qualitative research papers attached and for that paper, identify and comment on the theoretical perspective and key ethical issues apparent in the paper.


lable at ScienceDirect

Social Science & Medicine 163 (2016) 98e106

Contents lists avai

Social Science & Medicine

journal homepage: www.elsevier.com/locate/socscimed

“Halfway towards recovery”: Rehabilitating the relational self in narratives of postnatal depression

Meredith Stone a, b, Renata Kokanovic c, *

a New South Wales Institute of Psychiatry, 5 Fleet Street, North Parramatta, NSW 2151, Australia b Prince of Wales Hospital, Barker St, Randwick, NSW 2031, Australia c School of Social Sciences, Faculty of Arts, Monash University, 20 Chancellors Walk, Clayton Campus, VIC 3800, Australia

a r t i c l e i n f o

Article history: Received 15 February 2016 Received in revised form 21 June 2016 Accepted 22 June 2016 Available online 23 June 2016

Keywords: Australia Narrative interviews Postnatal depression Qualitative study Subjectivity Relational self Phenomenology Psychoanalysis

* Corresponding author. E-mail address: [email protected] (R

http://dx.doi.org/10.1016/j.socscimed.2016.06.040 0277-9536/© 2016 Elsevier Ltd. All rights reserved.

a b s t r a c t

This article explores expositions of subjectivity in accounts of postnatal depression (PND). It examines the public narratives of 19 Australian women contributing to a health information website (healthtalkaustralia.org), collected across two Australian qualitative research studies conducted between 2011 and 2014. For the first part of the paper we analysed narrative data using a combination of phenomenological and psychoanalytic techniques. We found that postnatal distress was described in embodied, relational terms and that women depicted their distress as a pre-verbal intrusion into ‘known’ selves. We interpreted this intrusion as a doubly relational phenomenon - informed at once by a woman’s encounter with her infant and her ‘body memory’ of earlier relational experiences. For the second part we examined how and why women classified this relational distress as PND. We drew on illness narrative literature and recent work on narrative identity to explore why women would want to ‘narrate PND’ e an apparently antithetical act in an environment where there is a duty to be a good (healthy) mother. We highlight the dual purpose of the public PND narration e as a means of re- establishing a socially sanctioned known self and as a relational act prompted by the heightened rela- tionality of early maternity. Our focus on the salutary aspects of narrating PND, and its links to relational maternal subjectivities, offers a novel contribution to the current literature and a timely analysis of a largely uninterrogated sociocultural phenomenon.

© 2016 Elsevier Ltd. All rights reserved.

1. Introduction

The distress of women diagnosed with postnatal depression (PND) has been considered from different vantage points. Histori- cally, biomedical researchers have tended to examine disease processes, hormones and psychological vulnerabilities (O’Hara and Swain, 1996; Wisner et al., 2002). They have also considered stressful life events, lack of social support and marital discord (Beck, 2001; O’Hara and Swain, 1996); however, largely within a positivist epistemology more concerned with the ‘nature of disease’ than women’s own language about suffering.

A counterpoint is the growing body of work using feminist and health sociology perspectives e frameworks that foreground both the lived experience of early maternal distress (e.g. Edhborg et al., 2005; Rodrigues et al., 2003) and the broader social context in

. Kokanovic).

which that distress emerges (Oakley, 1980; Nicolson, 1998). A third perspective has sought to integrate the psychic and the

social. This is reflected in the growing interest in how concepts such as identity change and ‘loss of self’ are linked to postnatal distress (Beck, 2002; Everingham et al., 2006). However despite the current popularity of these concepts, there remain theoretical gaps in our understanding of self and identity in the early maternal context. The increasingly sophisticated body of work on the psychosocial self that has been emerging elsewhere in the health sociology field is yet to be incorporated into studies of PND. In particular growing interest in the relational self (e.g. Fullagar and O’Brien, 2014; Ussher and Perz, 2008), embodiment (e.g. Freund, 1990; Williams, 2000) and intersubjectivity (e.g. Crawford, 2009) has, with the exception of work by social psychologists such as Natasha Mauthner (2002, 1999), remained largely absent from the PND literature.

PND as a sociocultural phenomenon has also received remark- ably scant attention. Current studies tend to either accept the concept uncritically or dismiss it altogether. Nevertheless there is an unspoken consensus across theoretical frameworks that women

M. Stone, R. Kokanovic / Social Science & Medicine 163 (2016) 98e106 99

do not want to be diagnosed with PND. This is reflected in the attention paid to PND and stigma in the biomedical literature (e.g. Edwards and Timmons, 2005), and conversely in the feminist literature seeking to dismantle the concept (e.g. Nicolson, 1998).

The proliferation of fora in which the diagnosis is openly dis- cussed and actively ‘claimed’, including ‘mummy blogs’ (see Circle of Moms, 2012), celebrity memoirs (Rowe, 2015; Shields, 2006), and PND support groups has in contrast remained largely unin- terrogated. Furthermore, despite asserting the centrality of identity to postnatal distress, the PND literature has yet to incorporate re- flections on the identity work associated with claiming a diagnosis, as discussed in the context of ‘generic’ depression (Kokanovic et al., 2013), terminal illness (Avrahami, 2003) and chronic illness (Adams et al., 1997; Asbring, 2001).

In this paper we address these lacunae in the literature and comment on their interplay. We reference the interdisciplinary turn in contemporary work on subjectivity (e.g. Butler, 2001) and combine feminist psychoanalytic and existential phenomenology frameworks with a poststructural reading of illness narratives. Our objective is to generate a more nuanced understanding of how distress in early maternity, and the process of diagnosing and ‘storying’ this distress, relate to maternal subjectivities as they exist in neoliberal societies.

2. Theoretical framework

2.1. Pre-verbal selves

Most existing PND literature describes maternal identity change as a linear, narrative arc from constructs such as ‘career woman’ to ‘stay-at-home mother.’ It attributes the distress associated with this move both to structural realities and to the dissonance between how a woman experienced herself before motherhood and how she sees herself, or fears others might see her, after maternity (Beck, 2002; Nicolson, 1998; Oakley, 1980). Importantly, it is a process that is seen to occur at a conscious, speaking ‘I’ level.

This is at odds with the increasing emphasis on pre-conscious or non-verbal aspects of the self in contemporary research on subjectivity. One thread is influenced by psychoanalytic thought. Judith Butler (2001), for instance, draws on object-relations the- ories in particular to discredit the Kantian notion of the unitary self. Butler holds that when we respond to changes in our lives, we do so both at conscious and pre-conscious levels. Furthermore, the pre- conscious is populated, or indeed ‘besieged’ (Butler, 2001, p. 74), by earlier experiences with an Other. Butler argues that before we become thinking, speaking beings there must be a ‘primary, inau- gurating impingement (…) by the Other’ (2001, p. 85), where the Other is at once material caregiver and the sociocultural norms (s) he transmits. This extends the Freudian notion that our earliest relationships are ‘carried over’ through Übertragung (transference) and inform how we relate to others and feel about ourselves in subsequent relationships (Freud, 2000). It also suggests the un- conscious, or pre-verbal, is an under-researched site for ‘identity struggle’ in women experiencing early maternal distress.

Another thread of subjectivity research focuses on embodiment. This references Heidegger’s idea that the body is at once K€orper (body-as-object) and Leib (lived body) - the vehicle through which we experience ‘being-in-the-world’ (Heidegger, 1967). The idea that the body can mediate experience has been important for recent thinking about illness and subjectivity. Drew Leder (1990) writes on how a body in pain becomes a body-as-object, disrupt- ing a person’s sense of lived body, and by extension harmonious being-in-the-world. There is also an expanding literature on the embodied mind. Thomas Fuchs (2012) combines phenomenolog- ical and psychoanalytic ideas to describe the notion of ‘body

memory’ e the corporeal storing of experience that re-emerges not as explicit memory but in posture and patterns of behavior. Simi- larly, Elizabeth Wilson (2004) locates the mind, and by extension subjectivity, as much in the gut as in the conscious self.

Finally, subjectivity scholars have in recent years advocated a ‘turn to affect’ (see Wetherell, 2013), seeking to expose the non- linguistic elements that shape subjective experience, and move between subjects. Authors concerned with affect focus not only on embodiment, but also on intercorporeality. Lisa Blackman (2010), for instance, has written extensively on how matter, affect and even non-linguistic traumatic memories circulate between, and consti- tute, embodied subjects.

2.2. Maternal subjectivities

That an embodied and relational understanding of the self might be particularly relevant to early maternity is not a novel idea. French feminist psychoanalysts such as Julia Kristeva and Luce Iri- garay, and more recently, maternal studies scholars such as Baraitser (2009), Stone (2012) and Hollway (2015) have devoted considerable attention to the relational maternal self. Both Irigaray (Irigaray and Wenzel, 1981) and Kristeva (1977) argue that the experience of maternity, starting with the two-in-one state of pregnancy and extending into the profoundly bodily, pre-verbal rhythms of early mother-infant ‘being-with,’ destabilises a wom- an’s preexisting conceptualisations of autonomous selfhood.

Bracha Ettinger (2010) calls this process ‘fragilisation.’ She em- phasises the intersubjective potential of such a state, but cautions that it is threatening for women in neoliberal societies, who usually function in an autonomous ‘phallic’ mode. Crucially for our un- derstanding of early maternal distress, it is especially confronting for women whose infantile experiences of relationality were trau- matic. Often such women were ‘defensively’ autonomous (see Hollway, 2015, Chapter 7) before becoming mothers, a character- istic at once prompted by a desire to escape the threat of rela- tionality and rewarded by patriarchal systems requiring autonomous workers.

This brings us to the ‘doubly relational’ (Stone, 2012) nature of the maternal self e a self not only constituted by a woman’s in- teractions with her infant, but also by her experiences of having been an infant herself. Maternal studies scholars suggest a partic- ularly potent form of Übertragung in early maternity (Raphael-Leff, 2000; Stone, 2012). Joan Raphael-Leff (2000) posits that the post- natal haze of smells, secretions and intimate encounters activates a woman’s pre-linguistic memories of infancy, which in turn ignite infantile states of being or, as Sara Beardsworth (2004, p. 263) terms it, an ‘upsurge of forgotten body relationships.’ A woman therefore encounters her baby for the first time, but also (re-)en- counters her self-as-baby, or more precisely, her self-as-baby- relating-to-mother. To Raphael-Leff (2000), early maternal distress is thus located as much in a woman’s past relations as it is in present maternal encounters.

The prominence of the pre-verbal and the infantile in maternal subjectivity is of course contested e Kristeva, for example, in her focus on the maternal semiotic, has been criticized for placing mothers ‘outside culture’ and reifying woman-nature/man-reason binaries (see Butler, 1990). However, recent maternal studies scholars have engaged creatively with the problem of the relational mother who also speaks. Lisa Baraitser (2009) depicts mothers who exist at the nexus (and in the interstices) between discourse, intersubjectivity and maternal materiality. She presents a pecu- liarly maternal subject but not one existing before culture or entirely pre-consciously. Instead, to paraphrase Blackman (2010, p. 174), she proposes a maternal subject ‘living multiplicity’ (both through mother-infant intercorporeality and the Foucauldian

M. Stone, R. Kokanovic / Social Science & Medicine 163 (2016) 98e106100

Other) whilst attempting to ‘hang together’ (Blackman, 2010, p.174) as her own ‘I.’ One way in which she might do this is through narration, a topic to which we now turn.

2.3. Illness narratives and subjectivity

Critical examinations of PND narratives are conspicuously scarce. Hillary Clark (2008) provides an engaging critique of Brooke Shields’ autopathography (2006), but focuses more on Shields’ readers than on Shields’ own negotiations with subjectivity.

In the broader illness narrative literature there have been a number of approaches to the narrating self. Early work in the field focused on how people use narrative to engage in ‘biographical repair’ when illness interrupts ‘storied lives’ (Bury, 1982; Frank, 1995). This approach emphasizes the ability of narrative to pro- ductively answer the ‘who am I now?’ question for people encountering unexpected illness.

Authors influenced by Foucault have focused on the regulatory function of illness narratives. They have highlighted how people speak of illness in order to demonstrate that they are good patients and to self-regulate the deviance associated with not being a fully- functioning neoliberal subject (Avrahami, 2003; Kokanovic et al., 2013). In the mental health literature the role of diagnosis has been explored by Kokanovic et al. (2013), who observe that people will often actively claim and narrate diagnoses so that worse forms of deviance (e.g. laziness) might be mitigated.

Recently the illness narrative literature has also been shaped by the turn to affect. Frank’s (1995) idea that non-narrated illness experiences represent ‘chaos’ has been questioned by authors re- orienting to non-narrative ways of negotiating illness (Woods, 2013). Others have turned attention to the relational function of illness narratives. Einat Avrahami (2003), for example, writes on the author Harold Brodkey’s HIV autopathography, exploring how Brodkey employs narrative to mobilise audiences to engage ‘emotionally and ethically’ (Avrahami, 2003, p. 183) with his predicament.

The question of how this scholarship might apply to PND nar- ratives is yet to be explored. To what extent is the PND narrative an attempt to re-instate biographical coherence? How does the PND narrative articulate with broader ‘good mother’ discourses? In what ways does it respond to the peculiarities of the relational maternal moment? How does it give voice to what can only be felt? We hope to begin to answer these questions in this paper.

3. Methods

In this article we draw on data from video and audio recorded narrative interviews undertaken to populate two sections of an Australian health information website (www.healthtalkaustralia. org). The narrative data were collected in two separate qualitative research studies conducted in Australia: Experiences of depression and recovery (conducted between May 2011 and June 2012) and Emotional experiences of early parenthood in Australian families (conducted between May 2013 and July 2014). Both studies used rigorous scholarly methods developed by the Health Experience Research Group (HERG) at the University of Oxford (Ziebland and Hunt, 2014). We performed a secondary qualitative analysis of women’s spontaneous accounts of PND e seeking to better un- derstand the nature of the experience and to explore the personal and sociocultural contexts.

3.1. Original research

Participants were recruited through newsletters, website advertising, support groups, word-of-mouth and healthcare

practitioners. Ninety women and men were interviewed across Australia. Each study aimed to capture wide variation in experi- ences (e.g. diagnosis, disclosure, treatments (first study); experi- ences of same-sex parents, IVF, surrogacy (second study). Both studies recruited participants from diverse socio-demographic and ethno-cultural backgrounds.

3.2. Current sample

In this paper we present our analysis of interview transcripts with 19 women from both studies, selected because they sponta- neously talked about having been diagnosed, or self-diagnosing with, postnatal depression. Women ranged in age from 23 to 51 years, and came from predominantly middle-class, Caucasian backgrounds although some recent immigrants and socioeco- nomically disadvantaged participants were included. The ‘PND subset’ of participants was significantly more homogenous than the overall sample. We were unable to explore this phenomenon further in the scope of this paper, but believe it warrants future research.

All 19 women identified as heterosexual and were either mar- ried or in de facto relationships. Eleven women had two children and seven had only one child at the time of interview. One woman had six children. Of the 19 women, 13 had received a formal diag- nosis of PND, while seven self-identified as having experienced PND. The discrepancy is because one of our participants self- identified as having PND after the birth of one child and received a formal diagnosis after the birth of another. All who had received a diagnosis (bar one) were prescribed antidepressants. Of the 19 women, 15 were in full or part-time work and one was undertaking full-time study. Three women identified as stay-at-home mothers. Most women had post-secondary qualifications and were employed in white-collar jobs. One woman owned a business.

3.3. Interviews

The narrative interview was designed to explore experiential aspects and interpretive practices situated within personal narra- tives (Potter and Wetherell, 1987). Participants were initially asked to tell their story of experiences of depression (first study), and emotional experiences of early parenthood (second study). A topic list was used in the second part of the interview to prompt par- ticipants to discuss study-related themes if these had not already emerged. These included: life before depression, diagnosis, treat- ment and recovery experiences (first study) and conception, pregnancy, birth, becoming a parent, and earlier life history (second study). Interviews lasted one to four hours and were mostly con- ducted in people’s homes. They were video or audio recorded with participants’ informed consent, transcribed strict verbatim and returned to participants for verification before being de-identified and coded using NVivo (for the purpose of analysis for the healthtalkaustralia.org website). The second author (RK) was the lead investigator of both studies. Importantly, neither of the studies explicitly asked about PND. Pseudonyms were assigned to partici- pants. Ethics approval was granted by the Monash University Hu- man Research Ethics Committee (MUHREC).

3.4. Analysis

In our analysis we undertook an in-depth empirical examination of the sociocultural and experiential reality of women experiencing PND. We initiated the research process by reading and re-reading interview transcripts where PND had been mentioned. We then began iteratively reviewing the social sciences literature on PND and related themes and returning to the transcripts with specific

M. Stone, R. Kokanovic / Social Science & Medicine 163 (2016) 98e106 101

questions/theoretical frameworks in mind. We became particularly interested in expositions of subjectivity after observing the preva- lence of themes of identity and ‘loss of self’ in the PND literature and noticing a) the limited reference to contemporary interdisci- plinary writing on subjectivity in the literature, and b) the rich descriptions of early maternal subjectivities in the interview transcripts.

In our literature review we were struck by pluralist approaches to the study of subjectivity as espoused by writers such as Judith Butler (2001), who manages to integrate psychoanalytic, Fou- cauldian, existential phenomenological and feminist frameworks when exploring the ‘nature of being.’ We found a similar theo- retical flexibility in the work of contemporary maternal studies scholars (Baraitser, 2009; Hollway, 2015; Stone, 2012). This inspired our interdisciplinary approach to data analysis using a combination of phenomenological and psychoanalytic frame- works. When coding we used both ‘hermeneutics of meaning recollection’ and ‘hermeneutics of suspicion’ (Ricoeur, 1970) e in the former seeking ‘an understanding of the experience on its own terms’ (Davidsen, 2013, p. 329) and in the latter trying ‘to shed light on the material from a more distant perspective, such as psychoanalytic (…) theory’ (Davidsen, 2013, p. 329). As well as using narratives as a rich source of qualitative data, we were interested in narrative performances, co-construction (between participant and researcher/imagined public) of narratives, and the ways in which narratives were both constituted by and constituted sociocultural realities.

Initially the transcripts were read independently by each author. The first author then conducted a preliminary analysis using the iterative process described above. The second author read the preliminary analysis, provided reflections on interpretations, and suggested additional literature to strengthen the analysis. This process was repeated over several months with regular meetings in person and on Skype. If there were disagreements about interpre- tation, each author read the relevant literature and returned with ‘substantiation’ for their interpretation. The authors then agreed on which approach had most coherence with other sections of narrative and the published literature.

In our first two sections of data analysis we use phenomeno- logical and psychoanalytic insights to present a relational, pre- verbal ontology for postnatal distress, as elaborated by the women we interviewed.

In our third section we use narrative inquiry to examine how the process of having this distress diagnosed (or self-diagnosed) as PND, and then ‘narrating PND’ related to early maternal sub- jectivities. In particular we explore how these processes enabled a (self-)regulation of the ‘disordered’ maternal body and a relational response to (re-)accentuated relational selves.

3.5. A note on terminology

Postnatal (or postpartum) depression is a contested term. It has recently been changed to depression ‘with peripartum onset’ in the DSM-5 (American Psychiatric Association, 2013), reflecting uncer- tainty in the psychiatric community about distinctions between ‘depression’, ‘prepartum depression’ and ‘postpartum depression’ (see Sharma and Mazmanian, 2014). We tend to refer to postnatal depression, as one focus of the article is to interrogate how di- agnoses were talked about, and postnatal depression was still the favoured term amongst participants. We are mindful of the con- ceptual difficulty with the term(s) (see Fullagar and O’Brien, 2014; Nicolson,1998). For this reason we refer to post/perinatal suffering/ distress where possible.

4. Findings

4.1. The speaking ‘I’ interrupted

4.1.1. Embodied affects In our study several women described the prominence of

corporeal experiences of PND. Amelia, a 33-year-old married mar- keting professional and mother of a 3-year-old girl, described her experiences after the birth of her daughter:

At first the symptoms of the actual postnatal depression were very much physical, um, I lost my appetite completely, I would sweat suddenly, um, and I - I had a dry mouth to a point that it didn’t matter how much I drank, didn’t matter how much I washed my teeth, my mouth was just dry and, um, not very pleasant.

In Amelia’s account we see both Leder’s (1990) foregrounding of the body-as-object and witness Amelia’s failed attempts to respond in a ‘rational’ way. This primacy of the body was echoed in Daphne’s description of experiences of PND. Daphne, a 36-year-old married customer service worker and mother of an 8-month-old baby boy, described her “episodes” as “purely physiological”:

That sickness, like in the pit of my stomach like I always have that, you know that nervous um sick feeling um in my stomach the whole - like all the time. Um and like just a burning in my chest, like this is just purely um physiological (…) and you just have that general sort of cloudy - like my mind goes all cloudy.

In both Amelia and Daphne’s narratives we discern a hyper- embodiment - a body that interrupts reason spaces, resisting Amelia’s ‘rational’ responses, and clouding Daphne’s mind. These descriptions are reminiscent of Sandor Ferenczi’s (in Wilson, 2004, p. 75) observation that when ‘the psychic system fails, the organism begins to think.’ They extend beyond biomedical notions of physical symptoms, instead invoking a more profound corporeal disruption of pre-maternal subjectivities.

4.1.2. Disembodied thoughts and visions Disruptions were not just described in physical terms. As in

other studies, women described feelings of guilt (Beck, 2002; Edhborg et al., 2005; Mauthner, 1999), inadequacy as a mother (Edhborg et al., 2005; Mauthner, 1999), and thoughts of self-harm (Beck, 2002; Mauthner, 2002) or harming the baby (Beck, 2002; Mauthner, 2002; Raphael-Leff, 2000) as core features of the condition.

In a number of narratives these themes appeared in the form of intrusive thoughts or obsessions, thoughts that women in our study and elsewhere (see Beck, 2002) described as being qualitatively different from other thoughts and particularly distressing.

Louise, a 37-year-old married writer with a 2-year-old daughter and a 10-month-old son, told us she had developed PND a few weeks after the birth of her daughter:

I just started to have funny thoughts, lots of strange thoughts about um, the harm that could come to her. And then it just sort of became an obsession (…) it stepped into territory where I was thinking that um, I would - you know, I could, you know, not only accidentally cause harm to her but purposely cause harm to her. I started to get plagued by - by this (…) it was almost like a 24/7 thing and trying to suppress them and push them away

M. Stone, R. Kokanovic / Social Science & Medicine 163 (2016) 98e106102

Louise was at pains to convey that the thoughts were not ones she had authored. They were “strange thoughts” she needed to “push (…) away.” Paradoxically then, although thoughts are generally considered to be part of the ‘I’ that emerges through language, these were thoughts that mocked such a linguistic self. They were not consciously generated, but rather experienced as intrusions. Louise suffered from similarly intrusive “visions”:

I could feel tension, fuzziness, in my wrist. I - actually this is all coming back to me now. There was a stage where I was really uncomfortable around being around knives. Um, not that I thought I was going to harm myself, but that I - I was just having these sort of visions of like my - jabbing myself in the stomach, like with a knife.

Implicit in this account is a connection between the visions and the fuzziness in the wrists - a potential for action conveyed along neural pathways not involving a conscious self. The visions were therefore at once disembodied and embodied e they occurred independently of the eyes, the bodily apparatus for seeing e but somehow found their way into the wrists, heralding the possibility of transformation from vision to action.

Louise’s account has an air of Heidegger’s (1967) Unheimlichkeit (uncanniness) e the sense of a body that has lost its resonance with mind and world and instead asserts its own malignant authority. Here, her narrative echoes Amelia’s description of her body following a visceral (il)logic impervious to ‘acts of reason.’ …