Auto-hermeneutics: A Phenomenological Approach to Information Experience

Abstract. The need for methodologically rigorous approaches to the study of human experience in LIS has emerged in recent years. Auto-hermeneutics is a research methodology that offers a systematic way to study one's own experiences. In LIS, auto-hermeneutics offers a way to approach emerging questions regarding information experience and allows researchers to explore yet-undocumented contexts, setting precedents for further work in these areas and ultimately widening our understanding of information. Auto-hermeneutics draws principles from autoethnography (perhaps the most well-known of automethodologies), self-study and systematic self-observation; prior studies of these types in LIS and allied fields are presented. A discussion of generalizability, validity and reliability in auto-hermeneutic research follows. Finally, an example of an auto-hermeneutic study conducted by the author is outlined for illustration.

The birds have vanished in the sky, and now the last cloud drains away. We sit together, the mountain and me, until only the mountain remains.

—Li Po, 8th century

Introduction

Though its scope of inquiry has been defined broadly, most information behavior research has focused on information seeking (Bates, 2010; Case, 2016; Fidel, 2012). For decades, scholars have been calling for more research on the outcomes of information seeking (Case \& O'Connor, 2016). Among these outcomes, people's in-the-moment engagement with information has been identified as an area in particular need of characterization (Kari, 2007). To this end, the research area of information experience has emerged. Bruce, Davis, Hughes, Partridge, and Stoodley (2014) offer a collection of scholarship on information experience, calling for the consideration of novel methodological approaches to information experience as a research object.

This paper presents one such methodological approach: auto-hermeneutics, which offers, in short, a systematic way to explore and describe the ontological nature of one's own lived experiences. For library and information science (LIS) researchers, auto-hermeneutics is an accessible way to consider novel questions regarding information experience and explore yet-undocumented contexts, setting precedents for further work in these areas and deepening our understanding of information phenomena. This paper first considers motivations for and unique contributions that can arise from the study of the self. It then discusses prior approaches to the study of the self in LIS---autoethnography and self-study---before introducing an auto-hermeneutic methodology that draws from systematic self-observation and interpretative phenomenological analysis. Finally, an auto-hermeneutic LIS study is described in order to illustrate the methodology.

Why Study the Self?

Know thyself, implores the Delphic maxim. These words have been carried, from a pillar in ancient Greece to the furthest reaches of cyberspace, by so many hands that today, over 1,600 years after the last priestess presided over Mount Parnassus, the phrase is common---even banal. What good reason is there to know oneself beyond vapid egoism? Is researching the self merely another instance of academic navel-gazing, yet another stop on the Sisyphean quest for hypertrophied abdominals, or does it have veritable utility?

If Western philosophy engendered know thyself, then perhaps the East can explain why. The Zen tradition, for example, is based on the principle that only by knowing the self can the self be forgotten. Renowned Buddhist teacher Dogen-zenji put it this way: "To study Buddhism is to study ourselves. To study ourselves is to forget ourselves" (Suzuki, 1970, p. 79). Similarly, Lao Tsu says in the Tao Te Ching: "Knowing the mother, one also knows the sons. Knowing the sons, yet remaining in touch with the mother..." (52). This ancient wisdom suggests that studying one aspect of reality (in this case, the self) can lead to insights regarding other, connected aspects of reality. Put differently, such research can contribute to understanding---appreciation of the relational structure of different pieces of knowledge (Kvanvig, 2003)---which has been proposed as the key epistemic aim in LIS research and practice (Bawden, 2016).

This view has been echoed by numerous researchers in the social sciences. Bromley (1986) argued for the single-case study as a legitimate and valuable method of scientific inquiry, both within and beyond the discipline of psychology. In particular, he emphasized the value of studying cases in their real-world contexts. Similarly, Bruner (1986) extolled narrative research, arguing that humans have two ways of knowing and reasoning: first, through deductive logic; and second, through narrative. Through logic, discourse begins with a general theory and move toward specific examples, arguing for truth. Through narrative, discourse begins with examples and works its way toward general theory, arguing for lifelikeness. Indeed, modern psychology supports the notion that it is through narrative that people form their identities---and, moreover, that researchers come to know human behavior generally (McAdams, 2001). Smith, Harré and van Langenhove (1995) argued that the formulation of theory should stem from intensive idiographic studies, moving from specific observations to general principles.

These are compelling validations for the single-case study in human behavior research, but there are also important reasons to study the self in particular. First, doing so is convenient, which should not go unappreciated. Researchers have ongoing access to themselves. This reveals the advantages of automethods for longitudinal, demanding and speculative---perhaps even invasive---research that might otherwise be difficult to conduct, especially when outside funding is unavailable. Researchers also, in general, have more access to their own thoughts than they do to the thoughts of others. Thus researching the self may potentiate deeper, more precise and more accurate data collection. Rigorous methods for doing so would respond well to Bates' (2004) appeal for information-related inquiry that allows and encourages participants to freely express their thoughts and experiences so as to be valid within the person-centered research paradigm. Moreover, automethodologies may be a uniquely effective way for making certain advances in the social sciences. For instance, philosophers of science such as Harding (2015) argue that each researcher (and research community) operates from a particular standpoint which cannot be overcome; obfuscating this standpoint under the guise of disinterested "objectivity" is merely deception, while embracing it can reveal phenomena and biases that were previously hidden.

Two Approaches to the Study of the Self

But even if it is accepted that doing so can be useful, how does one go about knowing themselves in the first place? The ancient aphorism offers no guidance in this regard. Fortunately, the academic study of the self has, in recent decades, been developed in a number of ways. In LIS, autoethnography is perhaps the most well-known of automethodologies; the automethodological toolkit in LIS has also included self-study. This section provides an overview of autoethnography and self-study as grounding for the introduction of auto-hermeneutics in the following section.

Autoethnography

Stemming from ethnography, autoethnography allows a researcher to explore their place in a culture, emphasizing the relationship between self and other (Chang, 2008). In general, autoethnography seeks to paint a picture of the amalgamate experience of living within a culture rather than a single, discrete experience. This is often accomplished through narrative analysis. As they rely so heavily on narrative, autoethnographies tend to be characterized by their evocative and personal writing styles (Ellis & Bochner, 2000). In conducting autoethnography, researchers draw on their richly nuanced understanding of the context under study; such understanding may not be available when studying cultures of which the researcher is not a member.

Though autoethnography is mostly used in anthropology (Chang, 2008), the approach has been adopted by some researchers in LIS. Guzik (2013) discusses the usefulness of autoethnography in LIS research, particularly as a way to help information professionals become aware of and analyze their cultural assumptions in order to improve the way they develop programs and services. Guzik gives the example of Michels (2010), who investigated his experience as both a PhD student and librarian in an academic library over the course of a year. As an exploratory endeavor, Michels chose to represent his findings through poems and videos, which were found to resonate with his audiences and reveal some of the hidden assumptions that information professionals have about the people they serve. Another autoethnographic study in LIS is presented by Polkinghorne (2012), who analyzed herself as an information literacy instructor in order to uncover the factors---such as a sense of unpreparedness---that contributed to how the instruction was carried out, again as a means for improving practice.

As these studies demonstrate, autoethnography in LIS is aligned with action-research and critical methodologies, which seek to effect improvement in practice (Mills & Birks, 2014). This, it seems, stems from the emphasis the autoethnographic approach puts on the relationship between self and other in ongoing cultural engagement. But for research questions that seek a deep ontological characterization of a phenomenon, or those that wish to take a purely descriptive rather than critical stance, other methodologies may be better-suited. Still, it would seem that all automethodologists can draw on the rich tradition of autoethnography, including standards of practice and critiques (e.g., Holt, 2003), in order to inform their own studies.

Self-Study

Educators have long employed automethodologies to improve their teaching, giving rise to self-study. This research tradition formally emerged in the 1990's as a systematic way to study the self in various roles of situated practice (Lassonde, Galman, & Kosnik, 2009). Self-study holds that only through the close examination of one's own practice can that practice be improved. Rather than comprising a specific methodology, self-study draws from a wide array of other methodologies and traditions, including interviewing, phenomenology, participatory research, artistic development and ethnography (Lassonde, Galman, & Kosnik). Because of its focus on improving practice, self-study is not dissimilar from how autoethnography has been employed in LIS.

One particular innovation of self-study is that it can be collaborative; a group of educators who individually engage in self-study can compare their findings and learn from each other, thus furthering the practice of the cohort as a whole (Lassonde, Galman & Kosnik, 2009). It should also be noted that collaborative autoethnography has also recently emerged (Chang, Ngunjiri & Hernandez, 2012), perhaps inspired by this aspect of self-study, and collaborative autoethnography has indeed found application in LIS research in a study on the information practices of caregivers (Anderson & Fourie, 2015). In a related way, self-study has also been employed on the organizational level---an organization studying its own practices. Van Cleave (2008), for example, documents the literacy assessment efforts of the San Francisco State University, in which a self-study was conducted using a survey that was distributed to teaching librarians, and the results informed organizational planning.

So far in this paper, it has been suggested that researching the self can offer a unique approach to building understanding of phenomena of interest to LIS. While this has been breached through autoethnography and self-study, neither of these approaches has been able to address ontological questions regarding the essence of information phenomena. In the following section, it is suggested that hermeneutic phenomenology can contribute to this end, as evidenced in recent LIS work that has adopted a hermeneutic-phenomenological approach. This work has suffered some limitations, however, which may be overcome through the hermeneutic-phenomenological study of the self, which will be described further on.

Hermeneutic Phenomenology and LIS

Human-centered research has become a focus of LIS in recent decades, but certain assumptions have been carried over from earlier positivist traditions (e.g., "information" as something impersonally observable and measurable) (Case, 2016). A number of scholars have emphasized the importance of clarifying key concepts in LIS, which requires untangling these assumptions (Day, 2000). One approach to doing so is offered by the philosophical tradition of phenomenology, first proposed for use in LIS by Graziano (1968) as a way to explore the essences and relationships among the phenomena of interest to LIS.

Phenomenology has been conceptualized and developed in diverse ways by scholars over the past century. Budd (2005) offers an overview of these developments as they concern LIS. In Budd's view, these seemingly-divergent approaches to phenomenology share a number of commonalities: They are inductive and descriptive, and they recognize that experience is richer than what our senses apprehend and that the world has no meaning apart from consciousness. Von Herrmann (2000/2013) sees phenomenology as falling in two schools: the Husserlian school of reflective phenomenology, and the Heideggerian school of hermeneutic phenomenology. While both phenomenological approaches go back "to the things themselves" (von Herrmann, p. 5) reflective phenomenology involves the analytical isolation of phenomena, while hermeneutic phenomenology involves analysis in situ, emphasizing the role of interpretation. While von Herrmann emphasizes that both approaches can yield valid (if different) results, Budd (1995, 2005) argues that LIS would benefit most from adopting hermeneutic phenomenology as an epistemological and ontological foundation. "Any serious examination of informing should take into account the interpretive ontology of Heidegger. The question, 'What does it mean to be informed?' is the kind of reflective questioning that is integral to being" (Budd, 2005, p. 49). In Budd's (1995) view, the interpretative nature of hermeneutics is ideal for LIS because it is not isolationist, and nor is it merely cognitive or behaviorist; it is holistic, contextualized and iterative, essentially based on multiple interpretations. In this way, inaccuracies and inconsistencies of understanding gradually wash away as understanding moves toward determinacy. Vamanu (2012) also rallies for the importance of hermeneutics to LIS, outlining how the hermeneutic approach can lead to understanding.

As developed by Heidegger (1927/2010), hermeneutic phenomenology pierces into the essence of phenomena (ontologically, offering an account of being) by first considering its outward manifestation (ontically, offering an account of beings). That is, ontological investigation begins empirically. In Heidegger's view, a researcher can know when they have uncovered the essence of a phenomenon because their account will, first, demonstrate the relationships among the elements involved and, second, be rich, detailed and coherent.

Though, as Budd (2005) describes, some research in LIS over the past four decades has adopted a phenomenological approach partially, a truly phenomenological (and hermeneutic) study would entail thick description of a naturalistic, discursive event in which information is exchanged, taking into account the actions and thoughts of the beings involved. Such research would, in Budd's (1995) view, allow prior research questions to be investigated more deeply, while revealing new questions that previously went unasked.

Since Budd's (2005) writing, one hermeneutic phenomenological research approach that seems to satisfy these criteria: interpretative phenomenological analysis (IPA), a methodology based around semi-structured interviews with a handful of participants. IPA originated in health psychology as a way to seek understanding of people's life experiences (Smith, Flowers, & Larkin, 2009), and it has since been applied in a number of studies in LIS, as documented by VanScoy and Evenstad (2015).

IPA seems to have been adopted uncritically in LIS for hermeneutic research, but some issues should be raised. As developed by Smith et al. (2009) and employed by researchers in LIS (e.g., Gorichanaz, 2016; Latham, 2014; VanScoy & Evenstad, 2015), IPA attempts to combine elements of Husserl's reflective phenomenology with Heidegger's hermeneutic phenomenology. For example, Smith et al. counsel researchers to search for commonalities and differences among participants; but differences would, presumably, be ontic in nature (that is, based on idiosyncratic characteristics of specific entities) and thus possibly lead the researcher astray from the ontological essences that the hermeneutic endeavor is concerned with. Moreover, Smith et al. advocate bracketing, a notion of Husserl's that involves setting aside one's preconceptions in order to analyze a phenomenon as it is on its own, but this seems irreconcilable with the perspective of hermeneutics, which sees one's preconceptions as inescapable and, indeed, valuable.

A hermeneutic study of the self could benefit from the discourse and guidance developed around IPA while overcoming these limitations. As such a study would have only one participant, the question of similarities and differences among participants would be moot. Indeed, Smith et al. (2009) advocate for single-case IPA studies in cases that warrant the depth of analysis that such studies can allow. Though Smith et al. do not elaborate on this, it can be surmised that these situations include those proposed by Yin (2014) as times when a single-case research design is reasonable: the critical case, in which a single case can confirm, challenge or extend an extant theory; the unusual case, in which an opportunity arises to study a rare phenomenon; the common case, in which cases share so many commonalities that comparison is unnecessary; the revelatory case, in which the opportunity arises to study something previously unable to be studied, offering an initial contribution to the literature; and the longitudinal case, which allows a view of change over time. The second issue can be overcome with a deeper commitment to the principles of hermeneutic phenomenology. For instance, rather than "bracketing," a hermeneutic phenomenologist might follow the methodology of Heidegger (2010) and begin by exploring the ontic characteristics of a phenomenon in light of their personal standpoint and then, considering the characteristics of that standpoint, penetrate into the ontological nature of that phenomenon. Heidegger, for example, approaches death ontologically by beginning with the ontic cultural discourse around the concept. Furthermore, the hermeneutic study of the self could be beneficial in other ways: IPA is further limited by the researchers' inability to have directly experienced the participants' experiences, which can also be overcome through the study of the self. Taking all this into account, the following section articulates an approach to the hermeneutic-phenomenological study of the self.

Auto-hermeneutics

Originarily, hermeneutics involved the interpretation of ancient or classical texts. In its modern incarnation as an empirical, phenomenological research approach, hermeneutics has tended to involve the interpretation of texts in the form of interview transcripts or participants' experiential accounts (van Manen, 1997, 2002). However, hermeneutic phenomenology need not involve texts supplied by an other, as Heidegger (2010) demonstrated in his hermeneutic phenomenological account of human being, which was based largely on his own life experience. More recently Dennett (1992) has argued likewise:

It is also possible for a person to engage in auto-hermeneutics, > interpretation of one's self, and in particular to go back and think > about one's past, and one's memories, and to rethink them and rewrite > them. (pp. 5--6)

Certainly, however, a number of precautions must be taken for auto-hermeneutics to be reasonable as a research approach in the social sciences. In the view of Benjamin (2014), a researcher engaging in auto-hermeneutics must have a capacity for self-awareness, have a concrete way to externalize one's inner experience and be trained in qualitative research, echoing the precautions regarding conducting autoethnography offered by Ellis and Bochner (2000). Cultivating self-awareness is beyond the scope of this paper (though counsel here is offered by van Manen, 1997, who suggests that the very act of writing induces a reflective attitude), as is general training in qualitative methods. Regarding a concrete way to externalize one's inner experience, however, a number of things can be said.

Data Collection and Analysis

Auto-hermeneutics can draw data collection techniques from systematic self-observation (Rodriguez & Ryave, 2002) and analytical techniques from IPA (Smith et al., 2009). These sources were chosen because they are well-suited to the task and both have engendered a broad literature of examples, guidelines and critiques. Of course, the particular methods for data collection and analysis, and the nature of the data itself, will depend on the research questions and objectives; researchers might also find suitable inspiration in the literature regarding self-study and autoethnography, as intimated above.

Systematic self-observation (Rodriguez & Ryave, 2002) stems from the insight that many experiential phenomena can only be observed by the person having the experience. However, many aspects of any experience are taken for granted and thus go unnoticed, unremembered and unrecounted. This explains some of the difficulty that researchers have had with using interviews and surveys in attempting to collect such data after an experience has occurred, and it highlights an opportunity for an automethodological approach to studying personal experience. Rodriguez and Ryave suggest two specific methods for bringing tacit phenomena to light through self-observation: interval recording (e.g., time-based) and free-format narrative recording. A key objective of both methods is to reduce the distance between occurrence and data collection, thus leading to data that is more accurate, vivid and free from the transformations of faulty memory.

As for data analysis, Smith et al. (2009) describe the inductive, iterative process to be used with IPA. First, transcripts are made from all audio data, and each transcript is read several times. Open coding is then applied to segments that seem to reflect aspects of the participant's experience of the phenomenon under study. In subsequent coding cycles, codes are amended in order to reflect emerging patterns, and then abstracted into more general themes that reflect the individual's experience. These themes can then be considered for their interrelatedness, one of Heidegger's (2010) tests for the primordiality of the determined essences, as described above. Most published IPA studies do not seem to make interrelation of the themes central in the findings; one recent example in LIS that does this is offered by Gorichanaz (2016), who describes the information experiences of Bible readers. The IPA methodology described by Smith et al. (2009), though designed for semi-structured interview data, is flexible enough to be useable with any textual data---and perhaps even other forms of data entirely. Further inspiration can be drawn from the newly (and heavily) revised Qualitative Data Analysis: A Methods Sourcebook (Miles, Huberman, & Saldaña, 2014). Indeed, when exploring a new area of inquiry using a novel methodology, a researcher is limited only by their (disciplined) imagination (Alvesson & Kärreman, 2011).

The Value of Auto-hermeneutics

Auto-hermeneutics would be advisable for researchers who, for example, have had information experiences that have yet to be documented in the literature (the revelatory case of Yin, 2014). It may also reveal novel perspectives on contexts that have already been studied. In this way, new research avenues can be explored dextrously, paving the way for further research opportunities and deepening our collective understanding of the world of information.

Auto-hermeneutics differs from autoethnography in that auto-hermeneutics generally seeks to characterize a discrete phenomenon through one person's experience, whereas autoethnography seeks to paint a picture of the amalgamate experience of living in a culture (Chang, 2008). The boundary, however, is admittedly a diffuse one; an auto-hermeneutic study could conceivably consider a phenomenon that develops over a span of time or even a set of phenomena, and elements of the researcher-participant's culture might arise as important aspects of that experience. The focus of auto-hermeneutics is a person's conscious experience rather than their culture, but lived experience and culture are sometimes inseparable. An example of this is illustrated in the autoethnographic study of Closet--Crane (2014). Though she describes her work as autoethnographic, the study does not seem engage deeply with the "culture" aspect that characterizes autoethnography and seems to be more phenomenological in nature (in the vein of the lived-experience written accounts given by van Manen, 2002).

In keeping with the recent trend toward collaborative automethodologies, collaborative auto-hermeneutics has also been proposed, though not by that name. Brown and Duke (2005) present a phenomenological self-study in which two researchers collaborated. Typical of self-study in its critical nature, their investigation was aimed at improving collaboration across disciplines to facilitate knowledge transfer in teaching and research in rural Alaska. They found the collaborative, phenomenological approach to self-study to allow for deep personal reflection while also stimulating open conversation among researchers and educators. Such a collaborative approach may also prove useful even for research applications that are not critical in nature.

Assessing Quality in Auto-hermeneutics

Determining the quality of qualitative research can be challenging---especially for single-case studies, and all the more for studies of the self. Scholars have traditionally assessed research according to the values promulgated by the natural science tradition: generalizability, validity and reliability (Maxwell, 2005). These values, however, originated from a positivist world view which may seem incompatible with qualitative research that operates within a different paradigm. Still, qualitative research must be assessible and assessed according to some standards. To this end, some value can still be found in the notions of generalizability, validity and reliability, though they must be conceptualized somewhat differently than they are within positivism. Another concern, aside from these three, which is critical in human research, is ethics. This section considers issues surrounding generalizability, validity, reliability and ethics in auto-hermeneutic research.

Generalization is often conflated with statistical generalization, which is not the goal of any single qualitative study. This is because, outside the positivist paradigm, it is not assumed that a single truth exists, so the notion of generalizing findings to an entire population is impractical. However, qualitative studies---even those involving single cases---can strive for analytical generalization, as described by Yin (2014). In this way, findings from qualitative research should be assessed for their transferability to other specific cases---and perhaps other contexts---on a case-by-case basis (Maxwell, 2005). Thus the value of a given qualitative study rests largely on the detail in which the context and cases are described. To be sure, a single qualitative study can still present a theoretical contribution. Still, there is a temptation to assume that the fewer participants a study has, the smaller its possible contribution. This is not necessarily true; even the study of a single case can reveal significant theoretical findings on its own (Yin, 2012, 2014). Moreover, several single-cases can be considered in concert to formulate more general theories. A strategy for this type of theory-building from multiple case studies is presented by Eisenhardt and Graebner (2007). It is important to note that such theory-building can only occur if single-case studies are recognized for their value in this regard.

It is true that any conclusions drawn from an automethodological study may be extremely limited in their transferability because of individual idiosyncrasy. Moreover, studying only a single individual may not reveal all the potential findings that using a wider sample would. Still, single-case research offers an accessible way to research new contexts, as discussed above, as well as the opportunity for thorough documentation of a particular context. Because of this depth, findings from an automethodological study may spring forth that would not have emerged in a group study. Researchers of the self, then, should be concerned with presenting a narrative that is authentic---true to the self---and doing so with as much relevant detail as possible.

Validity is often dismissed by qualitative researchers when understood as reflective of a single, objective truth. Indeed, this sort of validity is incompatible with the assumptions of most qualitative paradigms. Instead, qualitative researchers should view validity as reflective of trustworthiness (Riessman, 1993) and verisimilitude (Bruner, 1986). Riessman argues that researchers can improve the trustworthiness of their work by detailing how they go about collecting and interpreting the data, and by making primary data available to other researchers when possible. When research achieves this goal, readers should experience resonance with the written account of the study, which is known as the "phenomenological nod" (van Manen, 1997). Another way to conceptualize validity is whether the study measures what the researcher claims was measured (Babbie, 2007). Two forms of validity described by Babbie that are relevant for auto-hermeneutics are face validity---surface-level common sense of the findings---and construct validity---demonstrating logical relationships among the concepts under consideration. Both these forms of validity are necessary for a study to be analytically generalizable.

Offering a slightly different perspective, Maxwell (2005) discusses validity in terms of the extent to which the researcher has given alternative explanations a fighting chance. This suggests an interpretative dialectic, which is compatible with the nature of hermeneutic phenomenology. Maxwell encourages research designs that identify validity threats, emphasizing that such threats are heavily contextualized in qualitative research and therefore must be specific rather than general. One example of a validity threat he names is researcher bias---selecting data that support a preconceived theory, rather than letting the theory emerge from the data idiographically---which would be particularly pernicious for auto-hermeneutics.

Auto-hermeneutics seems ripe to achieve these values of validity. Indeed, the methodology itself offers a distinct advantage regarding validity: It eliminates the possibility for misinterpretation of the participant's account, which is one validity threat to phenomenological research in general. Beyond that, it is the prerogative of the individual researcher to collect data and conduct analysis in a sound manner. There are many methods researchers can use to ensure validity in auto-hermeneutics. Maxwell (2005) suggests collecting rich data, searching for discrepant evidence, triangulating with data from multiple sources and modalities, comparing findings between data sources and to those in the literature that may be applicable, and carefully and completely describing the context and background, values, etc., of the researcher. Holt (2003) lists pitfalls to avoid in conducting autoethnography, some of which are relevant to auto-hermeneutics. In particular, he cautions against overemphasis on narrative rather than analysis and excessive reliance on memory as a data source. Finally, researchers embarking on an auto-hermeneutic study should first ensure that their research questions are appropriate for the methodology; research questions should seek to explore elements of experience, not necessarily search for why or how something occurs, and they should focus on the experience of an individual rather than a group.

Reliability, in terms of data collection, is "a matter of whether a particular technique, applied repeatedly to the same object, yields the same results each time" (Babbie, 2007, p. 143). In qualitative research with human participants, the prospect of studying the "same object" is dubious; therefore, it may be more productive to ask whether similar results would emerge from a repeat study of a similar object. Reliability can also be considered in terms of data analysis: If the data from a study was analyzed anew, would the same findings emerge? It would seem unlikely that the exact same findings should emerge, but they certainly should not be contradictory with the original findings.

Finally, auto-hermeneuticists must ensure that they are conducting their research ethically. Even though no outside participants are involved, a researcher's institutional review board should always be consulted during the design phase of an auto-hermeneutic study. For instance, as Holt (2003) points out, automethodological data collection activities and write-ups may refer to other people; because the main participant is identified as the researcher, the identities of these other people may be revealed by implication (especially if they are family members or associates with distinctive characteristics). Every effort should be made to protect the privacy and dignity of these unwitting participants.

Bearing all this in mind, it should be noted that auto-hermeneutics is not cast in iron. There is no single, step-by-step way to conduct auto-hermeneutic research. Indeed, auto-hermeneutics may best be thought of as a research approach rather than a strict methodology. Researchers adopting this approach, then, should seek to adapt the principles of hermeneutic phenomenology to their specific research questions. In this regard, it is vital that researchers keep ever in mind the philosophical grounding of the hermeneutic perspective rather than applying concepts, ad arbitrio and uncritically, from incompatible frameworks (Osborne, 1994).

Example of an Auto-hermeneutic Study on the Ultrarunning Information Experience

Ultrarunning is one of my most impassioned hobbies. It involves running significant mileage in training and periodic participation in ultramarathons, which are events that involve running or walking longer than 26.2 miles. Ultrarunners face many challenges---navigational, nutritional, medical, psychological and more---which necessitate ongoing information access and use.

In 2015 I ran my first 100-mile race. It should not be surprising that the prospect of covering that distance on foot, especially for the first time, carries a lot of uncertainty. As the event was approaching, I decided to take advantage of the opportunity to conduct research. How would I experience information during the 100-mile race? What sources would I turn to for information while running? How would I choose and perceive those sources? What information sources would I take for granted? Because these questions were exploratory in nature, dealt directly with aspects of human experience and concerned my own personal experience, I adopted auto-hermeneutics¹ for this study. While I was designing the study, I drew principles from autoethnography, systematic self-observation and interpretative phenomenological analysis, as described above.

In order to identify potential validity threats, I primarily used the memoing techniques described by Maxwell (2005). To further ensure validity, I adhered to the principles of systematic self-observation (Rodriguez & Ryave, 2002), collecting data in the moment at prescheduled time intervals and composing my recollections as soon as possible following the experience. While doing so, I was careful not to introduce afterthoughts (my own later reinterpretations of the events) in the report, but rather represent my thoughts as they were in each moment as much as possible. I referred to the surveyed literature during both my conceptual modeling and my data analysis as a strategy for triangulation. The threat of researcher bias was curbed by ensuring that all codes and findings were grounded in the data, as well as searching for counter-evidence during my analysis. Drafts of the manuscript were reviewed by other ultrarunners, including one trained in qualitative research methods, as a way of ensuring external validity.

The methods of data collection included memoing before and after the event (to systematize and incorporate my own personal experience at those times), self-interviews during the event and free-format narrative recording after the event. Memoing was used before and after the race. To conduct the self-interviews, I scheduled alerts in my smartphone to prompt me every 45 minutes, beginning one hour into the race, with one of two alternating open-ended questions. These questions were: (1) What is a recent problem you had, and how did you deal with it? and (2) What are you thinking about right now? Answers to these questions were recorded on the run as voice memos on my smartphone.

In practice, this method was imperfectly successful. It was raining heavily for the first 10 hours or so of the event, which made it sometimes impossible to use my smartphone. At other times, I did not sense the alert going off. All said, I collected 22 voice memos, representing well-distributed points from the first hour to the twenty-seventh hour of my run. (I completed the course in 31 hours, 41 minutes.) To compensate for the uncollected data, I sought to include as much detail as possible in my post-race narrative. I wrote the majority of the narrative the day after the event, and I continued to add detail to the account over the following days as further details came to mind. In composing this narrative, I employed the phenomenological writing techniques described by van Manen (2002). I referenced my voice memos as a means of jogging my memory about certain segments of the event.

In analyzing my data, I followed the IPA protocol outlined by Smith et al. (2009) as described above. In my analysis, I first read the data multiple times to get an overall sense of it, and then I conducted an in-depth close reading, during which I open-coded for recurring characteristics. At this stage, some of the codes included planning, weather, looking forward, worry, overwhelm, disassociation, ambiguity, training, knowledgebase, other runners, and more. As I conducted further cycles of analysis, it became clear that some of these initial codes were connected; worry and overwhelm, for example, were both examples of negative affect regarding my prospects of finishing the race; and training, knowledgebase and other runners were all examples of information sources for me during the run. This process was iterative and occurred over the span of several days. In the end, my findings revealed that my chief source of information on the run was my own body, along with a knowledge base that I'd built up through training, collecting lore and event-specific planning. Based on this, I was able to develop a conceptual model of how my mental states, corporeal information (drawn from my body) and knowledge base interfaced as I ran (removed for blind review). Such an ontological account of the ultrarunning information experience would not have emerged had I taken an autoethnographic (which may have emphasized the self--group interplay of the ultrarunning small world) or self-study (which may have emphasized routes for improving my running practice) approach, and this model can be tested and applied in other situations of individual performance and then, from there, extended further still.

This study can be assessed according to the quality criteria discussed above. In describing my case, data collection techniques and analysis in detail, I have maximized the transferability (though this can only be truly judged by others). Regarding validity, my single-case design was warranted as this was a revelatory case of a novel context for LIS research, and I considered validity threats as described by Maxwell (2005) in designing it. Moreover, I shared drafts of this work with fellow ultrarunners (one of whom holds a Ph.D. in psychology and was trained in qualitative methods), who gave it the "phenomenological nod," echoing that it reflected their own experiences running ultramarathons. The reliability of my study remains to be seen, as follow-up research has yet been done, but nascent evidence appears in the similarity of the model I developed in my findings to that of Groth (2015), who studied her emotional and mental experience practicing the craft of pottery. Finally, to speak of ethics, I worked with my institutional review board to ensure that my study design would be sound from an ethical perspective, and I conformed to all other institutional requirements. In all these considerations, I strove to conduct the highest-quality auto-hermeneutic study possible.

In closing, it is worth remarking specifically on how an auto-hermeneutic study can contribute to knowledge in human information behavior. This can be apprehended through considering how its findings validate, challenge and extend models that have been proposed based on prior research, as models constitute the key theoretical development of a discipline (Fisher, Erdelez, & McKechnie, 2005). Such a comparison can be done between mine and any number of proposed models, depending on a researcher's particular interests and needs, but as an illustration the auto-hermeneutic study I've described can be seen as offering nuance to what many models of information behavior (reviewed by Wilson, 2010) simply refer to as information selection and processing. For instance, the model of Wilson (1981) goes directly from successful information seeking to information use; the model presented in my study offers color to the processes between these steps in the context of ultrarunning (this stage is called "information processing" in Wilson's revised 1996 model; Wilson & Walsh, 1996).

Another model to which this work contributes is Savolainen's (2008) model of information source preference, which frames how information sources are ordered within an actor's perceived information environment. Savolainen called for comparative data to validate his model; my study offers such data. My findings are consistent with Savolainen's in that human sources were the most highly preferred, but my study offers the further nuance of considering one's own self as a critical information source, prior to other human sources. Further, while Savolainen's participants referred to the Internet and other (e.g., print) sources, I did not go to the Internet or other sources during my run to solve problems, even though I had my smartphone on my person and doing so would have been possible---suggesting some particularities of the ultrarunning context. Another way in which my findings diverge from Savolainen's model is that in his model participants used criteria regarding information content as the most important. In my case, however, availability and accessibility of information seem to have been more important.

Conclusion

As the above example illustrates, auto-hermeneutics affords an accessible way to study information experiences in contexts that have not yet been explored from the perspective of information science. Auto-hermeneutics also provides access to participants that may otherwise be unreachable; as in my illustration, auto-hermeneutics offered the scholarly world a glimpse into the information processes at play in the lifeworld of an ultrarunner.

Regarding the usefulness of their findings, auto-hermeneutic studies may, in consideration with findings from other studies, contribute to theory-building. Above I sketched how the findings from one auto-hermeneutic study can allow reconsideration of extant models in human information behavior. Beyond their usefulness in this regard, auto-hermeneutic studies serve to identify viable and fruitful avenues for further research. And for those seeking to improve their own practice rather than contribute to the scholarly literature, auto-hermeneutics offers an economical and flexible way to address important questions.

Furthermore---though, perhaps, outside the realm of research proper---reading auto-hermeneutic studies may be sublime, poignant and inspiring in ways that reading "traditional" research is not. In the case of my auto-hermeneutic study, a number of readers contacted me after reading it, saying that it inspired them to take up running, offered them tips for managing information in their own athletic pursuits or helped them reframe stressful challenges they faced in their own lives. Though these results were unintended and not necessarily scholarly in nature, they demonstrate the wide-reaching value that auto-hermeneutic research can have. Any well-written phenomenological study should be resonant on some level, but auto-hermeneutics seems to invite an even deeper level of intimacy, which can, for readers, lead to greater self-understanding. As Lao Tsu wrote, "Knowing others is wisdom; knowing the self is enlightenment" (33).

References

Allen--Collinson, J. (2012). Autoethnography: Situating personal sporting narratives in socio-cultural contexts. In K. Young & M. Atkinson (Eds.), Qualitative research on sport and physical culture (Research in the Sociology of Sport, Volume 6) (pp. 191-212). Bingley, UK: Emerald Group Publishing.

Alvesson, M., & Kärreman, D. (2011). A methodology of sorts for theorizing from empirical material. In Qualitative Research and Theory Development: Mystery as Method (pp. 57--77). London: Sage Publications.

Anderson, T. D., & Fourie, I. (2015). Collaborative autoethnography as a way of seeing the experience of care giving as an information practice. Information Research, 20(1). Available at http://InformationR.net/ir/20-1/isic2/isic33.html

Babbie, E. (2007). The practice of social research (11th ed.). Belmont, CA: Thomson Wadsworth.

Bates, J. A. (2004). Use of narrative interviewing in everyday information behavior research. Library & Information Science Research, 26, 15--28.

Bates, M. J. (2010). Information behavior. In M. J. Bates & M. N. Maack (Eds.), Encyclopedia of library and information sciences, third edition (pp. 2381--2391). New York: Taylor and Francis.

Bawden, D. (2016). The noblest pleasure: Theories of understanding in the information sciences. In D. H. Sonnenwald (Ed.), Theory development in the information sciences (pp. 283--299). Austin, TX: University of Texas Press.

Benjamin, E. (2014). Life after death: An experiential exploration with mediums by an agnostic investigator. Lulu Publishing Services.

Bromley, D. B. (1986). The case-study method in psychology and related disciplines. Chichester: Wiley.

Brown, J. D., & Duke, T. S. (2005). Librarian and faculty collaborative instruction: A phenomenological self-study. Research Strategies, 20(3), 171--190.

Bruce, C., Davis, K., Hughes, H., Partridge, H., & Stoodley, I. (Eds.) (2014). Information experience: approaches to theory and practice (library and information science, volume 9). Bingley, UK: Emerald Group.

Bruner, J. (1986). Actual minds, possible worlds. Cambridge, MA: Harvard University Press.

Budd, J. M. (1995). An epistemological foundation for library and information science. The Library Quarterly, 65(3), 295--318.

Budd, J. M. (2005). Phenomenology and information studies. Journal of Documentation, 61(1), 44--59.

Case, D. O. (2016). Looking for information: A survey of research on information seeking, needs and behavior (4th ed.). Bingley, UK: Emerald Group Publishing.

Case, D. O., & O'Connor, L. G. (2016). What's the use? Measuring the frequency of studies of information outcomes. Journal of the Association for Information Science and Technology, 67(3), 649--661.

Chang, H. (2008). Autoethnography as method. Walnut Creek, CA: Left Coast Press. 

Chang, H., Ngunjiri, F., & Hernandez, K.--A. C. (2012). Collaborative autoethnography. Walnut Creek, CA: Left Coast Press. 

Closet--Crane, C. M. C. (2014). Listening to shells and discovering a lost world; epiphanic experiences at the museum. Proceedings from the Document Academy, 1(1), paper 9. Retrieved from http://ideaexchange.uakron.edu/docam/vol1/iss1/9

Day, R. E. (2000). The "conduit metaphor" and the nature and politics of information studies. Journal of the American Society for Information Science, 51(9), 805--811.

Dennett, D. C. (1992). The self as a center of narrative gravity. In F. Kessel, P. Cole, & D. Johnson (Eds.), Self and consciousness: Multiple perspectives. Hillsdale, NJ: Erlbaum.

Eisenhardt, K. M., & Graebner, M. E. (2007). Theory building from cases: Opportunities and challenges. Academy of Management Journal, 50(1), 25--32.

Ellis, C., & Bochner, A. P. (2000). Autoethnography, personal narrative, reflexivity. In N. K. Denzin & Y. S. Lincoln (Eds.), Handbook of qualitative research (2nd ed.). Thousand Oaks, CA: Sage.

Fidel, R. (Ed.). (2012). Human information interaction: An ecological approach to information behavior. Cambridge, MA: The MIT Press.

Fisher, K. E., Erdelez, S. & McKechnie, E. F. (Eds.). (2005). Theories of information behavior: A researcher's guide. Medford, NJ: Information Today.

Gorichanaz, T. (2016). Experiencing the Bible. Journal of Religious & Theological Information, 15(1/2).

Graziano, E. E. (1968). On a theory of documentation. American Documentation, 1968(January), 85--89.

Groth, C. (2015). Emotions in risk assessment and decision making processes during craft practice. Journal of Research Practice, 11(2), Article M5. Retrieved from http://jrp.icaap.org/index.php/jrp/article/view/502/424

Guzik, E. (2013). Representing ourselves in information science research: A methodological essay on autoethnography. The Canadian Journal of Information and Library Science, 37(4), 267--283.

Harding, S. (2015). Objectivity and diversity: Another logic of scientific research. Chicago: The University of Chicago Press.

Heidegger, M. (2010). Being and time (trans. J. Stambaugh). Albany, NY: State University of New York Press. (Original work published 1927)

Holt, N. L. (2003). Representation, legitimation, and autoethnography: An autoethnographic writing story. International Journal of Qualitative Methods, 2(1), 18--28.

Kari, J. (2007). Conceptualizing the personal outcomes of information. Information Research, 12(2), paper 292. Retrieved from http://www.informationr.net/ir/12-2/paper292.html

Kvanvig, J. L. (2003). The value of knowledge and the pursuit of understanding. Cambridge: Cambridge University Press.

Lassonde, C. A., Galman, S., & Kosnik, C. (Eds.). (2009). Self-study research methodologies for teacher educators (Professional Learning, Volume 7). Rotterdam, The Netherlands: Sense Publishers.

Latham, K. F. (2014). Experiencing documents. Journal of Documentation, 70(4), 544--561.

Maxwell, J. A. (2005). Qualitative research design: An interactive approach (2nd ed.). Thousand Oaks, CA: Sage Publications.

McAdams, D. P. (2001). The psychology of life stories. Review of General Psychology, 5(2), 100--122.

Michels, D. H. (2010). The place of the person in LIS research: An exploration in methodology and representation. Canadian Journal of Information and Library Science, 34(2), 161--183.

Miles, M. B., Huberman, A. M., & Saldaña, J. (2014). Qualitative data analysis: A methods sourcebook (3rd ed.). Thousand Oaks, CA: Sage Publications.

Mills, J., & Birks, M. (Eds.). (2014). Qualitative methodology: A practical guide. Thousand Oaks, CA: Sage Publications.

Osborne, J. (1994). Some similarities and differences among phenomenological and other methods of psychological qualitative research. Canadian Psychology, 35(2), 167-189.

Polkinghorne, S. (2012). Caught in the act: An autoethnographic analysis of the performance of information literacy instruction. In Proceedings of the Annual Conference of CAIS. Retrieved from http://www.cais-acsi.ca/ojs/index.php/cais/article/view/649/408

Riessman, C. K. (1993). Narrative analysis. Newbury Park, CA: Sage Publications.

Rodriguez, N., & Ryave, A. (2002). Systematic self-observation. Thousand Oaks, CA: Sage Publications.

Savolainen, R. (2008). Source preferences in the context of seeking problem-specific information. Information Processing & Management, 44(1), 274--293.

Smith, J. A., Flowers, P., & Larkin, M. (2009). Interpretative phenomenological analysis: Theory, method, research. London: Sage Publications.

Smith, J. A., Harré, R., & van Langenhove, L. (1995). Idiography and the case-study. In J. A. Smith, R. Harré, & L. van Langenhove (Eds.), Rethinking psychology (pp. 59--69). London: Sage Publications.

Suzuki, S. (1970). Zen mind, beginner's mind. New York: Weatherhill.

Vamanu, I. (2012). Hermeneutics: a sketch of a metatheoretical framework for library and information science research. Information Research, 18(3). Available at http://InformationR.net/ir/18-3/colis/paperS08.html

Van Cleave, K. (2008). The self-study as an information literacy program assessment tool. College & Undergraduate Libraries, 15(4), 414--431.

van Manen, M. (1997). Researching lived experience: Human science for an action sensitive pedagogy (2nd ed.). Albany, NY: State University of New York Press.

van Manen, M. (2002). Writing in the dark: Phenomenological studies in interpretive inquiry. London, Ontario: The Althouse Press.

VanScoy, A., & Evenstad, S. B. (2015). Interpretative phenomenological analysis for LIS research. Journal of Documentation, 71(2), 338--357.

von Herrmann, F.--W. (2013). Hermeneutics and reflection: Heidegger and Husserl on the concept of phenomenology (trans. K. Maly). Toronto: University of Toronto Press. (Original work published 2000)

Wilson, T. D. (1981). On user studies and information needs. Journal of Documentation, 37(1), 3--15.

Wilson, T. D. (2010). Information behavior models. In M. J. Bates & M. N. Maack (Eds.), Encyclopedia of Library and Information Sciences, third edition (pp. 2392--2400). New York: Taylor & Francis.

Wilson, T. D., & Walsh, C. (1996). Information behaviour: An interdisciplinary perspective. Sheffield, UK: University of Sheffield Department of Information Studies.

Yin, R. K. (2012). Applications of case study research (3rd ed.). Thousand Oaks, CA: Sage.

Yin, R. K. (2014). Case study research (5th ed.). Thousand Oaks, CA: Sage.