Why affiliation matters: A conversation analysis of complaints calls to the NHS

Objective: Callers making a complaint share their negative experience in complaint narratives that make relevant affiliation from an operator. We examined how call handlers ’ language choices affect both the progress of the call and the stance of the caller. Methods: We identified episodes where affiliation is displayed or noticeably absent in a dataset of 95 complaints calls to the NHS. Two single cases were closely examined using conversation analysis. Results: Affiliation at sequentially relevant moments in conversation helps progress the call and de-escalate the complaint while the absence or misplacement of affiliation may lead to escalation. The latter recurringly involves blaming whilst de-escalation includes practices that diffuse blame. Early intervention in the form of affiliation to the ‘hurt ’ component and the reasoning of the complaint is essential to de-escalation. Conclusion: Our analysis revealed three key functions of affiliation in complaints calls: 1) ratifying the reason-ableness of the complaint; 2) progressing the institutional requirements of the call; 3) de-escalating the complaint. Practice implications: Call handlers should listen for callers ’ cues for legitimization of the complainability of their concerns and seek to provide responses that express affiliation.


Introduction
A complaint is an expression of dissatisfaction for which an individual or organisation is held accountable.In the UK National Health Service (NHS), the right to complain is written into the NHS constitution: all NHS service providers and commissioners must publish their complaints procedures and make available both verbal and written channels for complaining.In the complaints procedures, complaints are formally recorded and acknowledged and then investigated before a formal response/resolution is provided.This paper focuses on initial calls to the complaints departments of two service providers (a Scottish Regional Health Board and a Northern Irish HSC Trust) in which the complainant is making their complaint for the first time.Our analysis examines how the language choices of call handlers (CHs) can affect not only on the progress and efficacy of the call handling, but also the scope, scale and emotional intensity of the complaint expressed by the caller.
Callers and CHs have different interactional objectives which sometimes come into tension.Our data show that what usually matters for the complainant is to tell their story, expressing the full impact of their negative experience, meaning that these are often long and highly emotional stories.CHs on the other hand need to extract the significant facts from the story and work out the next step in the procedure.But they must still be sensitive to the emotional content of the complaint and demonstrate that the complainant has been properly heard and taken seriously.These potentially competing objectives contribute to making complaints communication challenging.

Complaining as an activity
Two broad categories of complaint are usually acknowledged in interactional studies of complaining: direct complaints, where the complaint recipient is also the complaint target and indirect (or third party) complaints, where the complaint target is not a participant in the interaction.The distinction is not straightforward [1] but is negotiated in situ by participants through relevant responses to complaints.Specifically, direct complaints make relevant remediation or apology, while indirect complaints, particularly in mundane conversation, project an affiliative response [2][3][4].Prior research however also shows clearly that complaints oriented to institutional settings and roles/categories proceed differently to mundane complaints with respect to complaint recipiency.In institutional complaints, the direct/indirect distinction is dependent on the degree to which the caller orients to the CH as a representative of the organisation and accountable for the failures of the organisation [5].In addition, affiliation is often constrained by organisational requirements for some degree of professional neutrality [6][7][8].
The NHS complaints calls in our corpus occupy a category of complaining that is arguably distinct from both mundane and more conventional institutional complaints.First, while being indirect complaints about absent parties, they are produced to representatives of the institution complained about who may be treated as vicariously responsible for the wrong doings. 1 Second, very occasionally CHs orient to complainants' narratives as not only reporting, but indirectly seeking mediation of a particular outcome from the complained-about service.Third, whilst constraints about avoiding explicit agreement with the caller seem to operate in this setting, complaints to the NHS are frequently serious and sometimes life-changing and therefore involve high degrees of emotional and moral investment by the complainant with implications for complaint recipiency and affiliation.

Affiliative responses to indirect complaints
The analysis presented here focuses on the role affiliation plays in these complaints calls.Complaining is a risky practice as complainants may be perceived as 'moaners' or 'whingers' [1] and hence typically stress having adequate grounds for complaining [9][10][11][12], for example through a variety of practices which sustain their 'reasonable complainant' identity [9,13].This sensitivity of complaints both in terms of emotion and reason may prompt or require affiliative response by the recipient [14].Affiliative responses work to ratify a complaint [15] by attending to the negative stance conveyed therein.Affiliation is widely defined as involving "taking a stance that matches the teller's stance toward the event(s) being described" [16].While Stivers [16] defines stance as "the teller's affective treatment of the events he or she is describing" (p.35, emphasis added), stance and affiliation are sometimes given wider scope to refer to a broader range of attitudinal perspectives.For example, Stivers et al. [17] suggest that "affiliative responses are "maximally pro-social when they match the prior speaker's evaluative stance, display empathy and/or cooperate with the preference of the prior action" (p.21).Similarly, [18] refer to "communicating one's stance through evaluative or affectual displays"(p.117).In our own analysis, we include in the category of affiliative uptake not only responses matching the emotional stance but also responses matching the broader range of evaluative stances conveyed by complainants.
While some authors refer to "stance alignment" as synonymous with affiliation [eg.[19], in the interest of clarity we maintain Stiver's [16] distinction between structural alignment (with respect to the activity or project in progress) and social affiliation (which relates to stance as discussed).In this vein, we understand affiliation as a phenomenon which supports alignment of interactional projects and thus the progressivity of calls.As we will show in our analysis, in the context of complaints, aligned responses orient to complaint telling as ongoing but do not necessarily ratify the complainability.Rather it is affiliative responses that serve to ratify a complaint as legitimate by affiliating with the stance expressed in the complaint telling.
In focusing on affiliation, it is also important to note that (dis)affiliation is understood as a continuum of response options, where different forms of response convey differing degrees of affiliation [2,3], and whether a particular response counts as affiliative is entirely context dependent [20].Moreover, what exactly recipients affiliate with in the prior speaker's turn matters for interactional outcomes but has remained relatively under-examined [7,cf.[21].By shifting the analytic focus to a more precise specification of what recipients affiliate with, Pino [7] reveals how complaint recipients display an understanding of hurt and blame as distinct but co-occurring constituent components in the action formation of a complaint.Our data supports Pino's analysis of hurt and blame as constituent components of the action formation of complaining and further develops his observation that these components may be displayed or oriented to either implicitly or explicitly.Of particular relevance to our data is Pino's observation that complaint recipients may affiliate with the hurt embodied in a complaint while avoiding taking a position vis-à-vis the blame. 2 In our analysis, we show that adopting a similarly fine-grained focus on the object of recipients' affiliation in our data reveals that recipients may also selectively affiliate with the 'appeal to reason' embodied in a complaint.This is in line with Traverso's [3] observation that "affiliating does not necessarily imply siding with the complaint teller against the third party, since the complaint teller often evokes the complainable feature of her/his Fig. 1.Phase structure of a complaint call (brackets indicate optionality). 1In our data, the complaint department seems to be treated as a relatively separate body from the complained-about services.How complainants see the relationships between different parts of the institution may be consequential for complaining, but this question goes far beyond the scope of this paper.
volves affiliating with the displays of evaluative stance through which the hurt and blame components are expressed, however we will continue to refer to affiliation with hurt and blame for the sake of brevity and clarity.
B. Benwell et al. situation rather than a clear third-party's wrongdoing or guilt.But it does imply that the complaint recipient recognizes that the complaint teller's (negative) feelings are justified, i.e., that her/his situation is complainable" (emphasis added).
The research question of this study asks how the degree of affiliation present in complaint responses affects the interactional outcome in complaints to the NHS.The novelty of our approach is that we combine the idea of affiliation with the existing research on the action formation of complaints to present a detailed picture of how affiliation functions in this type of institutional complaining.

Methods
This paper presents a comparative conversation analytic case study of two complaints calls, drawing on a larger corpus of data from a project examining complaints to three separate NHS organisations in two nations of the UK over a period of ten years. 3

Data
The data are from a corpus of 95 telephone calls made to NHS complaints handling services collected over two time periods.Informed consent was secured verbally from callers at the start and end of the call, and subsequently in writing.Informed written consent was secured from CHs in advance.The corpus was transcribed verbatim, then extracts of interest were transcribed using Jeffersonian conventions [22].Anonymity for all participants (including individuals discussed within the calls) is assured by the alteration of potentially identifying details.
Making a complaint is the primary legitimate activity of this setting.In a typical call, the CH listens to and records the details of the complaint and explains the next steps in the institutional procedures. 4Complaints calls thus progress through a normatively expected phase structure (Fig. 1) in which caller and CH typically remain mutually aligned as complaint teller/complaint recipient [23] 5 .

Analytic approach: Conversation Analysis
Conversation analysis (CA) is a qualitative methodology aimed at revealing what participants treat as relevant during communication [24].It is based on evidence that people orient to the minutiae of how speech is produced (pauses, overlaps, intonation, etc.) to make sense of what social actions are being accomplished by co-participants.To study such relevance structures, it is essential to examine how participants respond turn-by-turn, as the interpretation of a previous action is revealed in the next turn [25].In CA, it is argued that for participants and for analysts "the locus of order is the single case" [26].Analysis of the single case may illuminate the interactional resources deployed to produce that single case or, as here, be illuminated by a cumulative body of evidence built from analysis of collections of candidate cases [27].Our study examined cases of caller pursuit of affiliation and cases of affiliative uptake across our corpus of complaint calls, focusing on sequential outworking in both collections.

Case selection
Throughout our corpus, we observed local (turn-by-turn) effects of affiliation/pursuit of affiliation at both sequence and phase levels [cf.[28].Here we present two cases that illustrate both the local patterns identified and the cumulative impact of repeated affiliation/pursuit of affiliation.The first illustrates how providing affiliative responses can effectively propel the conversation forward.The second shows the features indicating pursuit of affiliation and the incremental progress of pursuit and complaint escalation in the absence of affiliation.

Findings
In the two cases analysed, both callers' complaints center on a lack of service provision.In the first call, affiliation is demonstrated by the CH and the call progresses to service provision and closure.In the second, the CH does not sufficiently affiliate with the concerns of the caller, and the complaint escalates in scope, scale and emotional intensity.

Affiliation progressing the conversation
First, we focus on a case where the CH produces affiliative uptake in response to the caller's pursuit of affiliation while the precise timing of the affiliation helps the CH to progress the call.
This call involves a patient who broke her toe and has already been twice to A&E.She is calling to complain about having been refused a follow-up appointment at the fracture clinic.During the complaint narrative phase, the caller repeats three concerns several times: that the broken toe is causing pain (9 repetitions); that she has been passed back and forth between A&E and the fracture clinic (6 repetitions); that a fracture clinic staff member was 'very abrupt and rude' (4 repetitions).These concerns orient to the hurt and blame components of the complaint, the last involving explicit blaming of an individual staff member.In addition, we argue that the repeated references to pain show how the caller observably orients to the need to justify the reasonableness of her complaint in the context of a putatively minor injury, with assertions such as, "I know it's only a broken toe but it's very uncomfortable and I'm having like cramps you know in the arch of my foot".
In Extract 1a, the patient repeats her complaint about the rude staff member at the fracture clinic who refused her an appointment.This extract shows how the CH's display of affiliation achieves transition to the next phase of the call.Extract 1a. 6 'Real Complaints' is an NIHR-funded qualitative study (2020-2023) combining conversation analytic and ethnographic research, including interviews and diaries.The data for this paper comprises recordings of initial calls from an earlier pilot study (2010-2012) and the Real Complaints project.4 Calls may also involve discussion as to the status of the concern presented and whether it should be recorded as a formal complaint.This is particularly the case where calls are treated by the CH as presenting problems that they may be able to help to resolve. 5 Occasionally, CHs offer to do more than simply record the complaint or account for why they cannot do more (for example, commenting "we are not an emergency service").In these cases, the callers in our data typically orient to these CH moves away from complaint recipiency as unexpected or to the CH accounts as unnecessary, which points to shared orientation to the constraints on service delivery in this institutional setting.Indeed, in follow up interviews, some complainants express discomfort at the idea that complaining might give them an unfair advantage over other (non-complaining) patients in their care.
In other words, although callers may express desired outcomes for the complaints process, the participant orientation evidence indicates that they nonetheless remain aligned to the interactional project of complaining.
This extract occurs at the end of the complaint narrative.The caller repeats her complaint about the rude staff member, escalating the blame component by explicitly attributing to them a complaint-worthy attitude (ll.65-7).This attribution gets acknowledged by the CH with a token 'okay'.This minimal response does not affiliate with either the hurt or the blame and the caller makes a double move to justify both components of her complaint: (1) in line 70 she produces an imagined reason for the complained-about party's behaviour -that the injury is minor; (2) immediately after that she offers an alternative perspective by asserting the in-fact serious impact of this 'minor' injury (l.72).Through this elegant move, the caller takes an evaluative stance towards the incident as a minor event with major consequences.This transformation is amplified with the insertion of a laughter particle into the word 'toe' and a rising pitch which holds throughout both utterances.
The CH is very receptive to the work being done by the caller to justify her complaint as reasonable.She responds to the way the word 'toe' is produced with a short laugh and overlaps with the caller's second utterance, the impact formulation [7].Her turn begins with a strong agreement 'of course' produced with rising intonation.This adverb is commonly used to emphasize agreement [29], and, since previous turns presented the caller's line of reasoning, we argue that it shows affiliation with the reasoning thus endorsing an evaluation of the complaint as a serious matter.The CH then reformulates the complaint in such a way that it acknowledges the hurt but generalises the blame: the CH explains that, in general, it is the fracture clinic's responsibility to deal with such injuries.With this response the CH again affiliates with the reasoning behind the complaint, and the caller corroborates their interpretation (ll.77-8).In this way the CH's response affiliates with the impact of the complained-of events and at the same time diffuses the blame made relevant by the caller's criticism of "the girl who phoned".
Sequentially, the CH's affiliative move achieves phase transition in a context of incipient repetition by the caller of their appeal to reason.The CH's affiliation with the impact of the complained-of event (ll.73-6) frees the complainant from the need to justify it further.Indeed, the caller produces her 'yeah' in line 78 with a final falling intonation, after which the CH projects transition to the next phase with a complaint formulation paraphrasing part of the complaint narrative relating to the referral letter from A&E to the fracture clinic. 7Thus, the CH uses affiliation to progress the conversation.
A little later in the call the caller re-launches her complaint about the 'abrupt' staff member in the fracture clinic, which the CH has so far subtly avoided endorsing.We return to the conversation in 1b a few lines after the previous extract.The caller is animating her conversation with the person from the fracture clinic, using reported speech to recount her insistence that there are no details about the patient's referred pain in the referral letter: Extract 1b.
argue that lines 79-81 are framed as "reporting back" the caller's reasoning (seen in lines 62-64) rather than agreeing with it -the framing with "okay if you just bear with me so.hh" marks the utterance as animating the caller's perspective by drawing attention to the CH's role in recording what the caller is saying.
The caller's criticism in line 102 receives only a minimal response 'mhmm yeah' with the CH typing during the gap of 1.1.The caller then pursues affiliation in line 105 with a 'smiley' tone and a sharply rising pitch, prosodically and lexically stressing that the behaviour of the staff member was unwarranted.Without pause she continues with a characterological formulation [1,30] of self and reiteration of the impact (l.107).Characterological formulations position oneself or others as a particular type of person (e.g.tolerant, obnoxious) to support an activity at hand (e.g.complaining, accusing).Here the caller positions herself as a reasonable person and the fracture clinic staff member as unreasonable, thus, making the blame component of the complaint more explicit.
Like the previous sequence, the CH overlaps with a formulation displaying comprehension (ll.108-9) produced with a rising intonation which again curtails a further repetition of the impact of the broken toe.She then reformulates another part of the complaint (ll.109-14) which this time shifts the focus from the rudeness of the fracture clinic member of staff to the confusing recommendation by A&E staff.The CH's response thus evades the blaming by affiliating with the reasonableness of the caller's evaluation of the communication between A&E and the fracture clinic.This version of the complaint receives strong agreement by the caller (l.116) in overlap, with markedly rising intonation and increased volume.In line 117, the CH acknowledges again the strangeness of the situation with a discourse marker 'yeah no' the first component of which ('yeah') serves to agree with the stance of the caller, while the second ('no') projects a negative assessment which is never fully verbalized.Previous studies showed that verbally incomplete utterances are used to delicately convey a negative stance [31][32][33].Thus, the CH in part sides with the caller's account of events but avoids explicitly blaming particular individuals, all the while affiliating strongly with the complaint's impact and validating the complainability by ratifying the caller's reasoning.
Sequentially, the affiliation again helps the CH progress the conversation, this time transitioning to service delivery (ll.118-19) after which the call moves to closure without any further re-launch of the complaint.Thus, affiliation displays not only help de-escalate the complaint by diffusing blame, but also serve progressivity through the phase structure of the call.

Absence of affiliation leading to escalation
The previous example illustrates how callers' displays of hurt, blame and reasoning make relevant affiliation with the stance embodied in those displays.In the following case study, we focus on the interactional outcomes where there is a lack of affiliation evidenced by the caller's pursuit.The context of this call is that the caller has unexpectedly ended up with a urinary catheter following a minor operation and needs a medical test before the catheter can be removed.Initially, the caller's concern is that he wants the test expedited and the catheter removed, so in contrast with our first case, there is no overt blaming in the early phase of the call.Extract 2a shows how the caller's initial narrative about the catheter displays the caller's negative stance through affective prosodic features such as recurring marked falling intonation and audible outbreaths, making accessible the complainability of the catheter: Extract 2a.
These mentions of the catheter are thus hearably produced as displays of hurt which imply responsibility of the medical staff and make relevant affiliative uptake by the CH.Nonetheless, they receive no or minimal uptake, resulting in silences at lines 27, 31 and 34 that we argue are oriented to unsuccessful pursuit of affiliation.In the next extract, the caller narrates a second unsuccessful attempt to remove the catheter: At the start of this extract, the caller explains, with similar prosodic stance marking to Extract 2a, that his catheter is still in place and a further test is required.Following minimal uptake by the CH (l.43), the caller makes more explicit the negative impact of the wait for the test, displaying negative stance through affective prosodic marking and the adverbials "in the meantime" and "still", which again make relevant affiliative uptake.Although the caller's turn is lexically and prosodically complete, there is initially no uptake and then the CH responds with the continuer 'mhmm', which displays an orientation to the caller's prior turn as not requiring a more substantive response and returns the conversation floor to the caller.The caller, however, does not initially take back the floor (l.48), displaying an orientation to the CH's uptake as insufficient.
With no elaboration from the CH, the caller pursues affiliation (ll.49-50) with the complainability of his circumstance by referencing the date for his test -'July' (two months after the call).The reported date functions as a negative assessment of the timeline of his care [34], which again accomplishes a subtle display of hurt and implicitly conveys blame, making relevant an affiliative second assessment. 8However, again the CH responds with a continuer.The silences after the CH's minimal responses (ll.44,48,53) suggest that the caller expects a more elaborate response to the waiting time to validate his complaint.
The caller responds by making explicit his dissatisfaction with the timeline and the personal impact it is having on him: Extract 2c.
With no uptake from the CH, the complaint expands incrementally, evidenced by the increasing gaps (ll.56, 58, 60) 9 followed by caller expansions which initially add details to the personal impact and ultimately introduce a new complainable, namely that the caller has not seen a doctor (l.61).Repeated gaps/minimal responses plus repetition/ expansion are a widespread pattern in our corpus which we argue is oriented to pursuit of affiliation in the context of a noticeable absence of CH recognition of the legitimacy of the complaint [20,26,35].We therefore argue that the expansions in this extract constitute pursuit of affiliation rather than simple continuation of the complaint narrative because of the emerging silences between caller's turns.
In response, the CH maintains an institutional footing and recurringly orients to the caller's complaint about not having seen a doctor as a factual matter: Extract 2d.
The CH, here, poses a question that implies that the caller has seen a 8 Waiting times are a recurring focus of complaints in our data but there is no evidence either in this case or in our wider corpus that complainants expect CHs to resolve the issue.Complainants do however typically pursue acknowledgement that their assessment of the waiting time as "too long" is reasonable. 9There is no audible typing during these silences.This data was collected between 2010 and 12 and we unfortunately have no evidence about what is happening during longer silences.
doctor, thus pushing back against this element of the caller's complaint and we see how the caller, while conceding that he was under some doctor by mentioning his probable name (which the CH confirms, l.6710 ), resists the implication of this fact-checking through referencing formulations that negatively evaluate the doctor: "I presume it's a doctor"; "the guy" (ll.65, 68).The CH's factual institutional focus and his lack of orientation to the caller's lifeworld concerns thus lead to an escalation in the complaint and the introduction of direct criticism of a specific individual (explicit blame).Again, the CH does not provide any uptake of either evaluative stance but, instead, moves to a new information gathering task (l.71).
In the next extract, we see a final attempt by the caller to pursue affiliation in response to this complainable: Extract 2e.
Line 124 shows a long silence following the CH minimal uptake of the caller's repetition of his concern about not seeing a doctor.As in extract 2b, this silence is recognisable as a noticeable absence of caller continuation following the CH's passing turn and is hearable as oriented to pursuit of uptake of the prior complainable, which the CH does not provide.Instead, we see an escalation in the pursuit of affiliation by the caller who explicates the reason of his concern at not being seen by a doctor through the heightened negative assessment of the emotional impact: "it's frightening".This escalated display of hurt may serve as an appeal to empathy (affiliation with an emotional impact of the complaint) but in this context, as with the displays of hurt in Extract 1, we argue that it also crucially functions as an appeal to reason -a bid to support the legitimacy of the complaint through the seriousness of the hurt. 11t this point, the CH begins to explain the role of Extended Nurse Practitioners (ll.127-8).His turn works as a normalising account, justifying the care by "specialised" nurses.Although this is likely an attempt by the CH to mitigate the blame, it expands on the implication that the complainable is not legitimate because the patient has been seen by qualified staff.Moreover, the explanation is produced in response to an assertion of heightened emotion, so it negotiates blame but displaces relevant uptake of the explicit expression of hurt.The caller displays an orientation to the CH's account as inapposite through the interruption in line 132 where his complaint moves to not knowing what is happening with his own body.This move extends both the hurt and the blame by further resisting the implication that he was provided with proper care.
The complaint has thus now escalated from implicit to explicit and emotionally heightened expressions of hurt, along with further implied accusations of inadequate care (blame).
Again, the CH doesn't respond to the evaluative stance expressed in the caller's interruption.
In this final extract which directly follows extract 2e, the caller returns to the initial presenting problem but here the lack of CH's uptake leads to a further increment in line 137 that escalates the blame component of the complaint through an implication of clinical negligence.This is met with a 1.0 silence and a minimal response "mhm" from the CH.In other words, the CH again misses the opportunity to acknowledge the negative impact of the complained-of events and show that the complaint is treated as serious and legitimate.This leads to a further escalation that builds on the implications of clinical negligence (ll.140-142).
This case study illustrates how repeated noticeable absence of CH affiliation with expressions of hurt, blame and reasoning results in an incremental escalation of the complaint.We see that what starts as a relatively practical problem for a caller ends up as a highly emotional complaint that encompasses the details of his daily life, as well as implied accusations of inappropriate care and even clinical negligence.Two observations can be made about how the escalation in this call is organised.Firstly, the escalation follows a series of incremental steps, from initial primarily prosodic displays of negative stance (2a) through lexically implied displays of hurt (2b) that are subsequently made explicit (2c), extensions of the scope and scale of the hurt (2e) and finally implicit orientations to escalating blame (2d, e f).Second, escalation follows silences in two distinct sequential positions: 1) immediately following the caller's displays of negative stance in a position where affiliative uptake is relevant and its absence therefore a notable absence and 2) following the CH's minimal responses where the absence of continuation by the caller displays an orientation to the CH's response as insufficient.These patterns in the incremental progress of escalation and the features indicating pursuit of affiliation hold across our broader corpus, albeit that the degree of escalation varies with the level of affiliation displayed by the CH both in individual responses and moment by moment as calls progress.

Discussion
This study has explored how the conversational moves of CHs may be consequential for the efficacy of complaint calls.Through a close analysis of recorded calls, we found that affiliation displayed at sequentially relevant moments helps progress the call and de-escalate the complaint, while relevantly absent or misplaced affiliation, by contrast, may lead to escalation [6].
There is a potential conflict between institutional goals and the heightened emotional valence of a healthcare complaints call [36][37][38].The CH's institutional focus results in caller displays of negative stance being neglected.Instead of an affiliative response, callers get either no response, a minimal acknowledgment token or a continuer such as 'mhmm' [cf.39].Complainants subsequently exhibit, for example through pausing, repetition of the impact of complained-of events or the addition of new details to the complaint, that they are pursuing a different type of response [40,41].Thus, missing affiliation recurringly accounts for the escalation of a complaint [cf.9].
Our study builds on and extends Pino's [7] account of the action formation of complaints firstly by examining how the hurt and blame components relate to conversational dynamics and outcomes and secondly, by making the case for appeal to reason as a call for ratification of complainability.Our analysis shows how attending to hurt and blame as distinct components of the action formation of complaining supports more detailed analysis of the organisation of escalation or de-escalation of a complaint.Specifically, our first case, where blame was explicit from the outset, illustrates how de-escalation includes practices of diffusing and generalizing blame [7,42].In this case, blame was generalized from an individual to more non-specific contextual factors, a practice that merits further investigation.Our second case, which did not initially involve overt blaming, exemplifies how escalation may result in blaming particular individuals, but that the precursor of explicit blaming was the inability of the CH to affiliate with the negative impact of the complainable, i.e. the hurt.In other words, where hurt goes unrecognized, both hurt and blame are likely to escalate.This matters because, in contrast to physician-patient interaction, where criticism of another doctor's decisions may help build trust between doctor and patient [43,44], in complaints calls, CH responses observably avoid affiliating with blame [6,45].In addition to extending the evidence base for hurt and blame as essential components of the action formation of complaining, our analysis also shows that, at least in this institutional setting, repeated displays of negative stance are oriented to pursuit of affiliation with the caller's evaluation of the reasonableness of their complaint rather than affective uptake.Moreover, we also showed that blame can be diffused by affiliating with the reasonableness of the evaluative stance implicit in the blaming.In other words, our analysis makes the case that affiliation with an appeal to reason is oriented to ratifying complainability.

Conclusion
Our analysis revealed three key functions of affiliation in complaints calls: it can ratify the reasonableness of the complaint and thus help move the conversation from one phase to the next, facilitating institutional goals as well as potentially de-escalating the complaint.

Practice implications
The Real Complaints project includes the development of training materials based on observations about effective communicative strategies for addressing complaints that successfully manage patient expectations.CHs should listen for cues from callers that they are seeking legitimization for their complaint and seek to provide affiliative responses.Early intervention in the form of affiliation to displays of hurt and appeals to reason may also obviate the need to explicitly negotiate ascriptions of blame which could otherwise present a conflict of interest for the CH.