theory of personhood. Within this literature, there is an ongoing tension between the recognition of the interactional, situation-specific constitution of subjectivity, on the one hand, and the phenomenological experiences of continuity, coherence, and agency on the other. Also, despite the evidence that extra-discursive factors play an important role in human interactions, constructionist researchers and theoreticians have been rather slow in including the affective and embodied dimensions of human life in their study of interaction in therapy. At the same time, however, constructionist accounts have provided a powerful analytic tool, that of subject positioning, that can help us explore the relational and semantic processes through which identity is talked into being within psychotherapeutic conversations. In the next section, we selectively present and discuss studies that have used these tools to examine the reformulation of subjectivity in therapy.Footnote 1

Discursive Research on Subjectivity in Psychotherapy

As already mentioned, discursive accounts approach subjectivity as multiple, variable, and fragmented; a key conceptualization of psychological difficulties arising from this perspective assumes that these are associated with a narrowing of the repertoire of available subject positions for the client, which constrains lived experience and limits options for action. Accordingly, therapy aims to enhance the clients’ ability to flexibly adopt a wider range of subject positions. Following on from this, several discourse analyses examine the range of subject positions that clients employ in the course of therapy (e.g. Avdi, 2005; Frosh, Burck, Strickland-Clark, & Morgan, 1996), and it has been proposed that the flexibility with which clients position themselves in the therapy room can be used as a micro-outcome variable (Strong, Busch, & Couture, 2008) in psychotherapy research. Another aspect of the repertoire of available subject positions concerns the relationships between positions. For example, some of our subject positions are central to the way we define, experience, and present our self, and these tend to be in line with dominant ideologies. This dominance of some positions means that other subject positions may be thinly narrated, under-elaborated, or unassimilated. Indeed, some studies have shown that traumatic experiences, in particular, may remain unstoried and the corresponding subject positions unavailable, marginalized, or even dissociated (e.g. Bromberg, 1998). In such cases, the aim of therapy is that of fostering marginalized subject positions, thus expanding the person’s position repertoire. To date, there are only a few discursive studies that explore the organization of the client’s position repertoire, by examining, for example, the dynamics of dissociation and the degree of accessibility of different subject positions (e.g. Avdi, 2016; Guilfoyle, 2016). Finally, a few discursive studies have focused on the emergence through therapy of a superordinate, reflexive meta-position, that observes and talks about other subject positions (e.g. Georgaca, 2003).

A central issue in discursive research on therapy relates to agency. Several discursive studies have shown how clients manifest compromised agency, often associated with problem-saturated discourses, and have argued that therapeutic change is associated with clients coming to occupy more agentic positions; on the other hand, it has been argued that sometimes clients take on too much agency in the form of critical self-blame (Wahlström, 2006). Agency, however, is a thorny issue in social constructionist accounts; it has been problematized as reflecting the modernist psychological subject, and thus promoting an ahistorical, decontextualized, and over-psychologized representation of subjectivity (e.g. Henriques et al., 1998). On the other hand, agency is recognized as an important aspect of resistance to the dominance of discourse and, in many cases, an implicit aim of therapy.

Reflecting these theoretical tensions, the discursive studies that examine the negotiation of agency in therapy approach it in diverse ways. Some assume that agentic subject positions are indeed associated with psychological well-being and treat the emergence of increasingly agentic positions in the client’s talk as evidence of therapeutic change (e.g. Burck, Frosh, Strickland-Clark, & Morgan, 1998). Others examine the interactional processes through which clients display weak agency, the rhetorical work through which agency is promoted, and the processes through which clients adopt more agentic positions, without discussing the ideological or practical implications of these practices (e.g. Madill & Doherty, 1994). Another group approach the construction of agentic subjects critically, as part of the ideology that promotes certain ideals regarding personhood, and illuminate the active role of therapists in producing “psychological subjects” in need of expert intervention (e.g. Guilfoyle, 2001).

The negotiation of agency has also been explored in relation to accountability and responsibility for one’s actions. This is an issue that is particularly relevant in couple and family therapy, where questions around who the client is and who needs to change are often contentious and associated with blame (e.g. Avdi, 2015a; Kurri & Wahlström, 2005; Stancombe & White, 2005) as well as in situations where the client’s actions are associated with morally delicate issues (e.g. Pärtanen, Wahlström, & Holma, 2006) or where the context of therapy is semi-mandatory (e.g. Seilonen, Wahlström, & Aaltonen, 2012).

Discursive research reveals the therapist’s active role in promoting certain versions of reality and subjectivity, and this allows us to consider the ways in which specific sessions are embedded in wider systems of meaning, moral values, and ideology. Several discursive studies focus on the role of dominant discourses in the construction of particular versions of personhood. For example, several studies examine the implications of the medical discourse for the clients’ agency; psychiatric conceptualizations have been shown to compound the person’s experienced difficulties, as they provide pathological subject positions for the “patient” that constrain his or her agency, limit possibilities for action, and contribute to the person’s alienation (e.g. Avdi, Lerou, & Seikkula, 2015; Karatza & Avdi, 2010). Children are another category of person that has been shown to be awarded half-membership status in therapy talk; for example, a series of studies have shown how children are simultaneously involved in and marginalized within family therapy practice (e.g. O’Reilly, 2008). These practices can be shown to be in line with discourses around child development and the claims to knowledge, rights, and responsibilities attributed to the categories of “child” and “adult” within these discourses (e.g. Avdi, 2015b). Other studies have focused on the effects on subjectivity of discourses around gender, explicating how these discourses create sexually divided subject positions (e.g. man/woman, father/mother, stepfather/stepmother, son/daughter, husband/wife) with specific rights and responsibilities, expected behaviors, as well as moral status and how these positions are intricately implicated with constructions of problems and their solutions (e.g. Suoninen & Wahlström, 2009).

In sum, in discursive studies, the difficulties that clients experience are often seen to arise from the use of a limited range of culturally dominant but pathologizing discourses, which restrict the range of subject positions that can be adopted. Accordingly, therapy is seen as a process of shifting the dominance of these discourses and enabling clients to adopt more varied subject positions, thus enriching the client’s lived experience and sense of self. From this perspective, the role of a discursively aware therapist can be described as entailing, on the one hand, the deconstruction of dominant discourses that are implicated in problem-saturated subject positions for the client, and on the other hand, the co-construction of alternative discourses that sustain more empowering subject positions. This presupposes therapist flexibility in terms of discursive positioning and the conversational moves employed in the course of therapy, as well as the therapist assuming a meta-position of reflexive awareness of the complex, and power-imbued, discursive processes that take place in therapy.

Following this brief introduction to relevant theory and research, in the next section we present a brief analysis of an extract from one session of couple therapy, with an aim to illustrate the usefulness of a discursive approach to analysing the processes through which meanings and identities are negotiated in psychotherapy talk.

Case Analysis: Reformulating Meanings and Identities in Psychotherapy

The example we present is from a couple therapy that was conducted in a public mental health clinic in Greece, and the material was collected in the context of a broader research project.Footnote 2 All sessions were videotaped and permission has been granted for the use of the session material for research purposes. The therapy lasted 15 months, spanning over 15 sessions. There were two experienced female family therapists whose work included the use of a one-way mirror as well as reflecting conversations. The clients were a heterosexual, married couple. Both partners (Costas and Demetra) were white, Greek, in their mid-30s, and had been living together for several years before the birth of their son 10 months earlier. They came to couple therapy to resolve difficulties in their relationship that had intensified through conflicts in the division of labor following their transition to parenthood.

In this brief example of analysis, we focus on one of the issues in this therapy that concerns the difficulties Demetra experienced in the transition to motherhood; this was a highly emotive issue that dominated the initial sessions and was discussed throughout the therapy. We selectively focus on this topic, using material from the third session, to illustrate how the problem, initially diffuse and vaguely described, is gradually reformulated into a psychological issue that can be addressed in therapy—and note the therapist’s active role in this process; we also illustrate how culturally preferred ideals are implicated in personal distress.

In the first session, Demetra reported struggling with the demands of caring for a young baby. She became very frustrated with the baby’s demands, to such an extent that she often found herself screaming, kicking, and hurtling objects in rage. These outbursts were followed by intense guilt and her sense of being trapped in a terrible situation she could not manage. Early in the first session, she evocatively described how she sometimes felt like killing her baby, because his relentless demands for care led her to exhaustion.

Before turning to the extract, it is worth noting that Demetra’s account is in stark contrast to culturally desirable and socially acceptable behaviors expected of mothers. Despite the powerful critique of dominant representations of motherhood articulated by feminist scholars for more than 30 years and the increasing recognition that motherhood is a multi-faceted experience, ideals about motherhood abound and continue to influence women’s experiences and identities, saturating everyday practices and interactions, and promoting social processes of gendered stratification (e.g. Goodwin & Huppatz, 2010). Images of “the good mother”—in the media, popular culture, public policy, and social institutions such as education and work—remain prevalent and continue to regulate women’s lives, functioning as standards against which women are judged and judge themselves, as mothering continues to occupy a powerful position in women’s identity (Arrendell, 2000). Contemporary feminist theorists have illuminated the diversity and flux that characterizes current representations of motherhood and have illustrated how such variations on a theme (e.g. the good working mother, the good lesbian mother, shared parenting, etc.) operate in different contexts with complex and sometimes contradictory effects, as they intersect with social processes related to social class, race, ethnicity, heteronormativity, as well as gender (Goodwin & Huppatz, 2010; Sévon, 2011).

In the analysis that follows, we focus on one aspect of the good motherhood ideology that is particularly relevant to this case. Feminist research has illuminated the prevalence of a discourse of intensive mothering in Western societies, which rests on the claim that mothering is exclusive, wholly child-centered, emotionally involving, and time-consuming (Hays, 1996). The subject position of good mother in this discourse is that of a self-sacrificing woman, devoted to the care of others, with no needs or desires of her own, an intuitive nurturer, naturally equipped and always readily available to care for her children, setting their needs as her exclusive priority. The intensive mothering ideology both assumes and reinforces traditional gender-based division of labor and an idealization of the nuclear family (Goodwin & Huppatz, 2010).

The interaction presented below took place in the middle of the third session, in which the focus was on Demetra’s low mood and her sense entrapment. Until that point in the session, Demetra was reluctant to talk and responded to the therapist’s attempts to engage her with single-word answers or with a simple “I don’t know.” She stated that her main problem related to feeling bored and finding little enjoyment in her life; she complained of having no time to herself, no social activities and few interests, and no sense that any change was possible. This diffuse, rather vague sense of distress and boredom dominated the discussion in the first half of the session. In terms of positioning, Demetra seems to simultaneously resist and adopt dominant discourses of mothering, which associate motherhood with personal fulfillment and happiness. The therapist’s discursive agenda, that is, the overarching effects of her talk on the evolving conversation, can be seen to be that of reformulating this rather vague—and morally delicate—problem as a psychologically meaningful difficulty, while inviting Demetra in a less problematic and more agentic subject position. This is reflected in the therapist’s persistence in exploring the possible meanings and causes of Demetra’s experiences and providing links between her difficulties, her personal history, and dominant constructions of motherhood, evidenced in the extract below.

1 Th:

It seems like, especially you Demetra, like you are saying that, OK, you are bored with the things you used to do, it seems like you don’t really have a wish to do other things, but (.) you are also crying (.) and this confuses me, what exactly are you crying about? (.) is this crying disappointment?

2 D:

(5) I don’t know [crying] (7) I don’t know

3 T:

Costas, do you know why Demetra is crying now? (.) Do you have a fantasy about it?

4 C:

I don’t want to

5 T:

You don’t want what?

6 C:

To have a fantasy

7 T:

Why?

8 C:

Ehm, it is more dangerous

9 T:

Dangerous?

10 C:

Yes, better to know [laughter] what it is about

11 T:

Mmm, Do you want to know?

12 C:

Mmm, yes. Wouldn’t I want to? (9)

13 D:

I don’t know [crying]

14 C:

Mmm, OK, I didn’t mean that you must tell me, when I said that I want to know

15 D:

I generally feel, sometimes, that I am suffocating in this situation that (.) I don’t want any more [crying] he pisses me off (.) I don’t have any patience

16 Th:

The baby?

17 D:

Yes

18 Th:

He pisses you off?

19 D:

Hugely [cries] sometimes (.) I feel like throttling him (.) that would teach him (.) I don’t have any patience

20 Th:

He tires you, hmm?

21 D:

Yes, he tires me, I don’t know, I find it difficult to make the transition from the world of adults to this other world

[brief interchange about the difficulties all adults have relating to babies]

22 Th:

We talked about it briefly last time and I think that this is important (.) perhaps this has to do with an internal conflict you have between an imaginary, ideal mum and the mum that you actually are (…) to what extent would you allow yourself to sometimes be annoyed with a little creature who constantly demands (.) and to not be available? How likely is it that you’d allow yourself to experience this without feeling guilty (D- hardly at all [crying]) you wouldn’t allow it at all?

23 D:

No, because when I get annoyed, I get annoyed (.) and I start shouting and screaming and kicking things, so that I don’t bash him of course, so this is a very intense situation, it’s not like “listen now, I am annoyed so calm down before…”

24 Th:

Hmm, perhaps you reach (.) you reach your limits

25 D:

It’s not like “talk him through it,” it’s like (.) “go away, get away from me”. The dog runs away, the child runs away, everyone runs away, like (.) this is like (.) mental illness [cries]

26 Th:

Who do you consider an ideal mum, Demetra? How did you, how do you imagine a perfect mum?

27 D:

I don’t know [cries] My mum was a super-mum

[brief discussion about Demetra’s mother]

28 Th:

What is your fantasy of the ideal mum? Because you are perhaps chasing that ideal and that’s why you reach your limits

29 D:

Yeah, in this tv ad for (margarine brand), the one in the ad for (margarine brand) (.) who is tall, blonde, in superb shape, always smiling, with the perfectly laid table and the perfect breakfast, on time, and with perfect nails

30 Th:

Yes, but you know that this life exists only in ads (.) or don’t you know that?

31 D:

Yeah (.) OK (.) this life exists only in ads

32 Th:

What I mean is, if there is a constant struggle inside you, about “what kind of mum am I? Am I a good mum or not?” and you set yourself such high standards, I understand that you get disappointed, because you feel that you are not the mum you’d like to be, and then it seems that everything else becomes boring

In turn 1, the therapist summarizes Demetra’s perspective about the problem (boredom and lack of interest) and then invites further elaboration of her experience by addressing Demetra’s non-verbal expression of sadness. Until that point in the session, Demetra had been crying silently from time to time, but only responded briefly and non-committedly to the therapist’s questions about this. Demetra seemed to be in a position of someone who is fed up with her life, exhausted, and withdrawn—with no agency or desire. Her experience seems under-narrated and stagnant, with no history to her troubles, no progression, and no causes or meanings associated with them. This subject position is both articulated and enacted in the interaction, as Demetra is a reluctant, un-cooperative client until that point. In her turn, the therapist marks Demetra’s silent crying as puzzling (“it confuses me”) and then suggests a possible psychological explanation for it (disappointment), thus inviting Demetra to elaborate. Demetra maintains a helpless position of someone who is in distress but does not know why.

There follows a brief interchange, where the therapist uses circular questioning and invites Costas to assume a position of observer of Demetra’s distress. Costas responds from the position of partner, who addresses Demetra with gentleness and concern; in this way, Demetra is positioned as the primary client at this point in the session by both Costas and the therapist.

Interestingly, Demetra responds from the position of distressed mother, rather than that of a distressed partner, and starts to talk about her sense of entrapment, suffocation, and anger. We hypothesize that Demetra’s rage and aggression towards her baby are delicate issues, as her feelings and actions are in sharp contrast to socially expected maternal behaviors; in a sense, she assumes a position of “bad mother.” The therapist responds gently to this disclosure and Demetra describes her desire to “throttle” her baby.

The therapist ignores this strong statement and responds with a reformulation, reframing Demetra’s aggression as fatigue. Reformulations are powerful discursive tools that selectively focus on one aspect of what has been said by the previous speaker or put a particular spin on it, thus changing the previous utterance while seemingly accepting it. They are commonly used rhetorical strategies in therapy talk that cast clients’ complaints into the language of therapy (e.g. Davis, 1986; Buttny, 2004), and function to promote the reconstruction of the problem in line with the therapeutic agenda. As such, reformulations have an ideological dimension, in the sense that they promote a particular version of reality and subjectivity in line with therapeutic assumptions. In this case, through the therapist’s reformulation, Demetra is called into the position of an exhausted rather than an aggressive mum; Demetra accepts this position call, which she elaborates upon introducing the difficulties in relating to babies. This new construction is further elaborated in turn 22, where the therapist introduces the idea of “internal conflict,” a key notion in psychological discourse. At the same time, Demetra’s initial account is altered quite significantly: the intensity of Demetra’s aggression is toned down (she “sometimes” gets “annoyed” rather than “pissed off”), there are distancing markers (“you” don’t allow “yourself” to experience such feelings) and a psychological explanation is provided that moderates the rage and takes attention off it (the problem is not Demetra’s rage per se, but the fact that she doesn’t allow herself to experience it). In this formulation, there is a very demanding little creature and an exhausted mum, who naturally sometimes gets annoyed; in this way, Demetra is positioned again as a good mum, who struggles because she tries to be perfect.

Demetra initially seems to accept this formulation but soon returns to the position of “bad mum,” as she describes her anger as too intense and concludes that her response is abnormal (“mental illness”). The therapist again ignores this powerful statement and shifts topic; this shift is in line with her discursive agenda of constructing a non-pathological narrative, which is gradually built up into a complete interpretation in turn 32, and this interpretation becomes central in the remaining session. In turns 26–32, the therapist builds a formulation that links Demetra’s difficulties with her high expectations that she constructs as resulting from her personal history (having a super mum) as well as from culturally dominant ideals about motherhood (TV advertisements). The therapist challenges both these constructions in a playful and humorous tone, an indication that she is managing a delicate issue (Buttny, 2004).

So, through this sequence of talk, the problem is reconstructed from boredom, to frustration and aggression, to psychological conflict and Demetra is re-positioned from a bad to a good mum, who is struggling because she tries to be perfect. Thus, the problem becomes one of perfectionism in relation to “false” ideas about motherhood rather than maternal aggression. The ideology around “perfect mums” and Demetra’s “perfectionism” becomes the topic of conversation—and deconstruction—in the remaining session. In terms of discourses of motherhood, the therapist both invokes and deconstructs a caricature of ideal motherhood, which is consistent with the discourse of intensive mothering. At the same time, she brings forth a more moderate normative, discourse of motherhood, calling Demetra to assume a position within it and thus enabling a more acceptable subject position for her as a mother. It is worth noting that much less discursive work is performed in this therapy on Costas’ corresponding position as father, and we consider this to reflect the relative power of motherhood discourses in defining and regulating women’s subjectivity.

Discussion

It is clear from the discussion above that subjectivity is at the center of the psychotherapeutic process. Addressing the client’s concerns in therapy always implicates issues of identity—including responsibility, agency, accountability, and morality—and so the meanings of the “problem” that brought the client to therapy are intricately implicated with how speakers position themselves and important others. In other words, the semantic work of psychotherapy entails “joint work on identity projects” (Wahlström, 2006) that often take the form of a negotiation and reformulation of the clients’ identity, arguably in ways that open new avenues for experience and action.

We have argued, and demonstrated through the extract analysis, that the concept of subject positioning is a valuable tool for investigating the processes of reconstructing subjectivity in psychotherapy, as it enables the examination of the role both of interactional processes and of sociocultural discourses in shaping the client’s self. We hope to have also demonstrated the importance of investigating the role of the therapist in actively shaping the psychotherapeutic process and thus forming particular, culturally and therapeutically preferred, client subject positions. In line with recent trends of discursive work on psychotherapy, we think that the non-verbal, affective, and bodily aspects of human life should be incorporated in discursive analyses of psychotherapy. While broadly acknowledging the non-verbal aspects of subject positioning, we regrettably have not been able to provide an analysis of non-verbal aspects of the brief interchange analyzed here. We certainly hope the analysis of subjectivity in psychotherapy in the future develops along these lines.

Finally, we hope to have demonstrated the relevance of positioning theory for clinical practice. We contend that recognizing the centrality of positioning for subjectivity, the multiplicity of positionings—for both therapists and clients—as well as the interactional, constructed, and context-bound character of positioning can enrich therapeutic understanding (Parker, 1999). We would argue that a distinctive feature of discursively informed approaches to therapy is attentiveness to the role of ideology and culturally dominant discourses in client distress. Accordingly, discursively aware therapists strive to exercise reflexive awareness of the discursive processes that constitute the process of therapy and flexibility in positioning within clinical conversations, in the service of deconstructing problematic discourses and opening up space for alternative discourses that enable the emergence of more empowering subject positions for the client.