STORYTELLING AND ADVICE: CONSTRUCTING THE LIVED EXPERIENCE OF EATING DISORDERS ONLINE

Online peer support groups encourage individuals to tell their stories and to find validation and emotional comfort when reading about the stories of others. Coincidently, lived experiences are the kind of knowledge applied to solicit and to deliver peer advice. This study examines the relationship between storytelling and advice in an English speaking online forum that provides support for those with an eating disorder (ED). The results revealed a range of different types of narratives within the data, from more elaborate testimonials of the ED and the process of recovery to brief personal passages responding to the first poster. The Labovian narrative structure appeared in a number of the first stories, whereas two main configurations, contingent upon the kind of response offered, emerged in second stories: parallel assessments (or snapshots) and success stories. Parallel assessments constituted self-centred stories and did not include any advice provision. Success stories, instead, became an essential component of the advice-giving act since they were remedial. The solu-tion proposed by responders to the problem posed by the first poster was organized either to offer tips (that is, a series of practical recommendations to address a specific ED or recovery issue) or to deliver thoughtful advice through a resolutive story that introduced the state of recovery as a real possibility. Both parallel assessments and resolutive stories included contrasting resonances in relation to the first story. Resolutive stories encompassed resonating elements whose meanings were transformed and (re)signified from the positioning of a subject moving towards recovery. However, snapshots echoed specific key expressions from the initiating post. The goal was to display alignment with the first teller by describing a similar I-perspective

Online peer support groups encourage individuals to tell their stories and to find validation and emotional comfort when reading about the stories of others. Coincidently, lived experiences are the kind of knowledge applied to solicit and to deliver peer advice. This study examines the relationship between storytelling and advice in an English speaking online forum that provides support for those with an eating disorder (ED). The results revealed a range of different types of narratives within the data, from more elaborate testimonials of the ED and the process of recovery to brief personal passages responding to the first poster. The Labovian narrative structure appeared in a number of the first stories, whereas two main configurations, contingent upon the kind of response offered, emerged in second stories: parallel assessments (or snapshots) and success stories. Parallel assessments constituted self-centred stories and did not include any advice provision. Success stories, instead, became an essential component of the advice-giving act since they were remedial. The solution proposed by responders to the problem posed by the first poster was organized either to offer tips (that is, a series of practical recommendations to address a specific ED or recovery issue) or to deliver thoughtful advice through a resolutive story that introduced the state of recovery as a real possibility. Both parallel assessments and resolutive stories included contrasting resonances in relation to the first story. Resolutive stories encompassed resonating elements whose meanings were transformed and (re)signified from the positioning of a subject moving towards recovery. However, snapshots echoed specific key expressions from the initiating post. The goal was to display alignment with the first teller by describing a similar I-perspective experience. Taken together, the individual small stories contributed to the coconstruction of a multiple-lived story with regard to the ED in the online community.

Introduction
Internet-based resources afford individuals with health conditions with the means to tell their stories, as well as to find support and emotional comfort when reading about those of others (Walters et al. 2016). These resources have implications on how the peer-to-peer activity unfolds in online support groups, while shaping new politics of health and illness on the Internet (cf. Akrich 2010). By writing and posting stories that position the subject vis-à-vis the illness, users of online support communities become cognizant of the significance and meaning of her/his own personal circumstances (Widdershoven 1993: 7). Online illness narratives are a naturally occurring phenomenon, as opposed to the accounts elicited in clinical contexts. They are unsolicited, which makes them unconstrained by the requirements, language, and protocols of clinical practice. The subject's experience can be communicated from a more unfettered perspective (O'Brien and Clark 2012), a situation that induces the production and exchange of personal stories to negotiate and to construct identities (Stommel 2009;Armstrong et al. 2011).
In the domain of EDs, websites and online forums provide the stage upon which to make public the users' experiences with illness and recovery (McCormack 2010). Writing personal stories represents an exercise of personal interpretation that helps others in the online community to gain an insight into their own realities with respect to the condition (Walters et al. 2016). One strategy to convey these understandings is resorting to "second stories" (also known as "response stories"; cf. Norrick 2000). With this term, Sacks (1992) referred to the conversational activity of replying to a previous story with another story. The first story is shaped and mapped onto the responding story, since "tellers provide story recipients with interpretive templates that they use to monitor the events being reported […] prospectively in order to locate when the story arrives at its climax, the place where recipients are expected to provide a response to it" (Goodwin 2002: 27). This interrelation between first and second stories, together with the reciprocity that is often displayed in collective storytelling, depends upon the practice of stance-taking, which means adopting a specific point of view publicly (Simoraa 2012;Kääntä and Lehtinen 2016). Stance-taking is both a linguistic and a social act, and it always involves an evaluation, whether explicit or implicit (inferred) (Du Bois 2007).
Stance-taking explains why responses to the illness narratives shared in online peer support groups seem to be shaped by the practical and emotional frames invoked in the first stories (cf. Sandaunet 2008). Despite this connection, response stories are not just concerned with matching experiences, but with providing alternative contexts in which the issue, concern or problem formulated in the first story might be better understood. For that reason, in some cases the second teller takes a similar stance to that adopted by the first, whereas in many others she/he might opt to favour the opposite stance (Page 2012). Accordingly, second stories have the potential to perform face-enhancing relational work (Arminen 2004;Harrison and Barlow 2009), but they might also achieve face-threatening work, such as criticizing or expressing open disagreement (Veen et al. 2010).
The relational work performed in online support groups is the key to the activity of exchanging advice. According to DeCapua and Dunham (1993: 519), advising is an interactive process by which "opinions or counsel given by people who perceive themselves as knowledgeable, and/or who the advice seeker may think are credible, trustworthy and reliable" are presented. Expert advice has been widely explored in face-to-face institutional settings or in problem pages dedicated to providing professional guidance in magazines (e.g. Silverman 1997;Locher 2006;Locher and Hoffmann 2006). In recent years, however, there has been an increasing interest in how peer advice is produced and delivered online (cf. Morrow 2006;Vayreda and Antaki 2009;Kouper 2010;Veen et al. 2010;Sillence 2013;Stommel and Lamerichs 2014;Figueras 2020Figueras , 2021. The proliferation of peer support groups in which members exchange advice on health and illness has raised issues with the concept of experiential knowledge, as opposed, or complementary, to professional, socially sanctioned knowledge (Blume 2017;Faulkner 2017;Noorani et al. 2019).
The term "experiential knowledge", initially introduced by Borkman (1976: 446), is defined as the "truth learned from personal experience with a phenomenon rather than truth acquired by discursive reasoning, observation, or reflection on information provided by others". Experiential knowledge is holistic, comprehensive, and total, instead of fragmented and specialized in concrete domains (as is the case with professional knowledge). It is pragmatic, subjective, and oriented to the here and now of the lived experience. This more emotional, contextual, and unconscious understanding of personal issues has become the main source of "true information" for participants in self-help groups, in contrast to the kind of assistance found in patient groups led by professionals (Borkman 1976;Blume 2017).
One of the ways in which experiential knowledge is interactionally used to exchange information and provide emotional assistance in online health peer support platforms is through storytelling. It is common for participants who position themselves as advice-seekers to resort to personal stories to pose a problem through relating the background facts that justify the speech act of soliciting assistance (Kouper 2010;Page 2012;Sillence 2013;Thurnherr et al. 2016;Lindholm 2018Lindholm , 2019. Advicegivers, in turn, often play the role of "peer expert" by sharing personal stories that "encapsulate" and substantiate the advice being offered, in lieu of giving direct instructions to the recipient (Kouper 2010;Sillence 2013;Figueras 2021).
The important connection between narratives and peer advice has been examined in several recent studies (e.g. Morrow 2006Morrow , 2012Harrison and Barlow 2009;Kouper 2010;Sillence 2010Sillence , 2016Page 2012;Placencia 2012;Lindholm 2018Lindholm , 2019. As the analysis by Giles and Newbold (2013) indicates, second stories are an effective strategy when offering advice in mental health discussion forums, since experiences are better received than direct recommendations. Furthermore, response stories induce the empathy displays that sustain the advice-giving act realized in peer support groups (Figueras 2020(Figueras , 2021. Against this background, my aim in the present study is to contribute to the research on storytelling in online mental health communities from the perspective of discourse analysis. Specifically, my interest is in the types and functions that response (second) stories in those sites fulfill, and the kind of correspondences online communication draws between the first and second stories. I focus my analysis on the interaction in an online forum for recovery from an ED. The reason behind the choice of this platform is that it provides an invaluable source of accessible data, thus facilitating an examination of the interactive construction of the ED experience. As Jowett (2015) argues, mental health support groups are realistic settings, but without the presence of a researcher who might distort the kind of discourse created. Among the range of support communities, online discussion forums stand out as virtual spaces in which to question, debate, contest, co-construe and deconstrue the social and individual meanings, and also the boundaries, of health and illness. The exchange of information, affective support and practical directions facilitates the disclosure of personal stories about mental or physical conditions that have altered a person's life. Because of these features, mental health support groups "have proven particularly useful for critical psychologists wishing to examine naturalistic data on a wide range of social phenomena" (Jowett 2015: 287).
To conduct a qualitative analysis of the stories made public in the ED forum, I adopt a discourse-analytic approach. I examine the narrative passages in the dataset created for this project in the light of the following research questions: 1. What kind of stories are told and shared in an ED peer support group? 2. How are stories and advice combined in peer-to-peer interactions? 3. What kind of parallelisms and resonances can be found in first-second story pairs in an ED peer support group?
In the next section the decisions regarding the methodology are explained.

Corpus
The corpus for this study included 80 threads (91,484 words), posted between March and August in 2021, from an English-speaking online discussion forum that provides peer assistance for those recovering from an ED. The ED forum belongs to an extensive social networking platform with different topical moderated forums that has become increasingly popular with young adults. The discussion threads were selected after exploring the site in order to identify opening messages with more than five responses that explicitly posed a problem (troubles-telling), requested advice, or asked for information. The first posts in the sample usually introduced a storied account of the events that led the subject to make a request for emotional or informational assistance from the audience. Responses to these requests often incorporated personal experiences prompted by the first story. These intimate accounts were deeply ingrained in the interactional dynamics in which they were produced, so that key correspondences could be established between the stories inserted in the initiating messages and those provided in the responses.

Ethical considerations
Regarding the ethical issues raised by the present study, we adhered to the heuristic principles outlined by the Association of Internet Researchers (Markham et al. 2012). The site is public, so anyone can access the repository of threads and posts and read the message boards online. However, visitors are required to register and log in to publish posts. Recommendations established by the University of Barcelona's Bioethics Commission (CBUB) regarding online data were followed, specifically, those relating to preserving user anonymity by concealing the name of the site and by removing individual identifiers, in addition to details that might identify individuals or organizations. Ethical approval was granted by the CBUB.

Analysis
A qualitative study was conducted on the sample, that is the 80 first posts and their corresponding responses (510 messages). The approach adopted was the "digital" conversation analysis (CA) discussed in Giles et al. (2015), and developed in both Meredith (2017Meredith ( , 2019 and Mondada (2019). The basis for this method is the applicability of the CA sequential organization to asynchronous computer mediated communication, or CMC (e.g. Vayreda and Antaki 2009;Stommel and Meijman 2010). CMC messages or posts are thought to be sequentially related and, hence, the interaction in the ED forum is understood as participants taking turns asynchronically.
To conduct the analysis, I initially searched for those turns in which storied experience talk was embedded. Storied experience talk was defined as the (re)construction of personal events related to living with an ED and/or recovering from it. Narrating experiences is a contextualized activity arising from specific occasions for telling (Ochs and Capps 2001;Georgakopoulou 2007), such as responding to a first story in the ED forum. For the present study, a story was defined as the re-establishment of a lived experience according to the narrativity components listed by Herman (2009: 22): 1) situatedness, 2) event sequencing, 3) worldmaking/world disruption, and 4) what it's like (experientiality). These four elements are placed on a gradient that results in a broad spectrum of narrative schemas, from more to less prototypical (cf. Ochs and Capps 2001).
At this stage of the research, I placed the storied experience talk identified in the sample on a continuum of narrativity, with clear instances of narrative accounts at one extreme and non-narrative accounts at the other. To do so, following Kääntä and Lethtinen (2016), I considered the following parameters: whether the experience talk displayed the prototypical three-stage narrative design (orientation, narrative, and evaluation); whether there was a temporal change in the sequence of events; whether a single course of events or a recurrent one was reported; the organization of the sequence of events, from past discrete and punctual events to more atypical stories with repeated or habitual events, either in the past or in the present; how agency and personal voice were produced in the telling; and whether the responder was focusing only on her/his own experience, or she/he proceeded to (re)interpret the first poster's account.
The identification and the analysis of the recurrent experience talks in the sample were also conducted within the framework of small stories research. Initially developed by Alexandra Georgakopoulou and Michael Bamberg, this approach represented a new paradigm to analyze narratives and identities (e.g. Bamberg 2006;Georgakopoulou 2007;Bamberg and Georgakopoulou 2008). It questioned prevalent models of narrative studies (e.g. Labov 1972) that characterized narratives solely on the basis of textual parameters, and instead favoured one specific kind of narrative, that is the almost monological, long-storied account of past events in somebody's life, usually in the context of an interview (Georgakopoulou 2017a). Small stories, in contrast, incorporate "a gamut of under-represented narrative activities, such as tellings of ongoing activities, future or hypothetical events, shared (known) events, but also allusions to (previous) tellings, deferrals of tellings and refusals to tell" (De Fina and Georgakopoulou 2012: 116). Small stories are flexible, dynamic, open-ended, and characterized by a multiplicity of selves. Different kinds of narratives fit into the category of small stories: "stories to be told", "breaking news", "projections", and "shared stories" (Georgakopoulou 2007).
The present study extends narrative analysis to examine an array of experiential stories that are worked and reworked through the interactional activity of a site designed for discussion and support. My aim is to explore the features of those narratives, what role(s) they play in the communicative dynamics of the forum, and what correspondences, with either alignment or disalignment, can be identified in the first-second story pairs exchanged among the interactants. In the results section representative exemplars of narratives identified in the sample are subjected to a finegrained, small-scale, and local qualitative analysis.

Degrees of narrativity
Personal stories are a fundamental part of the interactional activity in the ED community. Both first posts and their subsequent responses are likely to contain different kinds of stories, ranging from a full sequence of past events that yields one or more complicating actions to small stories. Therefore, the narrative passages in our data exhibit different degrees of narrativity (Thurnherr et al. 2016). First posts may present full-fledged accounts with many of the Labovian narrative components; 1 namely, abstract, orientation, complication, resolution, and coda. This pattern, exemplified in (1), 2 could be found in sixteen first posts. 1 Evaluation is usually embedded in the telling and does not constitute a separate category (cf. Lindholm 2019). 2 The excerpts reproduced in the present study are verbatim and have not been corrected for grammar or spelling errors. Bold type is used in the text to highlight certain structures in the passages quoted.
( In (1), the teller opens her/his story with an abstract that summarizes and evaluates the content of what she/he is going to disclose. The abstract announces the story and provides the frame so that the responders can understand the reported ED episodes. The poster makes clear to the potential audience that the account that follows is a personal experience and that she/he is looking for validation. Next, the orientation move introduces the relevant information about the background (scenario, characters and temporal-spatial-actional context) against which the illness developed. Following the orientation, the complication action (an essential component of the Labovian model) incorporates the climax or high point of the story. In the post in (1), the complication action subsumes the sequence of events that ensued after the onset of the ED. The whole lived experience recounted in (1) is discursively divided into two phases or episodes. The first is the anorexia period, characterized by behaviours such as restricting and overexercising, whereas the second ED phase is defined by other practices (binging, in particular). Two different resolutions, corresponding to each phase of the ED lived experience, are part of the whole illness story told in (1): the first is the "unusual" restoring of "normal behaviours" following the restrictive-anorexic phase (the poster does not specify what she/he means by "normal" as far as her/his behaviour is concerned); the second is the resulting weight gain due to the cycles of binge eating. This second resolution does not entail the termination of the negative events listed in the complication phase, but it brings about the undesirable current situation that is described in the coda. The coda acts as the bridge between the past and the present: "[I] am now obese". These two extremes of the ED experience (being "extremely underweight" in the past vs being "obese" in the present) leave the poster with a feeling of confusion. She/he wonders how to mentally conciliate two behaviours that she/he conceives as being at odds with each other: restricting vs binging. This conflict is the basis for the advice-seeking act performed at the end of the post.
In contrast to these more structured narratives, most of the personal stories rendered by forum members can be considered small stories, in the sense that, although they contain some disruptive event in the process of living with and recovering from an ED, they are minimal. Small stories might implicate a temporal change, but not a fully developed story-like plot (Kääntä and Lehtinen 2016). The interactive nature of the forum forces tellers to be more flexible when designing the form and content of their stories, particularly if their goal is to provide informational and emotional support to the first poster by sharing their daily struggles with the condition. Thus, rather than eliciting monological narratives (more typical in a structured interview setting), the interactional dynamics of the forum induces the production of collaborative narratives (cf. De Fina and Georgakopoulou 2012). Meanings of the ED are discussed, evaluated, and disputed, while difficult episodes in the different phases of recovery are discursively construed as shared experiences (Norrick 2000).
In the context of the ED forum, first posters who disclose personal stories pursue one of two main interactional goals: they either seek advice to cope with the ED and the journey to recovery; or they look for someone else with a similar experience, to normalize and validate certain thoughts, feelings, and behaviours with which they are dealing. Thus, in 19 first posts (23.7% of the sample; N=80) the writer asked directly for a tip or recommendation ("Do you guys have any advice?"; "Really need your advice"; "Any advice on moderation?"), whereas in 21 first posts (26.2% of the sample; N=80) the person is searching for validation and normalization of their emotions and sensations ("Does anyone know why I feel this all the sudden…?"; "Does anyone else have 'safe clothing'?"; "idk if anyone will relate").
In both cases, ED first stories are typically problem stories (troubles-tellings). According to Lindholm (2019), the focal point of problem stories is a continuing and unresolved situation that impacts negatively upon the narrator's existence. The circumstances of this situation are then discussed and debated with other participants in the ED forum, who offer response stories (second stories), thus reacting to the first account. Response stories tend to share some similarities with the first story, such as the topic, the theme, and the roles of the participants (Lindholm 2019).
The responder might construe her/his own story as a plausible solution to the problem posed by the first poster. In this case we talk of "success stories" (Lindholm 2019). Success stories are designed to deliver certain remedies that tackle the health issue raised in the first place. The recommendations offered within these second stories are usually portrayed as useful, effective, and doable practices that worked for the responder (Norrick 2000). In the following exchange, the first poster prepares her/his request for assistance by summarizing a recent experience, namely hair loss, that is a side effect of her/his ED.
(2) First post I began recovery in January, recently I did relapse, however my doctor has put me back on track for the past week now and I have been eating quite a good amount for me. But, for the past couple of weeks, my hair has started falling out really badly. It did fall out before but not at the rate it is now. Everyday I seem to be losing more and more hair, and it's already clearly visible that my hair has thinned. All my family and friends have noticed. I had a breakdown today after a brushing it because so much hair came out. I don't want this to carry on happening, it's really scaring me (it's actually scaring me back into recovery tbh) Is there anything more I can do to stop this happening? How much longer will it continue? If you have any hair regrowth stories, it would be much appreciated. In (2), the author details the most recent events and main facts surrounding her/his current health problem, and frames them in the context of getting "back on track" in order to recover from her/his ED. Her/his feelings in relation to this experience are part of the evaluation performed at the end of the first paragraph when she/he describes a recent breakdown and confesses to fearing the progressive loss of her/his hair. This small story sets the stage to formulate the solicitation of assistance in the second paragraph. The advice seeking act is realized with two interrelated moves: first, requesting specific information and tips on how to manage the situation; and second, opening the floor to receive others' personal stories. This post makes explicit the recognition that forum members grant to the experiences of others. Expertise, in this context, is based on first-hand, experiential knowledge, and conveyed via narratives. The weight of experiential knowledge is the justification for the advice-seeker in (2) requesting that other users share their "hair regrowth stories". Compared to the ED account reproduced in (1), that in (2) does not develop all the phases in a high-point structure narrative. Instead it features a condensed summary of events pertaining to a current episode of hair loss. The contrast between post (1) and post (2) reveals some of the variations in narrative patterns that can be found in online forums. As several researchers have noted, narratives present different configurations and arrangements across distinctive computer mediated commu-nication modes (see Georgakopoulou 2004, for email; Arendholz 2010, for message boards; Page 2010and Georgakopoulou 2013Dayter 2015, for Twitter). This diversity is the result of differences in the spatial and temporal constraints of each medium. In relation, specifically, to discussion forums, the asynchronicity that characterizes the exchanges on those platforms affords the interactants the space and time to compose lengthier and more reflective messages (Lindholm 2019). First posts published by newcomers in mental health forums tend to contain some storied account to contextualize the request for advice and/or information (Figueras 2021).
Conversely, response (or second) stories in the ED forum usually take the form of "snapshots". Snapshots, adapting Lindholm's (2019) definition, are small stories that either focus on a specific past event, directly related to the problem, question or concern raised in the first post, or describe habitual states of affairs. In (3), the responder's advice is preceded by short sentences describing past events in her/his personal experience of hair re-growth. Each single past event is followed by displays of encouragement and emotional support directed to the recipient. A series of moves thus precedes the assistance being offered in the form of tips in (3). 3 The personal experience recounted in this response is a success story. It illuminates the ways in which condensed accounts of past events that mirror those in the first story, retell the experience of certain common ED symptoms (such as hair loss).

Advice and narratives
Personal stories are the fabric of the peer-to-peer interaction in online support communities addressing health issues, so the discursive construction of stance-taking via narratives constitutes a very effective strategy in order to create different identities within the ED forum. Narratives provide the structure to produce the self discursively (Thurnherr et al. 2016), according to a constructionist approach to identity (De Fina 2003;Hall 2005, 2010;Hall and Bucholtz 2013). Likewise, narrative stance-taking mobilizes "social indexicality" (Georgakopoulou 2017b), which means that, in the context of advice, the process of requesting and providing recommendations requires the participants to index themselves as help-seekers or helpgivers. These identities are the result of relational work (Locher 2015). While experience talk has to do with the self (Kääntä and Lehtinen 2016), advice is connected to expertise. Hence, advice and personal stories are often entwined. This association explains why single speech acts of advice provision might be interspersed throughout a response story (Lindholm 2019), as is the case in the reply in (4b) to the troublestelling reproduced in (4a): [This is really common bc weight gain is often the fear that sparks an ED in the first place, so gaining weight can be incredibly triggering and is often why relapses oc- [I found positive affirmations daily really really helpful. I have been doing those and daily gratitude every day for the past 6 months. While I still struggle with anxious and depressive emotions, I no longer have negative self talk. It's so relieving. Everyone has something different that helps them, but positive affirmations can be really powerful with negative self talk. Your brain learns thought patterns, and they repeat. By doing positive affirmations, you're creating a new pathway in your brain for new thought patterns. It feels really silly at first doing them, but i was really suffering from my thoughts and the affirmations changed the dialogue in my head. The other thing i've done is challenged my negative thoughts. Instead of just accepting them, i practice a form of mindfulness where i observe them, and i challenge them. Instead of just accepting that "i'm a POS"(example of negative thought) I tell myself reasons i'm not and remind myself that all humans have dark and light parts to them. Everyone has things they're not proud of and things they

VIOR (NEW PRACTICES IN RECOVERY: POSITIVE AFFIRMATIONS, MINDFULNESS)
[I'm wishing you the best. It sounds like you're really struggling and I am so sorry for what you're going through.] EXPRESSING SYMPATHY The author of (4a) focuses the discussion on one of the first and most important steps on the path to recovery from anorexia: eating more to gain weight. While weight gain is construed as a milestone in the healing process, the poster makes explicit the emotional toll that increasing food intake is taking on her/his mental state. She/he acknowledges three types of effects caused by weight restoration: the negative moods of anxiety and depression that she/he is now experiencing; the resumed practice of self-harm; and the flood of negative thoughts and "self-talk". This succinct description of her/his current state of mind prepares the speech act of advice solicitation, which is realized indirectly. The advisee poses an open question to the audience ("Have other people experienced this?"), and she/he asks for some useful tools to overcome ED self-destructive emotions, practices, and thoughts. As this example reveals, the information that is most valued in the ED forum is the experience-based knowledge, as opposed to the factual knowledge owned by clinicians and medical professionals. The response in (4b) reframes the specific ED problem addressed in the first post and takes a different (epistemic) stance to substantiate the directions being offered. The main purpose of this response is to deliver advice, a task that is carried out by resorting to two different tactics, one direct and the other indirect. The direct strategy incorporates two mitigated speech acts of recommendation: "I would honor the tiredness just like your hunger" and "If you can, maybe avoid looking in mirrors and wear baggy clothes for a while". The indirect method involves presenting the advisor's own techniques to challenge the characteristic ED "negative self-talk". The advisor's first-hand experience with certain psychological tools is used to resituate the recipient's positioning towards the ED via cognitive restructuring; that is, by questioning the "distorted" thoughts triggered by the illness, a process that might eventually lead to a modification of the sufferer's beliefs. By sharing her/his personal mental exercises the advice provider aims to indirectly model the recipient's ED cognitions and to guide her/him in the transition to recovery. These cognitive practices are of two kinds: positive affirmations ("I have been doing those and daily gratitude every day for the past 6 months") and mindfulness ("I practice a form of mindfulness where I observe them, and I challenge them. "). To show how this approach works, the responder provides an example of a typical negative thought ("I'm a POS. ") and the kind of reasoning that she/he applies to challenge and to dismiss it: "I tell myself reasons I'm not and remind myself that all humans have dark and light parts to them".
The comprehensive assistance offered in (4b) is carefully devised and structured to guarantee its successful reception. The advisor initiates her/his response with an explicit recognition and positive evaluation of the efforts made by the recipient to recover. This display of emotional empathy (Figueras 2023) is then followed by a general statement about the comorbidity in EDs that is validated with information concerning the advisor's personal background with mental illness: "I have CPTSD, PDD and anxiety. My ED developed alongside my CPTSD, as it was a form of control during the trauma i was experiencing". This revelation is used to claim legitimate knowledge to deliver a diagnostic opinion addressing the unresolved ED issues posed by the advisee. The second and the third paragraphs contain pieces of advice that are tactfully crafted and administered. The pattern adopted in both paragraphs is the following: GENERAL STATEMENT ABOUT RECOVERY + MITI-GATED ADVICE + REASONS TO SUPPORT THE ADVICE. These tips are explicitly directed to the recipient (as the use of the second person singular reveals), whereas the mental practices described in the fourth paragraph switch the focus from the recipient towards the emitter and her/his personal dealings with the ED symptoms. The advisor's own experience is construed in terms of a habitual story in the present tense.
According to Lindholm (2019), habitual stories are non-canonical narratives that encompass repeated activities realized by the teller. They have been considered as nonnarrative structures because they refer to "general events which have occurred an indefinite number of times" (Labov 1972: 361). Yet, such stories are a salient narrative type in the advice-givers' repertoire of responses in the ED forum. These habitual accounts of activities highlight the impact of repeated events in the narrator's life, while showcasing the normality or abnormality of certain incidents and states of affairs. The daily routine is purported as advice. Due to their routinizing nature, therefore, habitual stories could be conceived as a kind of procedural narrative (adopting the term coined by Trunk and Abrams 2009): they incorporate a string of regular events that may take place in the past or in the present. The account of the daily mental exercises performed by the responder becomes the roadmap for the advice delivered in the fourth paragraph of (4b). The description of the practices is interwoven with the evaluation of the effects that positive affirmations and mindfulness have brought to the emitter's life.
The different pieces of advice provided by the responder in (4b) are combined with a report of the personal tribulations experienced due to the illness (see Lindholm 2018). Narratives qualify advice providers as having the epistemic superiority and moral authority to influence and to achieve specific intended interpersonal effects (Thurnherr et al. 2016), such as impacting and modifying the recipient's state of mind (Figueras 2021). When the advice is enmeshed in the personal story of the ED, as in (4b), the emitter positions herself/himself as a credible and trustworthy advicegiver whose credentials are granted by first-hand experiential knowledge. The identity indexed with narrative passages delivering advice is of a peer expert who holds a certain epistemic superiority over the recipient. The emotional connection between both forum members is sealed at the end of the message, when the responder expresses genuine understanding of the psychological troubles faced by the recipient. This move represents a highly involved talk conducive to enhance rapport and solidarity within the group.
Both the first poster and the responder resort in the exchange in (4a-b) to habitual stories in the present tense, bringing to the fore the painful quandaries of recovery. These ordinary accounts represent small stories about recent or still unfolding events of the ED experience (cf. Bamberg 2006;Georgakopoulou 2007;Bamberg and Georgakopoulou 2008). It is, precisely, the common nature of the ED symptomatology that makes these narratives the product of high audience involvement; in other words, the participants cooperate with "substantial narrative contributions" to the telling of the story in the first post (Ochs and Capps 2001: 26-27). In doing so, personal stories become choral narratives, that is, projects of collaboration and participation that include all the participants in the ED forum. In truth, habitual stories play a key role in the concerted effort to make sense of the condition between individuals in a comparable situation. 4 The similarities and differences in the individual lived accounts of the illness can be determined by looking at the different ways a second story resonates in relation to the first. In this regard, and as Norrick (2000: 125) acknowledges, participants "routinely align themselves through matching their response stories with foregoing ones", an interactional operation that is examined in the next section.

Parallelisms and resonances between first and second stories
Often, the response stories in the sample echo the event or ED episode described in the initial post. They do so by depicting the successive respondents in parallel situations, conducting identical actions, suffering from and expressing comparable feelings, as well as rendering the same evaluation. This is observed in the messages in (6a-c), with which forum users reply to the author of (5): (5) First post My parents are watching really closely what i eat. They sometimes ask what i've eaten today and all of this. They do not know about my eating disorder but i guess they suspect it. So my father started making lunch everyday (he is in homeoffice) so i have to eat at least 2 meals a day. My family eats together for dinner because this is the only time where everyone is at home. Usually there was no lunch so dinner would be my only meal (with sometimes breakfast if my parents were watching) but now i have to eat two times a day which makes me really upset. So i started finding excuses like "i've just eaten something" or "i'm in a video lesson rn i'll eat later". Ofc i couldn't say this everyday so I always skip breakfast. To make it less suspicious i always ask my dad whether he would want some porridge too bc i was making some and then i would make porridge give it all to him except for a little spoon bc i would need that to make my bowl look used. When my dad doesn't want any porridge i just cook like a very little amount and then throw it away. I feel bad for wasting the food. Also when i eat I can't enjoy it i don't want these calories i hate it so it's wasted because other people would be really happy about it and i'm not. I just want to give all my food to them.
(6) Responses (a) I used to do this around my family and then I moved out and live with roommates now. I buy groceries then throw them away at the pace I'm supposed to be eating them so they don't get weirded out I don't have any food. Ana 5 is really wasting my OWN money and food now smh :( (b) omg super same my dad also makes breakfast and i just throw it away because i hate eating stuff, i feel bad too because he made it for me. just end up skipping the days meals of having low cal stuff. it do really feel guulty smh. daddy i love you (c) I relate. I'm in recovery and I have to throw away things when they turn away.
I feel bad but my ED is too strong The core theme of the troubles-telling in (5) is the narrator's routinary efforts to skip breakfast without being noticed. The repeated action of throwing food away to avoid eating is contextualized within the framework of an ED ("They do not know about my eating disorder but i guess they suspect it"). In this scenario, the detailed account of the concrete maneuvers to achieve the goal of not eating becomes the foundation for the habitual story detailed in (5). In this case, the poster does not explicitly ask for advice, but rather calls for sympathy and alignment from the rest of the members of the group. Hence, the replies are in line with the needs of the first poster (cf. Riccioni et al. 2014). Instances of those responses are the messages reproduced in (6a-c).
The fact that some of the ED members reply just with self-references (e.g. "I used to do this"; "my dad also makes breakfast and I just throw it away"; "I relate") and second stories that lack any (re)elaboration of the original poster's experience indicates the narrow focus of their reactions: instead of taking the other-perspective, responders focus on their own embodied experience (Figueras 2023). Yet, the act of advice-giving critically depends upon the advisor's capacity to take an "imagine-self " perspective (imagining how one would personally feel about a situation), and/or an "imagine other" perspective (imagining what the target of a story might be feeling or experiencing; Batson 2011). An elaborate act of advice provision requires projecting oneself into another's situation (Figueras 2020(Figueras , 2021. To do so, the advisor must be able to mentally picture two situations: that lived and experienced by the recipient, and that in which her/his troubles would be resolved or, at least, minimized. Usually, this operation requires contrasting the positionings of an ED sufferer with those of an ED individual who is on the road to recovery or has recovered. The authors of (6a-c) do not balance these two positionings. Instead, they confirm and consolidate the hardships of their EDs. The strategy here is to acknowledge their personal daily struggles with a similar ED behaviour as a display of alignment with the poster of (5); that is, they openly affiliate with her/him by endorsing her/his perspective. Their alignment is based on their shared knowledge about the ED. In a similar fashion, the messages in (6a-c) replicate the moral and emotional evaluation of the ED practice laid bare in (5), since successive tellers also embrace the initiator's evaluative and moral stance toward the condition. Thus, all the interactants regret wasting food and/or deceiving their loved ones. All of them deem this behaviour as morally reprehensible. These responses are meant to warrant those behaviours as ED symptoms, and to validate the shared and communal experience of the ED. Therefore, part of the conversational dynamics in the ED forum reproduces the structure of a troubles-telling followed by parallel assessments (Heritage 2011). This type of interaction (first part: troubles-telling; second part: parallel assessments) channels the emotional need to bond with others in a similar situation within the ED community. Members of the support group resort to parallel assessments when they seek to relate, to reassure the legitimacy of their struggles, and to experience a sense of belonging.
A comparison between the responses in (4b) and (6a-c) highlights the differences in the way second stories resonate with the first. Whereas the poster of (4b) owns a story of ongoing change and self-transformation to move away from the ED symptomatology, the parallel assessments in (6a-c) concretize the resounding effects induced by the first post in some members of the audience. In the former case, we find a resolutive second story, that is, a responder's account that draws on the key components of the first story to (re)create it within the format of a successful recovering story (Figueras 2023). This kind of response, containing a resolutive story, was found in 41 second posts (8% of the replies in the sample; N=510). By contrast, the reactions conveyed in (6a-c) are meant to corroborate the factualness of the initiator's experience while publicly contributing in order to co-construct a collaborative story of the everyday markers of the illness. These parallel assessments became the preferred response in 101 cases (19.8% of the responses; N=510). Both resolutive stories and parallel assessments are second stories that resonate with first stories, 6 but they are modelled and shaped in different ways with regard to the first poster's stance. A revealing aspect of how response stories resonate is the interactive practice of recycling key lexical, semantic, and syntactic material from the first to the second story (Kärkkäinen 2006;Simoraa 2012).
As Simoraa (2012) contends, these resonating elements forge cohesion between the first and the second story. Indeed, the response story in (4b) is anchored in that told in (4a), in such a way that the first narrative sets the parameters that determine the relevance of the second. The two stories stand as bounded domains of experience that are tightly connected thanks to their resonating elements. Concretely, the first account introduces a stance-taking that is then used as a reference point for responders to build their own positionings (Simoraa 2012). The nature of the linguistic 6 "Dialogic resonance" is defined by Du Bois (2014: 359) as "the catalytic activation of affinities across utterances" and between comparable linguistic elements at any linguistic level (see Kärkkäinen 2006;Du Bois and Giora 2014). Simoraa (2012) adopts the notion of resonance to account for the analysis of stance-taking in consecutive stories organized in a first-second pair. All these studies, however, have focused on lexical-syntactic, semantic, and prosodic resonance in conversational face-to-face interaction. In the present study, I am concerned with the resonance of certain lexical expressions between the first and subsequent posts in the ED online forum, and how these resonating elements reveal distinctive stance-taking positionings.
parallelisms between these individual accounts gives the researcher an insight into the dialogical and intersubjective nature of the ED storytelling in the forum. This joint account is explicated via communal evaluation and stance-taking. Participants in the ED forum recontextualize and redefine their own accounts and others' stories to construe a collective understanding of what the phenomena of EDs and recovery mean in that community. The comprehensive term "recovery" is a resonating element that is reinterpreted and reassessed in the reply in (4b). For the initial poster of (4a), "recovery" entails the following factors: "my ED", "anxiety", "eating (more)", and "gaining weight". These four expressions are introduced in the first paragraph, and their meanings are constructed from the standpoint of how she/he is experiencing recovery. Higher levels of anxiety (and depression), more eating, and weight gain are constitutive elements of recovery that bring about other issues to address, be these cognitive ("negative thoughts", "negative self-talk"), physical (feeling "exhausted") or emotional (having "emotions"). The first poster also reports that she has resumed her practice of self-harm, which was a regular action performed during the worst phase of "my ED". By linking the acute phase of her ED with the beginning stages of her recovery via a common detrimental activity, the narrator problematizes the boundaries between illness and health.
In the story pair of (4a)-(4b), these linguistic resources, invoked by the first teller to build her/his personal case, are then selectively and strategically reproduced by the responder (Du Bois 2014). The second teller retrieves and recycles part of the linguistic material deployed by the first and reworks its significance in the light of her/his own circumstances. It is precisely on the basis of her/his first-hand knowledge that the second teller repeats the term "recovery" and redefines its content by adopting a different stance. The individual ED experience shared in (4a) is recategorized by the second poster as a common and faceless reality lived by any person with this condition. Thus, the first-person singular used in the initial post (4a) is replaced by generic expressions such as "a lot of people" and "X is really common" in (4b).
A more fine-grained analysis of the induced resonances in the pair (4a)-(4b) indicates that some of the coupled components point to the same stance toward the subject of "recovery", although second posters tend to slightly modify what has been stated in the original message. This is the kind of subtle alteration found in two of the couplings identified in (4a)-(4b): In (7), as well as in (8), the responder rephrases the expressions used by the initiator to talk about recovery and reassigns them a congruent meaning that confirms the original interpretations. Whereas (7b) congratulates the first poster for initiating recovery, (8b) requalifies the overall experience. Although both responses take the same stance toward the ED as the initial teller, they generate "new affordances for the meaning" (Du Bois 2014: 360). In (7a), "push myself to eat more in recovery" is reassessed as "starting recovery" (7b), which implicitly conveys the assumption that, for the responder, the first teller is in the initial phase on the path to recovery. In contrast, the subjective evaluation of the struggles involved in recovering from an ED in (8a) is coupled with a general statement about the harshness of the process in (8b). These two instances of parallelism show that, even in those cases in which the elements are just re-uttered, the sequential context in which they appear shapes them in significant diverging ways.
In addition to the combinations introduced in (7) and (8), the situated meaning of "recovery" is further developed and reinterpreted in the story pair of (4a)-(4b) with other mapping expressions, such as "gaining weight" "anxiety and depression […] gone up", "eat more/eating more", "negative thoughts", "negative self-talk", "I feel exhausted", and emotions not as dulled. These elements are introduced in the first story as the experiential factors that link "my ED" and "recovery". The second post echoes them and dynamically reconstructs their significance in the life of any ED sufferer. This way, the linguistic resources deployed by the first teller are reassessed in the second post as the behavioural and emotional indexes that signal recovering from an ED. To successfully fulfill this task, the responder claims epistemic superiority based on her/his own diagnosis ("I have CPTSD, PDD and anxiety"], an operation that lays the groundwork to review the original meanings of the resonating items. This way, the new context to understand "anxiety", "eating (so much)", "feeling exhausted", "emotions out of control" and "weight gain/gaining weight" is the shared, standard, and collective nature of these indicators of recovery. To deal with these issues, the second poster formulates some recommendations.
As Ochs and Capps (2001: 210) contend, first stories shape the structure and the theme of second stories, because they provide "a template for interpreting the first story." By reutilizing certain linguistic material, the responder of (4b) does something more than normalizing and generalizing the challenging setbacks of recovering from an ED. She/he also aggregates her/his personal account of how to deal with ED symptoms such as "negative self-talk" and "negative thoughts", which are two more resounding elements borrowed from the first story. The habitual second story shared in (4b) (fifth paragraph) introduces an operative model of thinking to address the "anxiety/depression/self-harm issues" with which the first teller is afflicted. Hence, the alternative stance taking adopted by the second teller is formulated with the induced resonances linking the story pair. This set of key resonances is part of the argumentative moves oriented to promote social cohesiveness.
More precisely, resolutive second stories such as (4b) use resonating elements as a steppingstone to advance a solution to the problematized situation posed by the first teller. The complex operation of delivering a remedy to the issue at hand requires transitioning from the domain of the self to the domain of the other. Conversely, resonating expressions inserted in parallel assessments, such as those reproduced in (6a-c) mirror the first story by producing a self-focused account of the difficulties announced by the initiator of the thread. The key components of the original story are replicated but they are confined to the limits of each individual experience. Responders arrange the resonating elements as signs of co-alignment with the stance of the teller, without reinterpreting or reconceptualizing their meaning from a different epistemic or evaluative perspective. This dialogical strategy is displayed in the exchange of messages in (9) and (10), that are part of the story pairs of (5)-(6a), (5)-(6b) and (5)-(6c). For instance, the responses instanced in (9) contain expressions that map onto pivotal linguistic material deployed in the first post. The primary resonance occurs with the expression I throw [it, them, things] away, as the vertical alignments evince: The equivalences in this interaction are also dialogically established, which means that specific mappings can be identified between the first post (5) and each reply (6a-c), as the diagraph notation in (10) reveals: (10) First post (5) I feel bad for wasting the food Response (6b) I feel bad Response (6c) I feel bad Response (6a) Ana is really wasting my OWN money and food now As shown in (10), the three parallel assessments of (6a-c) are headed by an alignment expression that connects thematically the second story to the first and gives credence to the commonality of the experience among other ED sufferers, thereby contributing to the social cohesion. For instance, the poster of (6a) initiates her/his reply with the claim "I used to do this", in which the demonstrative pronoun this points to the key activity described in the first story (the business of throwing away food). In similar fashion, the messages in (6b-c) are prefaced by constructions that confirm the alignment with the first teller (e.g., "omg super same", in 6b; and "I relate", in 6c). All three responses constitute fitting second stories that shed light on the ways in which responders sympathize and emotionally relate to the preceding story, as well as how they attune their stances to that adopted by the first narrator. Therefore, these parallel assessments bring to the fore the role of small second stories to construct shared stance, alignment, affiliation, and emotional empathy between the forum users (Simoraa 2012: 528;Figueras 2023).

Conclusions
The point of departure for the present study was the determination of the kind of narratives produced in an online forum for those recovering from an ED. The analysis revealed that the personal stories disclosed to the group displayed a gradation in narrativity (Carranza 1998;Ochs and Capps 2001;Page 2012;Thurnherr et al. 2016). Few posts complied with the canonical structure of oral narratives encompassing abstract, orientation, complicating action, evaluation, resolution, and coda (Labov and Waletzky 1967;Labov 1997). Those narrators who resorted to the Labovian schema were mostly newcomers seeking help. As previous research has shown, detailed personal stories are instrumental in the relational work that new members introducing themselves to an online mental health support group have to perform to gain attention and to receive elaborate responses (Figueras 2021). Therefore, detailed accounts of the personal history with respect to the illness acted as a kind of "cover letter" aimed at claiming the necessary legitimacy to deserve a serious reply in the ED forum. The stories identified in our sample shared two key features: they all contained a "reportable event", using Labov's (1997: 398) term; and they all claimed a personal voice, as opposed to simply conveying depersonalized information (Thurnherr et al. 2016). Granted these factors, most of the narratives produced and distributed by forum members, whether they were first or second stories, fell into the category of small stories, that is, narrative activities located towards the end of the non-prototypical storied account of events (Bamberg 2006;Bamberg and Georgakopoulou 2008).
The examination of second stories highlighted two main configurations, contingent on the kind of reply offered by responders: parallel assessments (or snapshots) and success second stories. Parallel assessments were self-centred stories and did not include any advice provision. Success second stories, in turn, became an essential component of the advice-giving act since they were remedial. The solutions offered by responders to the problem posed by the first poster were organized in two different ways in our data: offering tips (that is, a series of practical recommendations to address a specific ED or recovery issue) or delivering thoughtful advice by means of adopting the other-perspective and describing the reality of recovery as a true alternative to the ED condition. I call the latter "resolutive second stories" (Figueras 2023).
Both parallel assessments and resolutive stories displayed contrasting resonances and parallelisms in relation to the first story. Resolutive second stories contained resonating elements whose meanings were transformed and resignified from the positioning of a subject in recovery. On the contrary, snapshots echoed specific key expressions from the first post to align with the first teller by describing a similar I-perspective experience.
In both cases, by introducing their own stories to the online mental health support group, the participants emphasized their lived experience of the illness, which included their perspectives and evaluations. The dialogical nature of the ED forum facilitated the development of a story throughout the whole thread, based on the requests for clarification, comments and remarks from the rest of the participants in the discussion.
Storytelling in the ED forum was thus a collaborative product created by the interactants, a cooperative enterprise built upon open participation and fueled by involvement. Taken together, the individual small stories shared by the forum participants contributed to the co-construction of a multiple-lived story of EDs. The foundation for this choral narrative was the experiential talk exchanged within the group. This highly personal and subjective narrative material became the gravitational axis of the virtual communication, since soliciting and providing advice were the core interactional activities on the site. Advising required the generation, and subsequently negotiation, of experiential knowledge about illness and recovery, a task accomplished by resorting to operations of intersubjetivity, stance-taking, rapport, empathy, and theory of mind.