“ If you start again , don ’ t worry . You haven ’ t failed ” Relapse talk and motivation in online smoking cessation

In this article, I explore how relapse following initial smoking cessation is discursively construed and how participants position each other to enhance motivation in two different settings: smoking cessation forums and websites from the UK. In my qualitative discourse analysis, I focus on identity construction and relational work to pinpoint how users are remotivated when they have not managed to reach their goal of becoming smoke-free. Results show an imbalance regarding how extensively relapse is covered in a selection of smoking cessation forums and websites. Relapsing is constructed as a normal part of a quitting journey and not as a deviation from it. Similarly, the moral obligation of making healthy lifestyle choices influences the construction of the relapsed self. My analysis also revealed that writers resort to face-enhancing relational work strategies to console readers and connect with them. Further, while referring to personal experience was a means of normalizing relapses in forums, websites used numerical evidence to back up their informational statements. In both settings, relapsing is transformed into a beneficial learning experience, thereby positioning quitters as having an advantage over new quitters. The findings suggest that there is a common discourse of how relapsing is conceptualized, both on professional and peer-to-peer sites. This paper adds to previous studies of online health practices, providing a different angle by not focussing on success stories. It adds interesting insights by comparing peer-to-peer practices to monologic websites, and it shows that an interpersonal pragmatic approach allows investigating how participants try to impact each other’s decision-making.


Introduction
In this study, I focus on how relapsing to smoking is discursively negotiated in different online practices in a UK context.It is my aim to uncover strategies of relational work and to see how relapses, which could be regarded as failure, are in fact assigned a motivational force and are reframed as a temporary set-back.According to the stop smoking website of the Isle of Wight "it often takes people four or five attempts to stop smoking for good".Thus, relapsing to smoke is a common event in the journey to quit smoking, likely to come up in public health websites or in online support groups.Smoking as a public health matter is considered to be linked to "behavioural factors" (Neuhauser/Kreps 2010: 9) that can be improved by individuals.Public health websites dealing with smoking cessation not only want to strengthen their readers' conviction that smoking is unhealthy but ultimately want to motivate them to stop smoking altogether by giving advice and information and by showing that to become non-smokers is an attainable goal for readers.Relapses pose a discursive challenge to the overall endeavor, as they are "[a] slip back into a state of addiction" (cf.OED online; s. v. relapse: sense 2b).Thus, website authors need to construct relapses in a manner that encourages readers to persevere, reassuring them that they can nonetheless become non-smokers.Members of online support groups assist relapsed help-seekers in accepting that their failed quitting attempts are in fact a step in the right direction as well as reinforcing their determination to start over.In website and peer-to-peer contexts, authors or contributors endeavour to motivate relapsed readers/posters to continue their quitting efforts, thereby trying to exert influence on the latter's decision-making process.Hence, there is a persuasive/motivational purpose to communication.
Previous research on persuasion highlighted that, apart from the rationality of arguments used, persuasion is linked to the creation of credibility, authenticity and expertise, and the emotional involvement of addressees (e. g.Cockroft/Cockroft 2005;Miller 2002).The creation of credibility, authenticity and expertise as well as the emotional involvement of addresses are integrally part of the interpersonal dimension of communication, which is why an interpersonal pragmatic approach is useful to explore these facets (see Locher/Graham 2010).Thus, this study is an important contribution to developments in research on relational work, calling for studies on linguistic patterns due to interpersonal variation in a holistic way (see Locher 2015).Moreover, it provides insight into how failure is dealt with in online health peer groups and public health websites, filling an important niche especially in view of the pressure to produce success stories in online support groups according to previous research (Koteyko/Hunt 2016: 66).Finally, this study adds to how topics commonly viewed as lifestyle issues are discursively established in online institutional and peer-to-peer practices and what idiosyncrasies can be found in view of the dominant discourse of self-responsibility (Hunt 2015;Koteyko/Hunt 2016;Lupton 2013).In section 2, previous research is revisited to establish the theoretical background.As a next step, more information will be provided on data selection and methodology (section 3).Section 4 presents my findings and a comparison of the different practices.

Literature review
To address how smokers are motivated to continue quitting and how relapsing is discursively constructed, I oriented to research on identity construction combined with relational work, health communication and computer-mediated communication.In the following, I will introduce my theoretical understanding of identity construction and its link to relational work to address the persuasive dimension entailed in motivating others.Secondly, I will comment on empirical studies that looked at the discursive negotiation of credibility, expertise and emotional involvement in online health contexts.Thirdly, I will discuss previous research that specifically dealt with the moral implication for participants of not adhering to the discourse ISSN 1615-3014 89 of self-responsibility.Finally, I will introduce two studies that have analyzed the discursive construction of setbacks in participants' health improvement projects.
I adopt a constructionist view of identity as proposed by Bucholtz/Hall (2005), who argue that identity is a social, cultural, discursive, relational and emergent phenomenon.In the same vein, I employ their broad understanding of identity as "the social positioning of self and other" (Bucholtz/Hall 2005: 586; emphasis removed).When positioning themselves or others, interactants are actually negotiating their relationship with each other, effectively engaging in relational work (see also De Fina 2010: 205-206;Locher 2008: 511).Relational work thus can be defined as "all aspects of the work invested by individuals in the construction, maintenance, reproduction and transformation of interpersonal relationships among those engaged in social practice" (Locher/Watts 2008: 96).In other words, relational work refers to how face concerns are handled in interaction -whether face is confirmed, enhanced or challenged -, which results in identity construction.).Richardson's (2003: 176) notion of warranting strategies pointed to discourse strategies of creating credibility, such as the use of a particular lexicon, referring to other sources/experts, as well as referring to one's own experience (see also Locher 2006;Rudolf von Rohr 2015;Sillence 2010;etc.).
The strategy of sharing one's own experience received a lot of attention in research on online health communication, particularly in peer-to-peer practices, due to its prevalence and its multifunctionality.It has even been argued that reading about and interacting with others' experiences is one of the main areas that can influence the decision-making process regarding one's health (Ziebland/Wyke 2012: 239).Drawing on one's own experience is a common means to legitimize one's identity as a knowledgeable or expert contributor (or as a legitimate helpseeker) and it can also have an involving, face-enhancing effect (e. g.Armstrong et al. 2011;Lindholm, this issue;Morrow 2006;Sillence 2010).Other researchers found relationship management important in an online advice context.For instance, Locher (2006: 137) observed that the fictional expert persona engaged in face-enhancing relational work (e. g. praising, bonding) to involve advice-seekers emotionally and to signal her investment in their wellbeing (Locher 2006: 137, 201).In an online depression group, Morrow (2006: 543) described discourse strategies with a face-enhancing effect (e. g. declaring to feel sorry for someone; whishing someone well) to be commonly used to build up a relationship and emotionally involve readers.
Previous research that looked at how people address and conceptualize difficulties when trying to change their lifestyles (offline and online) are particularly interesting to this study as they raise issues of the moral order and ideal selves.Hansen/Nelson (2016) analyzed their participants' stories of quitting or continued smoking after being hospitalized for cardiac problems in an Australian context.They highlighted that smoking after experiencing cardiovascular problems had a deep impact on participants.The stories of ex-smokers showed that escaping the stigma of smoking and being morally good in light of their serious health problems was a driving force for quitting (Hansen/Nelson 2016: 8).Conversely, participants who did not manage to quit smoking worried about themselves and their perceived identity, partly considering their habit as a personal failing (12).Hence, quitting smoking was strongly associated with being virtuous and responsible (10), which is an ideological position that is also relevant for my data.Similarly, Maor (2014) observed what moral implications obesity had for the participants of her study.According to Maor (2014: 89), being obese, as (smoking) addiction, is considered to be a sign of bad individual choices and of a weak personal character.Moreover, the fact that the power to change is located within the responsibility of individuals enhances the stigmatization of the self (90).In addition, obesity is constructed as a temporary, incomplete category that people continuously strive to leave behind (Maor 2014: 91) -a categorization that is also likely to be rhetorically mobilized in smoking cessation.
In the previous paragraph, the studies presented focused on the moral dimension inherent in health matters linked to lifestyle choices.The following studies point out how participants jointly normalized setbacks in their journey to adopt health lifestyles in online support groups.
Cranwell/Seymour-Smith (2012) studied how problems with weight loss (among other aspects) were discursively established in an online support group for people with bariatric surgeries.Advice-seekers routinely asked for normalization of reduced weight loss or stabilized weight several weeks after surgery.Advice-givers co-constructed these phases as temporary, referring to their own experience to reassure advice-seekers and to establish expertise.Other ways of normalizing consisted in generalizing episodes of weight stagnation as common for a lot of people and predicting its repeated occurrences over the entire weight loss journey (Cranwell/Seymour-Smith 2012: 876).Veen et al. (2010: 30) also attended to issues regarding the management of dietary regimes, showing that second stories (i.e. responses to an initial account, relating comparable experiences) were employed for their corrective and normalizing function in an online support group for celiac disease.Responding participants referred to their own experience to qualify the initial poster's desire to quit her diet after cheating as routine occurrence (31), rejecting the cheating incident effectively as a valid reason to stop.Further, users re-conceptualized lapses as a part of the diet instead of an instance when they had stopped the diet.Cheating was always portrayed as having undesirable health consequences, thereby providing a valuable lesson in "dietary compliance" (Veen et al. 2010: 34).
As shown, previous studies have explored discursive strategies of how to index credibility and create emotional involvement in online health practices, often linking them to identity construction or interpersonal effects.These discursive strategies can have an impact on other interactants decision-making process, which is why they need to be taken into account when looking at motivation/persuasion in relapse talk.Further, previous research that dealt with the moral overtone to health improvement projects pointed to the fact that failing to adhere to lifestyle changes and being a member of a stigmatized group is not only rhetorically mobilized but has an effect on how the identities of individuals are perceived.Moreover, the two final studies discussed highlighted that normalizing was a frequent strategy used by participants in online support groups to construct difficulties as part of self-improvement.This paper will answer the following three research questions: (1) What is the discourse of relapsing in the practices?(2) How do writers position themselves and others and with what relational effects in view of relapses?(3) How can these findings be linked to motivating readers to continue quitting?

Data/Methodology
The data consist of two corpora: (1) interaction from two online forums providing peer support for smoking quitters and (2) extracts dealing with relapse from anti-smoking websites by governmental, commercial and charitable institutions in the UK.These datasets were part of a larger study, in which discursive patterns and interpersonal variation were analysed through the lens of persuasion in smoking cessation online (Rudolf von Rohr 2017).
Both forums were openly available and shared a similar purpose in terms of forum users supporting each other to reach the common goal of becoming smoke-free, reinforcing each other's determination and exchanging advice/information: (1) SmokingisBad and (2) nosmokingday.co.uk.My ethical decision-making process was guided by the heuristic principles outlined by the Association of Internet Researchers (Ess et al. 2002;Markham et al. 2012).In both support groups, participants seemed to be aware of the public nature of their contributions as they explicitly referenced having left personal messages to other interactants.Moreover, the interaction was publically accessible without having to log in or register.Additionally, the selected data is of a positive nature, which focuses on affirmative and mutual motivation and does not contain inherently sensitive information.Both forums ceased to exist in 2015/2016.
To maintain the confidentiality of participants I use a pseudonym for the first forum.I use the original name of the second forum due to copyright reasons since it was associated with the British Heart Foundation.All location markers and nicknames were changed in both forums to protect all participants' anonymity.
The corpus for my research project was made up of 80 threads, dealing with the beginnings of the journey to quit smoking.The threads started between March and April 2012 and featured 10 to 20 posts.The sub-corpus of relapsing threads for this present study was established after a thematic analysis of the overall corpus, which showed that twelve threads (a1-a6; b1-b6) out of the entire corpus discussed relapses.This made it the second most common topic1 in the early days of quitting after explicit requests for help.The sub-corpus is made up of 8,056 words, and each thread features six to fourteen interactants.Relapse threads were always started by participants who are already part of the communities, implying that they departed from the shared goal of becoming smoke-free.Their membership of the community is made interactionally relevant by the initiating participants (e. g. by referring to previous help they had received).In other words, these initiating participants need to position themselves convincingly in view of their lapse to receive further support by responding participants. 2Similarly, responding participants have to react in line with the norms of the communities and construct their identities as credible help-givers.
The second corpus consists of seven websites and was established out a larger corpus of 30 websites.They were grouped together after a thematic and subsequent cluster analysis, which showed that these seven websites covered a broad range of similar themes regarding smoking cessation (Rudolf von Rohr 2017).3Due to their thematic similarity these websites were considered to be a good choice to detect possible patterns of relapse talk.Four websites had a governmental background, being maintained by the regional stop smoking services, the local NHS sites and the main governmental smoking prevention website in the UK.Another two websites had a clear commercial orientation, distributing nicotine replacement therapy.While the final website did not sell pharmaceuticals, it is also for profit, providing health information and platforms for patient support on a variety of topics.Passages dealing with relapsing were identified through close-reading, which showed that every website mentions relapsing in some form.The emphasis given to the topic varies considerably, ranging from one source featuring a specific sub-site to two or three sentences about relapsing within information on smoking and pregnancy.The fact that relapsing is mostly treated as a secondary subject is also reflected in the number of words (1,997).The little text that is dedicated to the issue resounds with Hunt's (2015: 84) study, in which he found that there was an absence of failure on diabetes pages on Facebook.He suggested that the fact that diabetic people are overwhelmingly portrayed in a positive light might have a motivating effect on readers.
To study how the discourse of relapsing is established and how users are motivated to continue with quitting in the selected practices, I opted for a discursive-analytical framework with a focus on the interpersonal side of interaction.For my analysis, I conducted qualitative closereadings, which led me to identify arguments that characterize relapse talk.In a next step, I investigated how identities and interpersonal effects are created in order to contribute to the credibility of writers and to the involvement of participants, whereby I gained insight into how interactants/readers are motivated to continue quitting.While I carried out these steps in both data sets, my main focus was on the forum corpus, where relapse seems to be a more pressing issue and is discussed more frequently -which is also supported by the fact that relapse as a topic makes up fifteen percent of the entire corpus (Rudolf von Rohr 2017).Even so, the website corpus gave me the opportunity to compare different social and technological set-ups.

Forums
My qualitative close-readings showed that relapsing is constructed as a normal event that the majority of members have experienced at one point.Similar to previous studies (e. g.Cranwell/Seymour-Smith 2012; Veen et al. 2010) help-givers reassure help-seekers by pointing out the commonality of so-called lapses.Since failures to adhere to lifestyle changes are normalized, they lose their edge as a personal defeat of contributors.In accordance with Veen et al.'s (2010) findings, participants construct relapses as no legitimate reason to stop trying to become a non-smoker.In fact, the dominant discourse is that participants only fail if they stop trying to become smoke-free, not unlike the ideology of obesity as a temporary identity cate-gory (Maor 2014).In addition, it underlines that endeavouring to become smoke-free is clearly linked to morality -a way of following the path of righteousness (see also Hansen/Nelson 2016).
Examples 4 (1)-( 3) are typical reactions to initial relapse (help-seeking) messages, and are representative for the entire corpus.They are responses to three different help-seeking messages but they show commonly employed discourse patterns.
(1) Sorry to hear about your lapse.It happens, unfortunately.(a1) (2) You didn't let anyone down hun and well done starting again so quickly we have all been there and the only time you're a failure id if you stop trying altogether [...] (a5) (3) Sorry to hear... been there, done that.If you jump back in and restart the quit its not too much trouble [...] (b6) The help-giver in example (1) bonds with the initial poster (carrying out face-enhancing relational work) by saying he is sorry about the help-seekers lapse.Then he indexes how frequent and widespread lapses are, by using a de-personalized it-construction, which has a normalizing effect.The help-giver in example (2) also reassures the help-seeker (You didn't let anyone down) and engages in face-enhancing relational work when she uses a term of endearment (hun).She gives more weight to her reassuring words by referring to the experience of the entire group (we have all been there).She reminds the initial poster (who said she felt like a failure) that relapses do not equal failing.Finally, in example (3), the help-giver also engages in face-enhancing relational work (sorry) before he refers to his own experience.Thereby, he not only establishes an emotional link but, as in the previous examples, stresses that relapsing is nothing out of the ordinary.When he advises the initiating poster to start over soon (restart the quit), he automatically assumes that she wants to try again, orienting to the ideology of smoking identities as a temporary category.
My close-readings revealed that initial and responding posters use a specific pattern of discourse strategies to co-construct and build up motivation.Initial posters index their identity as authentic, legitimate relapsed quitters who deserve to be given support again even though they did not manage to stay smoke-free with earlier help.For instance, they may appeal to responding posters by being self-deprecating or self-reflective and referring to their history with the community.In turn, responding posters invest relational work into adopting a position of empathetic help-givers, who are open to accommodating relapsed members back into their community.It is noteworthy that help-givers focus on tending to the initiating poster's interpersonal needs through face-enhancing discourse activities such as welcoming, wishing someone well, praising etc. 5 I will use excerpts from two exemplary threads to show how participants create their identities and how motivation emerges as a collaborative process.These threads are exemplary as they illustrate the specific pattern of use of discourse activities for the two scenarios in which forum participants start a relapse thread: (1) returning to the forum after a relapse and (2) relapsing as an active member.Whether initiating posters return to the community after a longer absence to start over or whether they relapse as an active member influences the discourse strategies used.The two selected threads are by the same initiating poster: In the first thread (b3), he returns to the community after a longer period of smoking and not posting.In the second thread (b4), he confesses to failing this previous attempt.While the two threads illustrate discourse patterns that are common in the entire relapse corpus, they also highlight the difference that being a returning or an active contributor has on the interaction.

Returning to the group (b3)
The initiating contributor posts after a two-year absence.In his first post, he sets the ground for positive reactions to his return, establishing his identity as authentic and legitimate quitter by referring to his previous posting history, being self-deprecating and announcing a new quitting attempt: (4) Initiator: This place did it for me before, and after a couple of years, I started back up, first on walks to work, and then at work, and then secretly as much as possible.I have to lose a bunch of weight and i hate myself for getting rehooked... yes, it was just one puff to find myself smoking all out very quickly, Tomorrow i take the plunge yet again.I am throwing all my embarassment about restarting out the window as I loved what this site did for me years ago.Its nice having a place to rant, to complain, and to have people around that have my back.[...] (b3, post 1) The initiator praises the community for previous help (this place did it for me).Thereby, he not only indicates that he is familiar with group practices but possibly also augments the chance of receiving favorable replies.He briefly alludes to his relapse story as he outlines how he became tangled up again in smoking, which makes his account relatable and marks it as legitimate.Further, the initiator is self-deprecating to get support (hate myself).Apart from rendering his relapse story authentic (see Harvey/Koteyko 2013), this strategy is also a way of pre-empting face-aggravating comments.His hard self-evaluation can also be set in context with the stigma smoking carries (see Hansen/Nelson 2016: 12).The initiator continues to be self-critical as he confesses to having succumbed to having just one cigarette (just one puff).The use of yes indexes that he is justifying himself to the other members and is anticipating further criticism but it also positions himself as socialized member.His use of the phrase just one puff indicates that he is aware of having infringed the group's mantra of Not One Puff Ever.Finally, the initiator constructs himself as a grateful member when he explains that returning to the community was difficult (throwing all my embarrassment) but that the support is worth it.By praising the forum, he engages in face-enhancing relational work, which sets the ground for positive responses.
The first respondent restricts himself to wishing him well in his endeavor.As example (5) reveals, the second respondent prioritizes creating a connection with the initiator: (5) Respondent 2: Wishing you luck for your quit, and thanks for that post, it has really helped me today!! I've been getting on fine with my quit for 2+ months and cravings are gone, but recently I've started wondering if there would be any harm in having one to prove that I'm not addicted anymore.... Stupid I know, but somehow my brain has been making it sound like an increasingly good idea!! Luckily you've just answered the question in your post [...] so I'll be giving the 'just one' a miss!! Again best of luck, you've done it once, you'll do it again!! (b3, post 5) He also wishes the initiator well and even thanks him for posting his story, both of which are face-enhancing and subvert the initiator's negative self-characterization.Secondly, the respondent refers to his own quitting journey, pointing out that he has been in danger of trying just one cigarette himself (if there would be any harm).Creating this relation of similarity between their stories has a strong bonding effect and positions the respondent as empathetic.Moreover, the respondent not only normalizes the initiator's slip-up by establishing a relation of similarity but he highlights how the latter's story had a positive effect on him, influencing his own decision-making process (I'll be giving the 'just one' a miss).The respondent repeats his good wishes for the initiator's next attempt (best of luck) and expresses his confidence in the latter's success.Interestingly, he argues that the initiator's previous experience of quitting successfully for several years is proof of his ability to succeed, which is a powerful motivation strategy that is mobilized in the entire corpus.
A case in point is the next example: the initiator returned to express his determination and motivation to tackle quitting in post 8 (not shown here).Consequently, the respondent in post 11 asserts her faith in the initiator's abilities, alluding to the latter's previous success (you've done it before): (6) Respondent 7: [...] you know you can do this, you've done it before.Chin up, chuck the fags in the bin, and wave bye bye to the choking smoker's cough Thank you for the reminder that I can never, ever, ever light up that first fag.It's a lesson that always bears repeating.NOPE, Not One Puff Ever, means exactly that.We're all one cig away from being addicts again![...] (b3, post 11) The respondent humorously advises the initiator to get back on track, which positions her as funny, hands-on supporter (chin up, chuck the fags).In addition, the fact that the respondent expresses her confidence in the initiator and that she advises him to start over immediately has a boosting, face-enhancing effect, motivating the initiator to quit smoking.Like in the previous example, she transforms the initiator's negative experience into something positive (Thank you for the reminder) as his relapse helps her adhering to the mantra of the group, Not One Puff Ever (NOPE).Finally, she uses the first person pronoun plural we to create a group identity and point out that all members are in the same boat with the identity of being a nonsmoker easily being jeopardized (We're all one cig away).Thereby, she actively constructs the group as a community of fellow quitters.

Relapsing as an active member -the confessional tale (b4)
Five days later the same initiator posts again, confessing to failing his recent quitting attempt.The initiator's assessment of how he felt when smoking has a self-critical undertone (Yay for bad taste), indexing that he deeply regrets his lapse.His self-criticism invites empathetic instead of face-threatening responses since he is clearly critical of having smoked.Further, he positions himself as having learnt his lesson -implying that he will know better this time (see also Veen et al. 2010: 34) -, which is boosted by the use of caps (IS a very good time).
Next, he pre-empts negative comments by acknowledging that others may see him as weakwilled, thus clearly orienting to a discourse of self-responsibility and the moral dimension to health (see Hunt 2015;Lupton 2013).He counters this categorization, as he positions himself as more eager to quit than ever, pointing to the penthouse as his next goal (this is a metaphor used in the forum, which equals one year of being smoke-free).Thus, he signals that he is still in need of support and worth the time and effort of the group members, which encourages further interaction.
The next example is the first response to the initiator's confession.The respondent begins by highlighting the similarity of their quitting journeys with the exclamation snap: She positions herself as empathetic to what the initiator is going through by referring to her own quitting experience (exactly the same).Creating this type of relations of similarity is typical in threads where initiators announce their relapses as active members.Afterwards, the respondent encourages the initiator, using her own success as motivating example (I'm now five months quit).She normalizes the relapse as part of the quitting process by sharing nearidiomatic wisdom (you have to take two steps back).Finally, she advises to get back on track (Stay with your quit) and positions the initiator as having gained additional insight through his experience (you know now…).
The initiator clearly understands respondent 1's post as a display of empathy when he posts again: (9) Initiator: Respondent 1, I will definitely be reading through your posts.Thanks for the kind reply, nice to know it happens.[...] I'm going to be ending posts with NOPE now to.NOPE! Here's to a smoke free tomorrow.(b4, post 4) He indicates that he will follow respondent 1's advice and follow up her journey for inspiration, which constructs his identity as a help-seeker who really wants to improve.Further, he shows his appreciation for her face-enhancing words (thanks for the kind reply), signalling that he feels understood.Finally, he underlines his determination by reiterating NOPE, which, since it is one of the key tenets of the group, is also a means of relating to the community.
Respondent 4 starts his post by expressing that he feels sorry for the initiator's relapse, which is a typical opening move in relapse threads of active members: (10) Respondent 4: [...] sorry you had a lapse.I know you were utterly tormented for a good few hours.
Stick with it, we will help you tech the penthouse., but let's take baby steps.It doesn't matter how we get there, it just matters that we do.[...] (b4, post 8) He shows his understanding for the initiator's experience, confirming the latter's identity as truly struggling (you were utterly tormented).Next, he gives generic motivational advice (stick with it), involving the initiator by promising the support of the entire group.He mitigates his promise by advising to break quitting down into units.
In the final post, the initiator updates the other members on his progress, which is a common way of highlighting that the group's support was not in vain: (11) Initiator: 1 hour, 15 minutes and I'll be past the 24 hour mark.One more stretch this long and those blips' traces of nicotine will be out of my system.. Bring it on day 2, I'm taking you out, and your friends 3 and 4 too...In fact, bring all your friends, I will trample all of them.(b4, post 14) The initiator reports that he has nearly overcome the first day.He uses time milestones as a measure of success (24 hour mark), whereby he shows that he makes sense of quitting as is practice in the community.Both forums have been programmed in a way that accompany their members' quitting attempt chronologically.Different sub-forums represent days, weeks, and months of a quitting journey, enabling users to switch forums after reaching a certain time milestone.Secondly, forum members also use time milestones to establish how successful they have been in quitting smoking in interaction (e. g. one day, 5 months) (see Rudolf von Rohr 2017).Similarly, the initiator indexes that he fully endorses the ideology of viewing relapses as an inevitable part of becoming smoke-free through the use of the word blip.He positions himself as motivated and feisty (Bring it on day 2), signalling that he is ready to fight the battle of addiction.Initiators regularly update other members on their progress in second or later posts in relapse threads.It highlights that initiators are authentic, legitimate posters (see also Harvey/Koteyko 2013; Sillence 2013; Stommels/Lamerichs 2014).Additionally, it is an essential element of the help-giving process as it encourages respondents to keep posting, give advice or support and, due to its face-enhancing effect, is greatly important for relationship management.

4.2
The website corpus -a comparison I will use the corpus of professional quitting websites as a means of comparison to the forum corpus.While no general statements can be issued regarding how relapsing is conceptualized in UK online smoking cessation practices due to the small data set, my analysis points to some tendencies in professional versus peer-to-peer practices.The close-reading analyses of the website excerpts helped me uncover some discursive constructs that to some part occur in the forum data or that have been observed in previous studies looking at how participants try to uphold lifestyle changes.I will draw comparisons between my selection of professional websites and forums at the end of each section.
First, in the website corpus, relapses are normalized and are viewed as part of a quitting process; for instance, by using lexemes, such as blip, lapse or slip-ups, all of which conceptualize relapses as minor, temporary set-back.Therefore, they do not count as justification to stop trying (see Cranwell/Seymour-Smith 2012; Veen et al. 2010).Secondly, websites refer both to the personal responsibility of readers/smokers and to the moral dimension of becoming smoke-free (see Hansen/Nelson 2016;Hunt 2015).Interestingly, it is former quitters who take up a judgmental position toward their own smoking in the form of testimonials in websites.
In other words, the relapses are reported on by "authentic" voices rather than by experts.Thirdly, not only the identity as smokers but also the one as non-smoker is constructed as a temporary category in the website corpus.Non-smoker identities are presented as fragile, which need constant work.Website authors warn readers that their identity as non-smokers is a transitory one, which can possibly be lost again by relapsing.
The following passages represent tendencies in the entire corpus.They are taken from three different websites: (12) Of course, we're not saying it's going to be easy.And you know what?You might slip up every now and then, but don't be discouraged.It's part of the journey and it's something to be learnt from.(Niquitin UK) (13) [...] Lillian Robertson has lost count of how long it has been since her last cigarette.She says she has had a few lapses but is now smoke free using nicotine replacement gum.'I've had a few blips, there's no excuse for it, sometimes when I've had the urge to smoke, I've gone into Quitstop.[...]' (NHS South Birmingham) (14) [...] once you become 'hooked', nicotine is so addictive that if you start smoking again after a period of quitting, you quickly escalate up to your original habit again very rapidly -even it's been years since your last puff (Netdoctor UK) In example (12), website authors explain that quitting smoking is hard, predicting that readers are likely to slip up.By combining the verb with the adverbial phrase every now and then, website authors construct relapses as regular events that readers may go through but which should not deter them from continuing their attempt to be smoke-free.In fact, they consider relapses to be part of the journey and not a deviation from quitting.Example ( 13) is an extract from a testimonial on a regional NHS site, in which a quitter praises the effectiveness of the stop smoking services offered.The quitter is constructed as a successful non-smoker despite the few lapses she has experienced, qualifying her relapses as minor bumps on the road to becoming smoke-free.Interestingly, when the quitter is given a voice (as indicated by the quotation marks), she is a lot more critical towards her relapses -not unlike the initiators in the forums who take a stern view on their own relapsing.She emphasizes that her behavior was disappointing (there's no excuse for it), acknowledging that she has deviated from the right path.Even so, the quitter manages to re-position herself as virtuous by stressing that she sometimes used the services offered when she faced the temptation of smoking again.Finally, in example ( 14), website authors outline the imminent danger to a non-smoking identity, as they argue how little it takes to be a smoker again (you quickly escalate up to your original habit).Thus, the three examples show the following tendencies that are common in the entire corpus.First, relapses are presented as normal and regular incidents in a quitting attempt (ex.12), which should not be viewed as an indication of a smoker's inability to quit.Secondly, critical perspectives towards relapsing is reserved for "authentic quitters" while authorial voices seem to be more lenient (ex.13), normalizing relapses as integral part of the quitting process.This has a face-saving effect as it could be off-putting and de-motivating if website authors openly judged their readers for relapsing.As a last point, there seems to be a need to remind readers that, even if they are smoke-free, they need to beware of succumbing to smoking again.
If we draw the comparison to the forum corpus, the following observations can be made.First, relapses are normalized in both corpora and their significance as a setback is generally attenuated.Secondly, the ideology of personal responsibility for one's own health (including the moral consequences of not complying) frames and characterizes the discourse of both corpora.The voices of former smokers are used to criticize relapses in testimonials in websites whereas initiators address the failure of being virtuous by being self-deprecating in forums.Thirdly, smokers are positioned as wanting to be smoke-free by all means in both contexts.However, the argument that readers only become failures if they stop trying is only present in the forums.Finally, in both corpora, there is an awareness of the danger of falling back to old habits.In fact, in the website corpus, the identity as a non-smoker is presented as equally transitory as the one as a smoker, reminding readers to keep on their toes.In the forum corpus, members assert the danger of trying another cigarette for themselves and do not directly admonish other successful quitters not to relapse.

Motivating readers after relapses in the website corpus
The way website authors position themselves and readers depends on how relapses are embedded in the larger context.If relapses are only present as an implicit possibility (e. g. when talking about the success rates of quitting methods), website authors are likely to stress their identities as trustworthy and knowledgeable sources through informing and assessing.There is little attempt to involve readers on an emotional level.However, if relapses are explicitly thematized, website authors attend to the face-needs of readers more actively.In this context, I identified six discourse strategies that are recurrently used and combined in different sequences and patterns to motivate readers: (1) console, (2) express confidence in quitters, (3) normalize, (4) refer to numerical evidence, (5) present previous experience as advantage and (6) reconceptualize failure.Similar to the forums, strategies with face-enhancing effects are an important element of discourse.This suggests that no matter what context (peer-to-peer or professional online practices) relapsing makes involvement and positive reinforcement necessary.Moreover, the idea of previous experience as advantage is an interesting motivational strategy, as it gives a positive spin on the non-virtuous relapsed smoker.In the same vein, websites reconceptualize this failure to stay a non-smoker in order for readers to move on with their quitting attempt and mark their relapse down as temporary "blip".In the following, I will illustrate how the interplay of these strategies used in the website corpus leads to presenting relapses as normal and ultimately useful experience with two examples.Consequently, readers are positioned as being on course to becoming smoke-free and are motivated to continue.
Example ( 15) is the final tip in an entire list of quitting advice, thus mentioning relapse as a sort of afterthought in case the recommendations were not fruitful.The if-clause in the header establishes the frame for a possible relapse: (15) If at first you don't succeed... Relapses can happen, so don't despair.Many ex-smokers usually try several times before being successful.Think of your relapse as a temporary setback, rather than a failure.Use the experience and learn from it so that next time you won't be tempted again.
(Nicorette UK) Firstly, website authors normalize relapses using an impersonal statement (can happen), which also points to their expert status.The following subordinate clause builds up on the previous information to console the reader (don't despair), which has a face-enhancing effect.Authors signal that they understand the readers' potential frustration.Next, website authors refer to numerical evidence for a warranting effect (Many ex-smokers), intensifying the faceenhancing sentiment of the readers not being the only ones to relapse.Moreover, as authors ground their statements in evidence, they contribute to their credibility as a source.They advise readers to reconceptualize their relapse (temporary setback), upholding the ideology of quitting smoking as a journey with ups and downs.Thereby, they encourage readers not to give up since the relapse cannot be counted as a final endpoint of the quitting journey and thus a failure.Finally, website authors position readers as being more knowledgeable due to their mishap, recommending readers make full use of their more experienced status.By transforming the relapse into an advantage for readers, it has a face-enhancing and motivating effect.
Example ( 16) is a question-answer sequence from a FAQs, categorized under the header Advice & information on this particular website.This section features a range of topics (Preparing to quit and go smokefree, Smoking and Pregnancy, etc.) from which readers can take their pick.Once readers have selected a topic, a variety of questions pop up where readers can click on the ones they find relevant.In other words, relapsing is not the main topic of the FAQs but is part of a series of questions from which readers can choose when they want to prepare their quitting attempt.The question-answer sequence is thus pre-established and of a static nature.This artificial interaction simulates a dialogue, which has the advantage that readers have the possibility to adopt the identity of the questioners and website authors can "directly" react to their distress.In example ( 16), the voice of the questioner (with whom the reader may choose to identify) is constructed as desperate and as stuck in their quitting attempts (always fail), looking for advice: ( Similar to example (15), website authors use an if-clause to set up the context of relapsing for the addressees.Then, they move on to console and assuage readers (don't worry), which is an involvement strategy and is face-enhancing.Thirdly, website authors reconceptualize relapses as practice runs (you haven't failed), thus effectively combining it with the strategy of presenting previous experience as an advantage when starting over with their quitting attempts.
As in example (15), reconceptualizing failure has a face-enhancing effect since it positions readers as going through (instead of deviating form) the normal quitting journey.In the next paragraph, website authors boost their claim by providing supportive evidence for their previous claims, indexing their status as a professional source.Moreover, their explanation of the benefit of experience is face-enhancing, motivating readers to try again.Finally, they close with some advice on how to succeed in the future, being careful to mitigate any promises with the modal verb can.
In sum, these two examples of how readers who have relapsed are motivated in the website corpus point to interesting commonalities with and differences to the forum corpus.In both contexts, the idea of previous experience as an advantage is mobilized as motivational advice.
Similarly, relapses are reconceptualized as beneficial training for a next attempt both in the website and in the forum corpus.Due to the fact that these practices diverge with respect to interactivity, this happens in different form.Website authors re-conceptualize failures for readers whereas initiators jointly co-construct relapses as a part of quitting with responding participants.Interestingly, expertise is more explicitly stressed in websites than in the forums, e. g. by referring to numerical evidence (see also Richardson 2003).This is maybe due to the fact that websites cannot immediately amend and construct their expertise whenever necessary as in the interactive peer-to-peer forums.Moreover, the organizational/professional backgrounds of the websites may make it necessary to index their evidence-based expertise in order to increase their trustworthiness as a source.Finally, even in the non-interactive context of the website practices, involving the readers through face-enhancing relational work is greatly present.Thus, creating a connection seems to be a crucial element in re-motivating relapsed smokers regardless of the medium.

Conclusions
This paper focused on how relapsing to smoking is viewed in a selection of online smoking cessation practices in the UK and how interactants construct their identities and negotiate face when relapsing.Further, these findings were linked to establishing a re-motivating frame for users/readers to keep quitting.Regarding the discursive construction of relapsing, my close qualitative analyses of forums and website practices revealed that relapsing to smoking was presented as part of a quitting journey, indexed by lexical choices that put these incidents into perspective.This clearly had a rhetorical and corrective function: since relapsing did not inherently endanger or question the decision to quit, smoking quitters were expected to continue on their quest to become smoke-free (see also Veen et al. 2010).Similarly, a smoking identity was viewed as a temporary category that participants/readers were striving to leave behind, thus exhibiting similarities to the discourse around weight loss (see Maor 2014).Additionally, attempting to quit (and consequently relapsing) was tinged with considerations of morality and responsibility, which exert pressure to become smoke-free.In the forum corpus, participants were considered to be virtuous so long they try to leave cigarettes behind even though relapsed members presented their relapses as a moral failure.In the website corpus, authors stressed the importance of the individual's willpower (see also Hunt 2015;Lupton 2013), withholding judgement themselves.Nonetheless, they could interpose the voices of quitters (in testimonials) who were self-critical towards their own relapsing.Moreover, it is important to point out that relapsing was given more weight in the forum corpus whereas relatively little text was dedicated to the issue in the website corpus, suggesting that the possibility of failure is downplayed.
In the forum corpus, dealing with relapses and remotivating each other was a collaborative endeavour.Relapsed members positioned themselves as deeply critical and reflective of their failure, which had an appealing effect.Additionally, they presented themselves as humbly appreciating and praising previous support, which encouraged respondents to re-engage in motivating them.Overall, responding participants adopted a position of empathetic helpgivers.While the fact whether the initiator was an active community member who relapsed or returned to the group after a longer absence had an influence on the type of discourse strategy employed, the face-enhancing dimension of interaction remained unchanged.Further, relapses could be re-narrated as a positive example insofar as it discouraged users to succumb to temptation.Since respondents mentioned how initiators kept them from making the same mistake, it had a consoling effect on initiators.Importantly, respondents referred to their own experience to normalize relapses at large.These stories tended to be of a face-enhancing nature only secondarily stressing respondents' expertise.In turn, while normalizing was also common in websites, it primarily highlighted their status as professional, expert source.Further, the use of numerical evidence also contributed to source credibility, which should encourage readers to comply with advice.In both settings, relapses were transformed into being a valuable lesson, presenting previous experiences as an advantage (see also Veen et al. 2010).Therefore, talking about them constituted a powerful, face-enhancing, motivational strategy.Generally, websites carefully built up their motivational attempt when talking about relapses, building rapport by consoling and normalizing before reconceptualizing failures, expressing their confidence in quitters and presenting previous experience as advantage.
Analysing relational work and identity construction proved a fruitful way to tap into how motivation is (co-)constructed in these diverse practices, revealing that relationship management is crucial in the face of set-back and potential failure.Further, this study uncovered interesting patterns of how the ideology of self-responsibility and the moral dimension to health is discursively negotiated when people do not succeed in their self-improvement projects.Future research needs to delve deeper into how the failure to follow lifestyle choices is dealt with online, especially in view of the dominant discourses of self-responsibility in public health websites.Moreover, the question of how relational work and identity work can be strategically employed to motivate deserves further study.

Corpus
Netdoctor UK. www.netdoctor.co.uk [01.05.2012 and a huge overwhelming paranoia, guilt, and sadness.I'm going to chalk this experience to letting myself know that this IS a very good time to quit...[...].Some of you may see this as another weak willed person fluttering around the forum I guess..But I'll be reaching the penthouse now, no matter what it takes.[...] (b4, post 1) He shares his story, explaining what his, admittedly mistaken, reflections were: (7) Initiator: Tomorrow begins... Well, day 1 revisited.Somehow talked myself an hour ago into not being ready to quit, got smokes, smoked two.Thought process was to set a cap on smoking, and wean myself until I felt truly ready.Yay for bad taste, instant headache, a reminder cough, 16) Q: I've tried to give up many times, but always fail -what else can I do?A: If you've tried before and it didn't work out don't worry.You haven't failed, you have just given yourself more practice for the next time you quit.Research has shown that the more attempts you have made in the past, the more likely you are to succeed in the future.This is because every time you are getting more experienced in how to quit.