The liminal landscape of mentoring—Stories of physicians becoming mentors

This study explores narratives of physicians negotiating liminality while becoming and being mentors for medical students. Liminality is the unstable phase of a learning trajectory in which one leaves behind one understanding but has yet to reach a new insight or position.

In several learning philosophies, sensations of frustration and confusion are considered key to learning, such as Biesta's notion of education being "risky." 7With risk, there may be anxiety and uncertainty, which can be necessary for learning beyond the trivial. 8,9Within the threshold concept framework, this trajectory can be understood as a state of liminality. 10Liminality is the provisional stage in a learning trajectory where one way of understanding is left behind, and a new insight or position is not yet arrived at. 11,12The mentoring context has its special meaning, relevance and significance, which make the ways in which liminality is recognised, perceived and negotiated by the individual unique and necessary to explore. 13en facing liminality, believing that talent and abilities are dynamic and can develop over time, may help allocate motivation to tolerate confusion and embrace challenges and criticism.In contrast, own traits can be understood as immutable, efforts as futile, and hence constructive negative feedback may be ignored, and challenges avoided. 14,15Even though mentors often are expected to cultivate professional becoming in their mentees, there has been limited attention given to how mentors approach their own liminality and the possibility of learning through struggling.
In the existing literature, it is hard to come by detailed narratives about the experience of becoming and being a mentor.Narratives are powerful representations of lived experience.Where other methodological approaches often aim to reduce complexity and identify general principles or categories, narratives have the potential to convey interpretations, situatedness and complexities in experiences. 16These qualities make narratives suitable for exploring the flexible and fluid construct of mentoring. 17 this study, we analyse narratives of becoming and being a mentor for medical students in group-based mentoring programmes.We sought to explore: How do mentors negotiate liminality?

| Context
This study draws on empirical data gathered among mentors from three undergraduate medical programmes, two in Norway and one in Canada.There are certain similarities between the universities' mentorship programmes; they offer longitudinal, group-based mentorship programmes for students; for each mentor group, there are six to eight students, with one or two mentors; and mentor groups meet approximately four to five times per year, generally outside the clinical setting (e.g., in a mentors' home or at university campus).The mentorship programmes have common goals: To offer a safe space for reflection, to assist students in their transition from layperson to professional and to develop a person-centred mindset.All mentors have been offered preparatory instructions and intermittent workshops.

| Study design
This is a qualitative interview study, placed within the socioconstructivist paradigm, applying dialogical narrative analysis on mentors' stories of becoming and being a mentor. 18The study informants were recruited from a comprehensive survey in June 2017, in which we invited all mentors from three universities (N = 461); 148 out of 272 survey respondents agreed we could contact them for a possible semi-structured interview.Through e-mail, we recruited a convenience sample representative of the study population by gender, university affiliation and mentoring experience (N = 22) (see Table 1).Among the 22 interviewed mentors, 13 were full-time clinicians, whereas nine were engaged in research, teaching or administration.
Between October 2017 and April 2018, three authors (ES, EHO, MK) and two research assistants conducted semi-structured interviews.
They did not interview mentors from their affiliated university.The interviews were designed to elicit stories of becoming and being a mentor, asking, for example, "what was your first experience as a mentor like?"; "how did you learn to be a mentor?";"what is a good mentor?" and "has being a mentor affected you or changed your ways of being a person, teacher, or professional?"Interviews lasted 40-70 min, were audiotaped and later transcribed verbatim before analysis.
The dataset has previously been analysed, resulting in identification of threshold concepts in mentoring. 4However, there was an unredeemed potential in the data for exploring how mentors negotiate the liminal phases of mentoring.The present study provides additional contributions as it introduces a new research question, using a different theoretical framework and methodology, while filling a significant new knowledge gap and having distinct implications. 19

| Methodology
Narrative inquiry is the study of stories in which "one thing happens in consequence of another" 18 A whole story is a unit of analysis in which parts relate to the whole and cannot stand alone.Such a stance enables a recognition of the situated nature of learning to be a mentor.Through narratives, people "convey to themselves and to others who they are now, how they came to be, and where they think their lives may be going in the future" 20 Dialogical narrative analysis (DNA) is a specific method for exploring aspects of identity. 21With its roots in Bakhtin's dialogism, DNA enables exploring how people construct identities through narratives within a given context.
We searched for "hidden dialogues" in the informants' narratives, drawing on Bakhtin's ideas on the multiplicity of dialogue, narratives always being "populated by the voices of others." 22Dialogue is, in this perspective, seen as a meeting place in which own and other's opinions, values, attitudes and intentions reverberate. 23Avnon points to the idea of "hidden dialogue," meaning how the narrator, consciously or not, incorporates impulses from the surroundings, contexts and communities into the narrative. 24Thus, a narrative may have layers of content and significance, depending on whom the narrator wants or needs to be in dialogue with and which identity they take on.For Bakhtin, a narrative is not entirely anyone's own but is told as a dialogue between multiple voices of self with rearranged elements of others' stories.The self is seen as unfinalisable, dynamic and everevolving.The identification of hidden dialogues provides essential information about the narrator's identity, the shaping forces of this identity and the negotiation of the identity in an ongoing narrative.
The narrative may thus contain contradictions and inconsistencies, as the process is continuous, partly nonconscious, dynamic and changing.

| Data analysis
Three researchers read all transcripts independently (EAV, LMH, MK), identifying mentors' strategies for negotiating liminality before collectively discussing these various approaches.We identified four stories suitable for deeper analysis due to their apparent, variable features that addressed the research question differently while displaying the complexity of mentoring.In DNA, the selection of narratives is an act of practical wisdom (phronesis), basing the decision on insights developed as we analysed the material. 21 then performed the DNA, anchoring it in the following characteristic questions: (1) What multiple voices can be heard in any single speakers' voice; how do these voices merge, and when do they contest each other?( 2) What is at stake for whom?(3) Who is holding their own in the story?(4) Why is someone choosing to tell a story?
(5) How does the story teach people who they are, and how do people tell stories to explore whom they might become? 21To help analysis, we created a table of who the stories were in dialogue with and the corresponding identities of the narrator.Liminality functioned as a sensitising lens, giving us a tool to explore narratives that otherwise may come across as chaotic and dispersed, with a focus on learning and identity formation.We revisited the rest of the material frequently, verifying that we could understand other stories as closely related to the features we had identified and reconsidering our choice of narratives for deeper analysis.At this point, we decided to present these four narratives in the results section.They were chosen due to their varied illustration of the essential aspects within the complete material that we found relevant to the research question.
As we explored theoretical literature, new nuances of meaning became more apparent.Therefore, continuously revisiting the data during the writing process was necessary.The act of writing the article became an essential part of the analysis, being a dialogue between us (the researchers), the data (interview subjects), existing knowledge (theory) and the expected readers of this article (you).

| Ethical considerations
Choosing to present excerpts of narratives has ethical implications.
We have taken considerations to ensure that the mentors' stories we present are not recognisable to others.All stories were given pseudonyms before analysis, and in this article, we omitted information making re-identification possible from quotes.Still, there is always the possibility of a mentor recognising what they believe to be their own story.
We have tried to avoid passing judgement on the stories, including not seeing one way of developing as a mentor better than others.
Nor has our focus been to conclude about the mentors' success.During our research, we were aware of the risk that some mentors might have differing views on how we interpret their narrative.However, as a socio-constructivist study, the narratives we present are the results of our analysis of texts of mentors' narratives and should be read as that.At best, we present part of the study participants identities. 25

| Reflexivity
We tried to be aware of what we, as researchers, brought to the research, be that our ontological position, theoretical and methodological assumptions and awareness of our interaction with the research objects.This relates to how we framed our thinking and positionality and took ourselves into the research process. 26,27Positionality refers to the "examination of place, biography, self and other to understand how they shape the analytic exercise." 28In this, we sought to establish "ways of seeing which act back on and reflect existing ways of seeing." 29r research group represents a variety of competencies and

| RESULTS
In the following, we present excerpts of four narratives, each telling a story about a mentor's journey.The narratives are complex, internally overlapping, dynamic, convey unconscious thoughts and ideas and can contain contradictions and inconsistencies.We present the narratives with orientation around three components: negotiation of liminality, narrative coherence and dialogues between identities.
Mentors from all universities are represented in the following narratives.Some characteristics are disclosed in the narratives when relevant, whereas others are withheld to protect their identities.To help present the narratives and how they compare, we have created a condensed summary (see Table 2).

| Fred's narrative-Narrative coherence
Even with only a few years of experience as a mentor, Fred enjoys it.He sees mentoring as a way of helping students with what he refers to as the "non-knowledge medicine," helping them develop "soft skills" and accompanying them through all their "concerns and issues" as students.
As Although he seeks to reject learning, he is in a liminal space where learning can happen.Importantly, this conflict generates powerful, difficult emotions for Ian.For instance, he is frustrated and angered by feeling "pretty stupid" as a mentor.He also displays shame for being a mentor, hiding his educational responsibility for colleagues.

| Vera's narrative-Dodging liminality
While having been a mentor for only a few years, Vera has decades of experience as an educator.She is aware that she is known for being honest and direct-a personal characteristic that she takes pride in, although it has given her "many nose stiffeners along the way.The statement that she has "actively avoided" courses on how to teach indicates that she is likely to refrain from liminality.It also connects with how she, in her view, always has been the way she is.She does not seek to change; she avoids such identity formation when possible.Accordingly, her suggestion for changing the mentoring programme is to "abolish it."Her struggles as a mentor are external to her; they are due to the mentoring programme's design, instructions and goals.
Vera's narrative becomes incoherent as she tries to be in dialogue with several of her identities simultaneously.As she tells her story contradictions to her portrayed identity of being "self-taught" become visible.While denying that she ever had any role models, she tells of being influenced by others.While portraying herself as always going her way, she also tells stories of a different Vera in the past who complied with the norms and social culture of medical school.She is not consciously aware of her own identity formation.

| DISCUSSION
We have presented four narratives of different approaches to liminality while becoming and being a mentor.During our analysis, we identified three components of the narratives that enriched our understanding of becoming and being a mentor: Negotiation of liminality, narrative coherence and dialogues between identities.There seems to be a link between mentors struggling with taking on the mentor role and being unaware of their professional identity formation, claiming always to have been like they were and denying any role models.These mentors proclaimed that their struggles were due to the mentoring programme's structure, not something within their control.In contrast, those mentors that were aware of how their own identity had formed as part of a culture focused on what they could do differently to succeed as a mentor.This resonates with previous research, which has indicated that mentors need to be prompted into thinking about how their own identity has been formed to be able to help their mentees better. 31desire to avoid liminality can come from focusing on increasing own comfort, confidence and efficiency. 32With that focus, it is rational not to want to enter liminality and potential for experiencing discomfort, confusion and frustration while trying to reach a new understanding.It can be more comfortable and easier to employ known solutions to fix their struggle with being a mentor than to find new ways of being and doing. 33cusing on their dialogues, we found that some mentors were in a 'dialogical struggle' between different selves. 34This struggle is most salient in Ian's story, where the hidden dialogues with multiple identities give the narrative several internal contradictions.This finding shows some of the struggles of integrating different identities.The hidden dialogues reflect how they stage themselves for others, changing their presentation of self between the "stages" they enter. 35As they move between different stages, inconsistency enters the narrative.As they try to please several audiences simultaneously, it becomes visible that they have other identities and layers at stake than what they try to show.Contrastingly, Fred and Dan tell coherent stories without diverging dialogues.Those with a mentor identity integrated with their other identities can tell one coherent story of themselves to all their dialogue partners.With greater integration between their identities, they need only one stage.
Those struggling with integrating their mentor identity with other identities, like Vera and Ian, may experience themselves as inauthentic mentors, not enacting their true selves. 36Previously it has been suggested that a connection between personal values and the work as a medical educator is helpful for the individual to invest in integrating identities. 1 In this study, we strengthen that argument as we find that tension between identities can stem from mentors understanding of medicine.The mentors who find the humanistic and emotional aspects of practicing medicine vital seem to have an easier task connecting their identities.They can "stand in the spaces between realities without losing any of them," with "capacity to feel like one self while being many." 37In contrast, if mentors have an understanding of medicine in which the medical humanities are deemed peripheral, other goals might become more central, making it harder to integrate a mentor's identity with other identities.
This study has implications for the practice of faculty developers who support clinicians that teach.They should be aware that those seeking faculty development initiatives are likely to accept and welcome liminality, perhaps seeing it as an unavoidable affordance as they grow and develop as professionals. 38Others avoid these initiatives, maybe interpreting liminality as a sign of failure and rejecting the integration of a mentor identity.Some are not yet ready to become a mentor and occupy liminality while being "betwixt and between" their previous and new understanding of self. 39In our study, the sociocultural surroundings influence some mentors to the extent that we question if it rarely will be productive to target them individually in the hope of helping them develop.We suggest that faculty developers focus on context, as the contradicting dialogues show the importance of the professional environment for determining which strategies clinicians that teach resort to when facing liminality.
For instance, faculty developers can focus on creating a culture where mentoring is deemed valuable and central for all doctors. 3If that helps clinicians expect and tolerate liminality, they may be rewarded with learning "journeys" in unexpected landscapes, possibly becoming a self-reinforcing spiral, increasing their ability to search for new solutions.
There are potential limitations to the study.The selection of four narratives for deeper analysis could have been done in multiple ways, which would have yielded different analyses and shed light on different aspects of becoming and being a mentor.Any narrative can only partially portray the richness of the complete material.While our choices are in concurrence with the DNA method, we recognise that further analyses could expand the dialogue. 18In addition, this study was conducted in two countries, with three different contexts and two languages.While there are cultural similarities, there are also differences that may have impacted mentors' narratives.Future studies may investigate how individual and contextual variables can contribute to being a successful mentor.

| CONCLUSIONS
Liminality is an uncomfortable, ambiguous, but often inherent and necessary aspect of learning and adapting to new roles.Mentors who have not integrated a mentor identity with their other identities may experience themselves as inauthentic, with feelings of discomfort, confusion and insufficiency while becoming a mentor.They might want to avoid this liminal experience, thus reducing their chances to integrate identities and render their narratives coherent.Cultivating a culture supportive of mentoring in medicine may help these physicians develop internal dialogues that may reconcile their clinician and mentor identities, making them more likely to embrace liminality and grow as mentors.

AUTHOR CONTRIBUTIONS
lenses for observation and analysis, respectively, having medical (EAV, ES, EHO) or educational (LMH, TS, MK) training, with a range of experience from doing educational development, teaching medical students and faculty and doing research focused within medical education on identity formation, threshold concepts framework and liminality, affective dimensions and mentoring.We divided the research team into two groups: The main group (EAV, LMH, MK) was hands-on with the data, whereas the coresearchers (ES, EHO, TS) participated in discussions about tentative interpretations of the data.The co-researchers focused on questioning the findings and their meaning, discussing implications and relevance and suggesting ways to move forward.This structure aimed to ensure that we could hold different perspectives salient throughout our work with the article.
While she cares for students, she struggles to grasp what the mentor role expects of her.She says: A mentor is someone that people know who are, I think almost every student at our faculty knows who I am.They know my pros and cons, I think.They know that I mean well.But mentoring a group of eight student, I have struggled big time with doing that.[…] Well, I am thinking, up to 10% of medical students have a problem.And you need to identify who they are in a good way and help them.So, the group that I have mentored haven't had that.They have all been wonderful people without problems; they talk and share and are very engaged in our meeting.[…] But they don't have any problems.[…] So I don't feel I have been … or had much value for that group.While she finds it challenging to understand how she can help her students, Vera tends to give up on finding out how she can be of value.For her, there needs to be a problem to fix.When asked how she learned about teaching, she replies: I have never learned it.I have never taken any courses in how to teach, never did anything.I am very selftaught.I think that works too.[…] I have actively avoided it.I use myself as a reference for what works and what doesn't work.
Fred interacts with the students, he gets new insights, even Fred enters liminality willingly.He could have rejected reflecting on the situation, by underplaying the effect and importance of what he said.Instead, he dares to feel uncertain and uncomfortable, putting himself consciously in a position to learn.He is unlikely to say such a thing to medical students again.Meanwhile, he is protective, not wanting such cynical statements to influence the students.
reflect on what is uncomfortable, he can learn from situations: I was telling them I became a specialist because I liked (…) care and this and that ….And I said "you know what, at the same time I'm going to be honest with you; I'm not gonna mind doing two extra years to be Throughout the interview, Fred tells a coherent narrative.He seems untroubled by being a mentor alongside being a physician, seeing his mentor and physician identities as connected.Therefore, he can take learning from one arena to the other, making his physician capabilities useful as a mentor, and vice versa.The coherence is also displayed in his dialogue with his surroundings as he tells his story, not shifting between positions or identities that contrast each other.3.2|Dan'snarrative-EmbracingliminalityDan is an experienced physician having several years of experience as a mentor.He understands his own identity as being formed through sociocultural interaction.Role models and culture have influenced him, and he has lived through a change in culture from one way of understanding to a new understanding.As a mentor, his "driving force was to be able to give back" to this collective: I was coming out of the old culture where if something wrong happened it was your fault.Because you weren't good enough.Because you hadn't learned.His high tolerance for uncertainty and his tendency to search internally for what he can do differently helps him tolerate liminality until he has learned.So, clearly, I said that there is something here and just because I don't understand it [the mentoring T A B L E 2 A condensed summary of the results.Like Fred, he tells a coherent narrative, displaying a broad understanding of medicine which includes being a mentor.He is in one consistent dialogue with his surroundings, as his identities are connected.3.3 | Ian's narrative-Conflicting identitiesIan is an experienced physician who struggles with being a mentor, as he feels he has "very little to contribute with."He is unaware of his own processes of professional identity formation and does not see that he can develop further as a mentor to help students form their identity.Ian considers the problems he experiences as external to him, characteristic of the mentoring programme, which he thinks should try to prepare students for the "coming disaster" of working life and not on "all those soft areas."Even though he claims to not having mastered the mentoring role, he actively avoids courses that might help him.Ok, it's zero reward [from mentoring].There is no reward.It's just stress.There is no joy, that way.(…) Yes, I'm not saying I'm not having fun, but it's just because of her [co-mentor].Yes?Not bec … and the group is fun.It's fun to see these younger students and … how eager they are and idealistic they are and such.