Morality and emotion in therapy clients’ accounts for suicidality

ABSTRACT

Suicide is a substantial public health problem, yet little is understood about the ways in which clinicians and clients discuss suicide attempts. This study investigates commonalities in psychology clients’ accounts of suicide attempts that were survived, or aborted after being enacted or planned and prepared. We used conversation analysis to investigate conversations within psychosocial (risk-assessment) sequences during four initial psychology consultations (interviews) recorded in an Australian voluntary outpatient multidisciplinary healthcare clinic. Analysis showed that clients’ suicidality accounts routinely included reports of morally problematic treatment by others, with clients typically seeking clinician responses to these elements of their accounts. Clinicians were not, however, observed to engage the moral dimensions of clients’ accounts and instead routinely oriented to the impact of these behaviours on clients. We conclude by discussing the implications and consequences of these differing orientations for psychologists, clients, and future suicide-risk research.

KEYWORDS:

• Suicide risk
• suicide risk assessment
• therapist communication
• conversation analysis
• psychology consultation

Introduction

Globally, suicide is reported to account for the death of over 700,000 people annually (World Health Organization, 2021). In Australia, in 2020–21, 1 in 6 people (16.7%) aged 16–85 reported experiencing suicidal ideation or behaviours during their lives, including 7.7% who had made a suicide plan and 4.8% who had attempted suicide (Australian Bureau of Statistics, 2022). Additionally, over 1 in 10,000 Australians suicide annually, making suicide the leading cause of premature death (Australian Bureau of Statistics, 2023). On average, those who die by suicide are reported to have three to four known risk factors (such as relationship or economic problems), with around 85% having at least one known risk factor (Australian Bureau of Statistics, 2023). Consequently, conducting a suicide risk assessment (which considers risk factors; Steele, Thrower, Noroian, & Saleh, 2018) is considered to play an important role in suicide prevention (Ryan & Oquendo, 2020). A meta-analysis of 365 studies that longitudinally examined risk factors for suicide-related outcomes (i.e. suicidal ideation, plans, attempts or deaths) found that predicting suicide is difficult, with existing risk factors failing to achieve clinical significance (Franklin et al., 2017). Nonetheless, it identified that a prior suicide attempt is among the top predictors of suicidal attempts and deaths (Franklin et al., 2017).

Similarly, preparatory suicide behaviours and aborted suicide attempts are considered suicide risk factors and are therefore included on prominent risk-assessment questionnaires, for example, the Columbia-Suicide Severity Rating Scale (Posner et al., 2011), which is commonly used by psychologists in the United States (Clay, 2022), and the Suicide Assessment Kit (Deady, Ross, & Darke, 2015) used in Australia. Despite researchers emphasising the importance of the client-practitioner relationship in suicide risk assessment (Sommers-Flanagan & Shaw, 2017), few investigations have examined the conversational elements of suicide risk assessments. It remains unknown what considerations clients and psychologists routinely orient towards when discussing suicide attempts that were survived, or aborted after being enacted or planned and prepared (henceforth referred to as survived or aborted suicide attempts).

Studies of conversations that occur more broadly within suicide risk assessments may provide insight into the considerations that we might expect psychologists and clients to orient towards during accounts of prepared or attempted suicides. We begin by reviewing research regarding conversational openings in risk assessment (i.e. clinicians’ elicitation questions), in order to furnish an understanding of the sequential environment into which clients’ accounts of suicidality are delivered in such assessments. Although our analysis focuses on accounts of suicide, we also discuss research on self-harm risk assessments as it has been found that, in practice, clinicians may use the term “self-harm” to refer to both suicide and non-suicidal self-injury during risk assessment and that questions regarding self-harm and suicide are often packaged together (Ford, Thomas, Byng, & McCabe, 2021). McCabe, Bergen, Lomas, Ryan, and Albert (2023) analysis of 46 self-harm and suicide risk assessment sequences in emergency departments showed that when asking closed-ended questions, approximately 50% of psychiatric liaison clinicians used leading questions that invited a “no, not suicidal” response. Such “not suicidal” responses were invited through the use of words such as “any,” “ever,” or “at all” (e.g. “Do you have any thoughts of harming yourself at the moment?”) or through negative declaratives (such as “You’re not planning on doing it again?”; pp. 3, 6). These leading question forms were associated with less frequent self-harm disclosures from clients than when questions were designed to lead to “yes, suicidal” responses. Similarly, in a study of 77 psychiatric consultations, 75% contained negatively phrased initial suicide questions from clinicians (McCabe, Sterno, Priebe, Barnes, & Byng, 2017). Together, these findings suggest that clinicians treat the possible disclosure of suicidal thoughts and behaviours as dispreferred and that soliciting suicide stories is a sensitive interactional matter.

Perhaps for this reason, one of the two main ways that mental health practitioners have been observed to preface suicide assessments when working with young people is by building incrementally on emotion-related talk (O’Reilly, Kiyimba, & Karim, 2016). Similarly, rather than directly soliciting talk on suicidal behaviours, clinicians often design their initial questions with a focus on thoughts or feelings before moving on to suicidal behaviours (Ford, Thomas, Byng, & McCabe, 2021). On a suicide helpline, for example, call-takers were observed to formulate suicidal ideation as a feeling, which they subsequently empathised with without endorsing suicidal ideation or impulses (Iversen, 2021). Together, such studies suggest that an emotion orientation in professionals can often appear at a variety of conversational junctures in suicide talk and suicide risk assessment. Additionally, in cognitive behaviour therapy (CBT), emotions are reasoned to be connected to clients’ thoughts and behaviours (Beck, 1970, 1976). However, while this prior research and CBT itself might lead us to anticipate some form of emotion orientation in suicide risk assessments in psychology consultations (operating within a CBT paradigm), it is unclear exactly how (or if) emotional orientations emerge from clinicians during clients’ accounts of survived or aborted suicide attempts.

Although some research exists on clinicians’ orientations and actions in suicide risk assessments, comparatively less research has focussed on the content of clients’ disclosures during risk assessment or how clinicians orient to such disclosures. Ford, Thomas, Byng, and McCabe (2021) conducted perhaps the most detailed analysis of patient-clinician interaction in instances where clients responded affirmatively to primary care doctors’ suicide and self-harm questions (which, as noted above, focussed on thoughts and feelings). In such cases, clients typically delivered affirmative responses in narrative form or with additional contextual information. In these descriptions, clients often positioned themselves as aware of the moral implications of their suicidal thoughts (e.g. by orienting to the way that it would adversely impact others) and justified their decision not to act with reference to such concerns. Ford, Thomas, Byng, and McCabe (2021) argued that this reflected an orientation towards, and distancing from, a stigma of selfishness associated with suicide. They also reported that doctors displayed little interest in pursuing the moral elements of patient’s responses and focussed instead on risk management (by transforming the object of clients’ moral concern into protective factors). However, it is unclear what orientations towards moral matters, if any, clients display when accounting for survived or aborted suicide attempts. In such cases, clients face different phenomena to account for (suicidal actions [including preparations] and suicidal thoughts that lead to action, rather than suicidal thoughts that did not lead to action). Similarly, it is unclear how clinicians treat clients’ orientations during these accounts if, indeed, clients regularly display an orientation to moral matters.

This study aims to address several gaps in the literature. One involves examining the considerations that clients and psychologists routinely orient towards when discussing survived or aborted suicide attempts. We also address how (or if) therapists orient towards emotion during clients’ accounts of such suicide attempts, what orientations towards moral matters, if any, clients display when accounting for such suicide attempts, and how therapists treat clients’ moral orientations during these accounts. To address these gaps in the literature, we use conversation analysis (CA) to examine whether an emotion (or experiential impact) focus is evident from psychologists/provisional psychologists when clients describe survived or aborted suicide attempts. We also examined if and how clients oriented towards moral matters during their accounts of survived or aborted suicide attempts, and the extent to which psychologists/provisional psychologists take up such matters. CA involves turn-by-turn analysis of talk to examine the social actions (such as offering, accepting, or requesting) and orientations of interlocutors (understandings and treatment of conversational elements, as well as observable focusses in conversation). CA is therefore well-suited to investigating the presence of an emotion (or experiential-impact) focus from clinicians and the extent to which clients orient towards moral matters during their accounts of survived or aborted suicide attempts.

Method

Research design

The session recordings used in this study are part of a wider examination of interaction in psychology consultations. Data analysed and reported here are unique to this study. Here, CA was used to explore interactional patterns in recorded suicide and self-harm risk assessment sequences during initial psychology consultations. CA is a sociolinguistic research methodology that analyses naturally occurring social interaction in institutional and everyday settings (Couture & Sutherland, 2006; Lopriore, LeCouteur, Ekberg, & Ekberg, 2019; Smoliak et al., 2022). A central pillar of CA is that social interaction is orderly as it is constructed moment-by-moment by speakers who draw from a shared set of culturally available practices and sense-making strategies (Smoliak et al., 2022). CA therefore works to uncover the “rules, procedures, and conventions” that organise conversations by analysing social actions, shared knowledge, and social context (Goodwin & Heritage, 1990, pp. 283–284).

In order to uncover such conventions in the present study, the first author examined clients’ suicide accounts during suicide risk assessment sequences, focusing on accounts of events leading up to the suicidal behaviours. First, a single instance of a conversational phenomenon of interest was observed. Then, additional instances were searched for in the corpus and compared with one another as we looked for patterns in features of talk, such as speakers’ orientations (treatment of talk, topics, or actions), social actions, and design of turns at talk. Through repeated listening to suicide risk assessment sequences and reading and re-reading of accompanying transcripts, we identified such patterns. Throughout this analysis, we grounded our analytic understanding of the unfolding conversation in the observable understandings of speakers.

Extracts were transcribed according to Jefferson’s (2004) conventions. Some elements of talk, including gender pronouns, have been altered or removed to preserve participants’ anonymity. Extracts are labelled with a randomly generated psychologist ID(s) (from A-E) and a client ID (from 1–4). For anonymity reasons, we cannot report the registration status of therapists in a corpus of this size (i.e. provisional psychologist or registered psychologist). Throughout this article, unless otherwise specified, we use the term “psychologist” to refer to both provisionally and generally registered psychologists. The Human Research Ethics Committee at The University of Adelaide approved this research (approval no. H-2021-136).

Participants

Participants consisted of four psychology clients, one registered psychologist, and three provisional psychologists. The four recorded sessions consisted of two dyads and two triads. The triads contained a psychologists/provisional psychologist who primarily conducted the session along with a provisional psychologist in a primarily observational capacity. One dyad contained a provisional psychologist as an observer who did not speak or participate in the research study (and whose demographic information is therefore not provided). Provisional psychologists were placement students enrolled in Clinical and Health Master of Psychology degrees. Clients had a median age of 35.5 years and psychologists had a median age of 36.5 years. The most commonly reported ethnicity for clients and psychologists was “Australian.” Individual participants’ ages and genders are not reported due to ethical constraints and to preserve anonymity.

Participant recruitment and selection

Researchers made contact with the Psychology Clinical Field Supervisor at an Australian voluntary outpatient multidisciplinary healthcare clinic and were invited to present the research opportunity to psychologists and incoming provisional psychologists. Provisional psychologists were on placement at the clinic. Psychologists and provisional psychologists selected and approached potential client participants on behalf of researchers in accordance with ethics requirements. Staff at the health clinic primarily practised CBT. All participants provided informed consent for participation and use of their information.

Data collection

The materials used in this study were selected from a corpus of 13 sessions recorded in an Australian voluntary outpatient multidisciplinary healthcare clinic. Participants also completed demographic questionnaires. Suicide account sequences were included for analysis if clients described a suicide attempt that they had fully planned and prepared but aborted, partially enacted but aborted, or enacted without halting and survived. This criterion resulted in four sequences for analysis. In two cases, clients reported surviving the attempt by aborting it before (but after fully planning and preparing for it) or while enacting it. In the other two cases, clients reported that they survived the attempt despite their intentions. The risk assessment sequences in which suicide accounts typically occurred took place in the final 10 or 20 minutes of sessions, sometimes involved a break between risk assessment and safety plan implementation (for example, providing clients with helpline numbers), and ran for approximately 6 minutes (two instances), 13 minutes (one instance), and 24 minutes (one instance), respectively.

Results

In accounting for their suicidal ideation and/or suicidal behaviours, clients routinely reported the precipitating circumstances as involving morally problematic (transgressive) behaviour towards them. Clients did this either by referring to events or actions that had been previously described (and characterised as morally problematic), or by describing new events during risk assessments in session. These assessment sequences occurred in various locations during the session but typically appeared towards the end of the session. Despite clients routinely searching for and displaying an expectation of a response from psychologists regarding the moral elements of their stories, therapists did not provide talk on this topic. Instead, therapists engaged a project of searching for evidence that such events had sizeable impacts on clients.

When describing morally problematic events in these sequences, clients sought non-minimal responses from psychologists. As we show below, they did this throughout their descriptions of moral transgressions by extending their own hearably complete turns and by pausing for psychologists to take the floor. Psychologists, however, typically provided minimal responses to client turns that described moral transgressions, thereby indexing less interest in these topics, and instead pursued other matters. This pattern is illustrated in Extract 1 below, where the client references a previously discussed issue regarding their experiences in high school. During information gathering, the client had described how they were “picked on because I have arm hair, I have leg hair, I have braces.hhh uh I’m the ugly friend.” The client has referenced circumstances for which they were not responsible – arm hair, leg hair, and braces are all routine elements of being a teenager, and their appearance was beyond the client’s control. By attributing the “cause” of their bullying to such elements of their personhood, they positioned the behaviour of perpetrators as unjust. They subsequently described how they “could not handle” this bullying and dropped out of school, with “could not handle” conveying a general sense of the event’s emotional impact on the client. From line 15 of Extract 1, the client describes a sexual assault they experienced as a “dumping issue” (an issue that they are “dumping” on the therapist):

Extract 1 (1CD)

Consistent with prior research (Ford, Thomas, Byng, & McCabe, 2021), the psychologist’s question on lines 6–8 asks about prior suicidal thoughts, but the client’s extended response includes talk about prior suicide attempts. The client pauses after introducing the first transgressive antecedent to their suicidality (“high school” on lines 11–12). Here, they allude again to the emotional impact of their experiences (“hating high school”). After the go-ahead from the psychologist on line 13, the client again pauses, suggesting that, in response to bringing up these transgressive incidents, they anticipated or sought something besides a go-ahead.

The client then introduces another morally transgressive incident, a sexual assault from their uncle (lines 15–22). Once again, we see a pause, go ahead, and pause pattern on lines 23–25. This pattern again suggests that the client expects or seeks further talk from the psychologist while the psychologist expects or seeks further story information from the client (by taking only a minimal turn on line 24). The client then recompletes their turn (“yeah!”) before pausing briefly on line 26. The client’s “on top of” construction in this extract (line 27), paired with the “already” categorisation of their high school issues (line 27), positions the abuse from their uncle as having an adverse impact by appealing to a previously described circumstance that negatively impacted them (“high school”). The client thus renders their suicide story as hearably the result of two problems for which they were not responsible and which had a significant emotional impact. In addition, they do moral work in describing the people who perpetrated these actions against them. Their uncle’s actions are attributed to his identity as a “really bad potheadhh” (line 19), a label that categorises him as morally deficient. This attribution is achieved through sequential positioning; by describing him as a “really bad potheadhh” right before they describe his behaviour, they work to attribute his actions to his pejoratively characterised drug use and thereby casts him as morally problematic.

After this description, the client once again pauses, receives a go-ahead from the psychologist, and silence ensues before the client re-takes the floor (lines 28–30). Again, the client displays an expectation or search for a more substantive response from the psychologist. After this silence, the client moves on from line 31 to describe their self-harm and suicide attempts. By moving directly from describing transgressive behaviours and their impacts to describing their self-harm and suicide attempts, the client treats the transgressive events and their impacts as a sufficient explanation for their attempted suicide. Overall, the client has worked to describe their suicide attempt (lines 33–45) as the result of impactful things done to them by morally at-fault (and deficient) transgressors. They have also displayed an interest throughout the account in receiving non-minimal contributions from the psychologist. Throughout this account, the psychologist did not pursue morally transgressive elements, instead offering minimal responses and continuers. In the remainder of the analysis, we will attempt to gain insight into what the client may have expected or sought in these silences, and why the psychologist contributed only minimally when the client disclosed morally problematic behaviour.

In the next extract, we see evidence that the feature of these events that clients sought a response towards was their morally transgressive nature and the feature that psychologists looked for was the impact of these events on clients. Extract 2 begins as the client is describing events at a party they did not want to hold, but their partner decided to hold at the request of a friend:

Extract 2 (2AB)

When discussing their “very deep” and “private” conversation, the client aborts their utterance on line 6 (“we w-”) and initiates an insert repair. The aborted utterance appears to have been “we were jumped over,” which the client redeployed on lines 6–7. The insert repair (“in that conversation”) is itself repaired (“in mid-conversation”). The effect of these repairs is to position the speech of the third party (the one who jumps over them) as intrusive or interruptive and, therefore, as morally problematic. This function of the repair is confirmed by the client’s subsequent use of the term “jumped over,” rather than a more neutral alternative such as “joined in” or “spoken to.” When asked to explain what they mean by the fact that they were “jumped over,” the client’s explanation again confirms that the words “jumped over” were used to convey a sense of moral transgression. This is done by describing the third party’s actions as unjustified (“all of a sudden” on line 13), self-motivated (“they decided” on line 13), and inconsiderate (“just gonna come” on lines 14–15). Like the client in Extract 1, after describing this morally problematic event, the client pauses, suggesting an expectation or orientation to a response from the psychologist. However, like the psychologist in Extract 1, the psychologist does not take the floor. Having failed to secure a response, the client continues to describe the transgressive nature of the incident (“just interrupt” on line 17). Although this move is delivered with sentence-closing intonation, the client immediately appends an upgrade to the transgressive nature of the incident (“straight over top”), before yielding the floor. This appended upgrade suggests that it is the morally transgressive nature of the incident that the client seeks or expects a response towards. When the psychologist provides only acknowledgement of this incident (“right”), the client recompletes their turn on line 19, and thereby provides the psychologist with another opportunity to formulate the expected response. The clients’ pursuit of a response indicates that a response is not just expected by the client but is being actively sought. The psychologist, however, does not engage with the transgressive elements of this story.

We also note that, following the client’s initial attempts to secure a response from the psychologist, it is only after the client volunteers the impact of these transgressive events (“it upset me,” lines 9–10) and after an extended period of silence while the psychologist types, that the psychologist retakes the conversational initiative. In doing so, the psychologist displays an orientation towards impacts as the sought-after component of the clients’ problem story at that juncture.

We have seen how clients worked to convey precipitating events as transgressions and sometimes volunteered that these had substantial impacts on them. When not volunteered by clients, psychologists searched for a display of the impact of these events on clients. In the following extract, we see further evidence for this as well as evidence that psychologists’ focus on impacts in these accounts reflects their orientation towards that which constitutes a sufficient account for suicidality. On line 1 of Extract 3, we join the conversation at a point where the psychologist first seeks an explanation for the client’s suicidality (“that” on line 1). The client’s use of the indexical “he” (from line 3) refers to their husband. At this point, they have not yet confirmed that their suicidal ideation progressed to a plan or preparation:

Extract 3 (3C)

The client’s initial response to the psychologist’s question about why they were suicidal involves delivering details regarding their husband’s morally transgressive behaviours towards them. They first complete this account on line 7, indicated by their falling intonation and subsequent pause. The psychologist does not take up the floor at this juncture. The client then resumes their description of their husband’s transgressions, this time leaving a substantial pause after using direct reported speech to convey their stance that their husband was acting in a sanctionable way (“I’m not your mother” on line 12, which positions their husband as inappropriately expecting them to act as his mother). Given the function of direct-reported speech in facilitating affiliation at the apex of stories (by allowing tellers to clearly display their stances; see Guardiola & Bertrand, 2013), we might assume that the client was searching for a display of affiliation from the psychologist. The psychologist, however, does not take up this opportunity to speak, and the client continues. When the client reaches a possible completion point in their story (line 24), they pause briefly, and we see the same pause-continuer-pause pattern that we have seen in prior extracts. As in previous extracts, this again suggests that the client was searching for psychologist talk that oriented to the transgressive behaviour of the husband.

Up to this point, the client’s account for why they felt suicidal has focused only on the morally transgressive behaviour of their husband. The psychologist, however, treats this as an insufficient account for suicidality by resoliciting an explanation on lines 29–30. This time, the psychologist asks a narrower question, which presupposes that “loss” is the appropriate pre-cursor to suicidal ideation. The client subsequently delivers further information regarding their marital situation from line 31 and further talk on their husband’s transgressions, which concludes on line 34. Once again, the psychologist does not accept this response as adequate, nor the response to their formulation on lines 35–38, as indicated by their later solicitation on lines 50–53, aimed again at understanding why the client was suicidal. Just prior, on line 29, the client had provided the emotional impact of their husband’s behaviour – that they felt “useless” - but as we will see, this likely failed to account for their suicidality as it failed to convey a substantial impact. The client’s initial response to the psychologist’s re-solicitation (lines 54–59) concludes with an upgraded account of the event’s impact, “it was just getting to me, badly” on line 59. With this, the client finally succeeds in accounting for their suicidality, as the psychologist subsequently pursues a different matter (data not shown), suggesting that this account of a substantial emotional impact was treated as sufficient. Overall, we can see that clients worked to elicit a response from psychologists regarding the morally transgressive behaviours they experienced, sometimes also offering the impacts of these behaviours when initially accounting for their suicidality. Psychologists, meanwhile, displayed little interest in pursuing or affiliating with the topic of morally transgressive behaviours, instead working to find evidence of significant impacts from these behaviours that could account for the client’s suicidality.

The pattern described above – that psychologists are in search of strong impacts during clients’ accounts of suicidality – is also shown in the following extract. The psychologist’s question on lines 1–2 addresses the clients suicide attempt:

Extract 4 (4BE)

The psychologist’s question on lines 1–2 is oriented towards understanding what gave rise to the client’s behaviour. Prior to this, the client had mentioned that the suicide attempt occurred just after their children were removed from their care by protective services (an organisation that they described early in the session as using deception to justify this removal, thereby characterising the removal as morally problematic). The “massive stress,” therefore, was hearably caused by this removal and is explicitly described by the client as the reason that they attempted suicide (lines 6–8). Subsequently, the client does additional work to characterise this stress as being unusually high and position themselves as highly troubles-resistant (lines 14–15). In doing so, the client positions themselves as only having attempted suicide due to the substantial impact of this incident. At this point, the psychologist discontinued questioning regarding suicidality precipitants, indicating they found the client’s account of their decision to suicide to be satisfactory only after the client included references to substantial emotional impacts. This extract therefore supports the interpretation that psychologists are searching for the impacts of events in clients’ suicide accounts.

A final piece of evidence that psychologists were in search of substantial impacts comes from Extract 2, previously presented above, where the client initially offered impacts that were not upgraded in any way. As we noted previously, the client initially offered the emotional impact of the transgressive behaviour on lines 9–10 (“it upset me”). Despite this, the psychologist resolicits the emotional impact of the events on line 27. This time, the client responds with an upgraded account (“pretty upset,” line 28) appended with an additional impact (“and pretty down,” line 31). The psychologist orients to these as upgrades (demonstrated by their reformulation that the client was “so upset and down” on line 33) and treats this as an acceptable answer to their question from line 27 (as they do not ask the client to provide further impacts of the event). A similar example was seen in Extract 3, where the client’s initial impact statement on line 51 (“I just felt useless”) was not treated as sufficient, unlike their later upgraded impact statement. Together, these extracts provide further support for the observation that psychologists were searching for substantial impacts from clients during their suicide accounts.

A consequence of interlocutors’ differing concerns

As we saw, clients sometimes treated impacts as unnecessary in accounting for their suicidality, but psychologists worked to elicit them when such information was not volunteered. We conclude our analysis by discussing a possible consequence of psychologists’ interest in pursuing adverse impacts in favour of the morally transgressive dimension of these stories. In one instance, a psychologist appeared to seek an impactful upshot prematurely, which resulted in the client resisting their solicitations. This sequence, shown in Extract 5, occurred later in the session that Extract 2 comes from:

Extract 5 (2AB)

On line 4, the client describes a transgression committed against them (that they were “laughed at”). The client confirms that they orient towards this event as a transgression by casting their husband’s movie choice as having been a “dig” at them (line 18). As in previous extracts, the client pauses after describing the transgression (line 4), but the psychologist does not take the floor. The client then continues describing the events that preceded their suicide attempt (“walked off to the my bedroom” on lines 5–6). The psychologist, however, disrupts the progressivity of the client’s story with a question in overlap (line 7). Notably, this was not at a transition relevant place, which therefore characterises this overlap as competitive (Abbas, 2020). The psychologist does not allow the story to progress without first clarifying and then investigating the impact of the transgressive incident (“how did it feel … ” on lines 12–13). During these clarifications, the psychologist formulates part of the transgression on line 25 (“people laughed”). There is silence on line 26, suggesting that the client expects or searches for further talk from the psychologist here, and the psychologist expects confirmation from the client. The client provides this confirmation on line 27 but then yields the floor to the psychologist (line 28). The psychologist does not take the floor, indexing a search for expansion from the client. The psychologist further displays this orientation through “and” on line 29. The client provides further information, which the psychologist receipts and seems to type into their computer. The psychologist, however, does not yet seem to have acquired the expansion they were searching for since they go on (lines 34–35) to elicit the emotional impact explicitly from the client. The client’s response, although providing an impact-related upshot (“just hard”), conveys a clear disinterest in discussing their emotions. This is in contrast to their prior pursuit/interest in eliciting transgression-relevant talk from the psychologist, suggesting it is not talk in general but this topic in particular that they do not wish to pursue. This is evident from their immediately subsequent dismissal of the importance of their emotions (and hence the topic of talk; “like whatever” on line 36). The psychologist provides an empathic formulation of the client’s emotional state, likely designed to elicit further emotion talk from the client (suggested by the psychologist’s delay in taking back the conversational reigns on line 39). This attempt is unsuccessful as the client passes the floor back to the psychologist after only minimal acknowledgement (lines 42–43), further indexing the client’s disinterest in impact-focused talk. The psychologist’s formulation on line 47 and mirrored “yeah” on line 49, followed by extended silence on line 50, appear to index their question on lines 44–45 as a second failed attempt to secure further impact-related talk from the client. Through their questions from line 51, the psychologist elicits progress in the client’s story. On line 59, the client’s narrative has now returned to the moment where it was interrupted by the psychologist on line 6 (c.f. “I went to my room” and “walked off to the my bedroom”). Here, the client volunteers impact-related information when asked to continue the story (“cried” on line 61). As it lasted for a “couple hours,” this “crying” forms a significant part of the narrative and, therefore, was unlikely to be omitted in the original interrupted telling of this story (although we cannot be certain). It seems, therefore, that the client has delivered the impact of the transgression without direct solicitation from the psychologist, which retroactively renders their prior resistance as marking the inapposite placement of the psychologist’s requests for an impact upshot. This extract therefore demonstrates the kind of interactional trouble that can arise (extended silences, resistance) due to clients’ and psychologists’ differing orientations in these sequences. The net result of this trouble is that the impact of the event was delivered much later than it may have been if the psychologist had not overlapped the client in pursuit of such impacts.

Discussion

In the instances observed in the corpus discussed here, clients displayed a concern with conveying the morally transgressive nature of behaviour perpetrated against them when accounting for their suicidality, and searched for responses from psychologists that addressed the morally problematic nature of such behaviour. We also saw evidence suggesting that it was affiliation with their stance towards these events (as being morally transgressive [problematic]) that clients were in search of. These findings extend prior research (Ford, Thomas, Byng, & McCabe, 2021) showing that patients often work to display themselves as aware of the moral repercussions of a potential suicide, as the present results show that clients attend to matters of morality, albeit in a different way, when discussing survived or aborted suicide attempts. Additionally, by positioning the behaviour of others as transgressive in the events that preceded their suicidal behaviours, clients could be seen to reduce their own accountability (as accountability for transgressive events rests with the transgressor, not the transgressed). Since, as we have seen, these behaviours were sequentially positioned as the precipitants of suicidality (or the precipitants of affective states that also precipitated suicidality), clients therefore indirectly reduced their accountability for their suicide plans and attempts. In so doing, clients carved out an environment that could be seen to diminish the possibility of a sanction for self-injurious behaviour. That is, clients appeared to orient to a stigma surrounding their culpability for suicide. This stigma orientation aligns with the work of Ford, Thomas, Byng, and McCabe (2021), who suggested that primary care patients worked to avoid aligning with the stigmatic notion that suicides reflect selfishness and that patients do this by attending to issues of morality. Together, these findings suggest that issues of morality, accountability, and stigma are salient for clients during suicide risk assessments, regardless of whether they are confirming suicidal ideation or suicidal behaviour.

We also observed that, despite clients’ efforts, psychologists did not engage clients’ moral claims (for example, they did not affiliate with clients’ stances that others’ behaviour was morally transgressive, nor did they provide extended responses that oriented to the morally problematic nature of events as clients described them). This finding similarly accords with the work of Ford, Thomas, Byng, and McCabe (2021), who found that doctors did not engage the moral dimensions of patients’ talk (but rather focussed on an institutional agenda). Conversely to our findings, it is a routine feature of troubles-tellings in mundane environments that a troubles-teller will receive affiliation from interlocutors regarding the troubles source (Jefferson, 1988). This feature of troubles telling in mundane environments may explain the apparent expectation and search from clients for talk on the moral dimension of their stories. Similarly, this may explain psychologists’ reluctance to engage these topics, as offering such affiliation may deviate from their institutional agenda and, for this reason is similarly avoided in other medical settings (e.g. Benwell & McCreaddie, 2016). Our findings then, appear to align with Ford, Thomas, Byng, and McCabe (2021) by indicating that psychologists’ reluctance to engage moral matters reflects their orientation towards the institutional task at hand.

Psychologists in our corpus oriented towards disclosures of substantial impacts of morally problematic behaviour towards clients, and engaged a project of uncovering such impacts when they were not volunteered. Impacts described by clients were, broadly speaking, emotional or emotion-implicative. This pattern aligns with and extends prior research (Ford, Thomas, Byng, & McCabe, 2021; Iversen, 2021; O’Reilly, Kiyimba, & Karim, 2016) showing that, during suicide talk and suicide risk assessments, clinicians at various times display an orientation towards emotions. In the cognitive model, which underpins CBT, events themselves are considered neutral, and it is the way that a person evaluates an event (through their automatic thoughts and prior beliefs) that gives rise to emotions or behaviours that may influence each other (Beck, 1970, 1976). As psychologists in our sample were primarily trained in CBT, it may be that their focus on the impacts of events reflects an orientation towards better understanding the precipitants of clients’ suicidality (that is, clients’ emotional states) in order to inform risk assessment. This may explain why psychologists treated accounts without impacts as insufficient explanations for why clients were suicidal. Once again, this risk-assessment orientation also parallels the results of Ford, Thomas, Byng, and McCabe (2021) as it suggests that clinicians were primarily focused on the institutional task at hand rather than engaging with clients’ accounts of morally problematic behaviour.

Limitations and clinical recommendation

A limitation of this research is the small size of the corpus. However, there are some inherent challenges in gathering recordings of risk assessments wherein clients provide accounts of preparing for or attempting suicide. It is common practice to request, and for clients to agree to, the recording of assessment and therapy sessions in training clinics to aid trainee development (Brown, Moller, & Ramsey-Wade, 2013). However, clients are less often asked to allow their sessions to be used in research, and many clients invited to participate in this study declined to have their sessions recorded (although exact numbers are not available to us). Additionally, patients report concealing suicidal thoughts (30%) and attempts (10%) from their therapists (Blanchard & Farber, 2016). Preliminary evidence suggests that such concealment can be motivated by shame and embarrassment (Blanchard & Farber, 2020), which may mean that clients who have attempted or prepared for suicide are less willing to be recorded than other clients. An additional challenge for our corpus is that training clinics (like the one where our data was gathered) often accept clients with lower risk and, therefore, may assist clients who are less likely to report suicidal ideation or behaviours. While small sample sizes are common in conversation analytic research, future replication attempts may nonetheless benefit from a larger corpus, which could be obtained by gathering data from a clinic that specialises in higher-risk populations.

A further limitation of this research was our use of audio-only recordings. Using such recordings prevented us from conducting a detailed analysis of embodied features of conversation, including therapists’ interactions with their computers or notepads. In addition, while not specific to trainees or the presence of observers, some research suggests that embarrassment and shame are barriers to suicide disclosure (Blanchard & Farber, 2020; Hom, Stanley, Podlogar, & Joiner, 2017). Therefore, we cannot exclude the possibility that the presence of non-speaking trainees in some sessions impacted talk around suicide due to increased shame or embarrassment from clients. Future studies could expand on this research by addressing the above limitations.

The results of this research are likely transferrable to CBT initial interviews in cultural settings similar to Australia. The multidisciplinary approach of the clinic did not appear to have a bearing on any of the conversational sequences analysed in this article. However, it is possible that these results are applicable only to outpatient settings as inpatient interviews may involve distinct exigencies and considerations for clinicians to manage during risk assessment (such as information from other clinicians, reduced time for interviews, and involuntary admissions). Replication in other contexts could elucidate this in future research.

We saw that the different focuses of psychologists/provisional psychologists and clients (on impacts and transgressions, respectively) led to a sequence where the psychologist “prematurely” solicited the emotional impact from a client, which resulted in resistance from the client in producing the emotional impact of transgressive events. Such resistance could potentially have been avoided if the psychologist had allowed the client to proceed with their story uninterrupted. Consequently, we suggest that it may benefit therapist-client alignment if psychologists display active listening while waiting until clients have finished delivering their stories before eliciting the impact of events from clients. This advice is applicable beyond psychology consultations to other risk-assessment settings where clinicians focus on emotions or emotion-implicative talk.

Practical implications

This article demonstrates that clients and therapists may adopt different orientations during clients’ accounts of suicidal ideation or behaviour and how this may lead therapists to prematurely solicit emotional upshots from clients. Our results suggest that it may be beneficial for psychologists to wait until clients have finished their accounts and then elicit the emotional impact of events from clients.

Author contributions

Thomas Bradshaw: Conceptualisation; Methodology; Investigation; Formal Analysis – Lead; Writing – Original Draft

Amanda Le Couteur: Conceptualisation; Methodology; Formal Analysis – Supporting; Supervision; Writing – Review and Editing

Melissa Oxlad: Conceptualisation; Methodology; Investigation; Supervision; Writing – Review and Editing

Disclosure statement

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

Additional information

Funding

The work was supported by an Australian Government Research Training Program Scholarship.

Notes on contributors

Thomas Bradshaw

Thomas Bradshaw is a PhD candidate in psychology at the University of Adelaide. He has undergraduate degrees in psychology and theoretical physics from the University of Adelaide. His research uses conversation analysis to better understand the interactions that occur in therapeutic settings.

Amanda Le Couteur

Amanda Le Couteur is an Adjunct Associate Professor at the University of Adelaide. She has published in the areas of racism, education, gender, health, and sport. Her current research involves analysis of interaction in medical and counselling interactions.

Melissa Oxlad

Melissa Oxlad is a Clinical and Health Psychologist and Associate Professor at the University of Adelaide. Her teaching has been recognised with multiple awards, including the Australian Psychological Society Early Career Teaching Award. She has published in the areas of health professional education, gender, health, chronic illness and law reform. Her current research employs quantitative, qualitative, mixed methods and review methodologies.