Methodology

2.1 The theoretical background to ethnography

Ethnography investigates social groups and makes the ethnographer’s interpretation of findings accessible to members of other social groups. In common with all methods of investigating aspects of the world, at its foundation ethnography has various assumptions about the nature of reality, and about the relationship between the researcher and that reality. Ethnographers assume that their subjects (social groups) have unique understandings and impressions of the world that inform their behaviour, their interactions with others and their attitudes towards the world. This is not to say that ethnographers deny the existence of a world of shared “objects, events and processes” (Guba and Lincoln, 1988: 81), but that this world of ‘things’ does not compare in significance to the world of constructed ‘facts’ when studying groups of people. Believing that the world of facts is more interesting than the world of things, ethnographers place themselves firmly in the research paradigm known variously as the naturalistic (Guba and Lincoln, 1988), the interpretive (Geertz, 1973) or the constructivist (Williamson, 2006) paradigm (see Lincoln and Guba, 2000, for a discussion of these and various other names for this paradigm). Although they do describe differing approaches to research, ethnography can be seen as exhibiting aspects of all of these approaches (Williamson, 2006: 83-84) and they shall be referred to as ‘interpretive paradigms’ in this dissertations, after Geertz, 1973.

As well as their assumption as to the existence and importance of differing interpretations of the world, methodologies informed by the interpretive paradigms are also united by an assumption about the role of the researcher. Traditional scientific research is located within the ‘positivist’ paradigm and holds that the researcher is an instrument which designs experiments to test hypotheses, records the results from these experiments, and reasons from these results to larger generalisations. Researchers in the interpretive paradigms have a more “messy” view of their relationship with the data (Guba and Lincoln, 1988: 82). Interpretivists embrace the fact that all representations of data are constructed from someone’s (the researcher’s) experiences. Ethnography in particular emphasises the crucial role that the humanity (the non-mechanical nature) of the researcher plays in the collection and presentation of data. The ethnographer accepts that the human mind is fallible, easily moved, error prone and extensively forgetful, but argues that the insight and synthetic abilities that the prism of another human mind can apply to the facts of someone else’s world are entirely invaluable and irreplaceable. What an ethnographer presents is his or her “own construction of other people’s constructions of what they and their compatriots are up to” (Geertz, 1973: 9).

The central role given to the researcher’s interpretations in ethnographic studies clearly raises questions about the objectivity, the validity and the authenticity of ethnographic findings. Some authors have argued for the replacement of these typically positivist terms with other concepts that better convey the nature and the strengths of interpretive research, such as “dependability”, “openness” and “fairness” (Guba and Lincoln, 1988: 84-85). When applied, ‘dependability’ means that ethnographers must explicitly record and keep referring to his or her method; ‘openness’ means that the ethnographer must be honest to the point of ‘nakedness’, with clearly no embellishment or fabrication of details, and a frank exposure of his or her own feelings and impressions; and ‘fairness’ means that the ethnographer must ignore any personal prejudice in reporting the data. By occupying a privileged position inside the group to the studied, the ethnographer has access to events and occurrences of interest. Through a thorough application of principles of the method, a successful ethnography can be produced if the ethnographer succeeds in “fixing” his or her impressions of these occurrences and turning them into an account which others can experience (Geertz, 1973: 19). The fully embraced importance of the researcher’s own interpretation, the privileged position of the researcher and the constant reference to method make ethnography “less vulnerable than empirical [sociology] to demands for positivist legitimacy in methods” according to Okely, 1994: 19.


The rest of this section will explain what ethnography involves at a practical level, outline some relevant applications of this method, explain why this method was used and any ethical or epistemic concerns raised by its use, and the precise research procedure will be described.

2.2 What ethnography involves

Ethnographies provide “rich and contextualised understandings” of social situations (Smith, 2001: 220). The process by which these understandings are reached is an involved and complex one. Any separations that this account projects onto the ethnographic research process are only present for pragmatic reasons, and in reality, ethnography is a thoroughly integrated and holistic process.

‘Ethnographic field work’ describes what the ethnographer does when he or she is directly exposed to the raw data of the study. The first step is to decide upon a subject for study, and then to secure entry into this social group – a feat that is treated briefly here purely because the author’s entry into the staff of the Information Commons was not difficult, but many authors go to extraordinary lengths to gain access to groups which are initially unwilling to admit researchers into their midst: see Smith, 2001: 220-221. The ethnographer must then “penetrate” the group (Goffman, 1989: 129) and immerse him or herself in the activities and day to day lives of those being studied. Geertz asserts that only by close and prolonged exposure to a group will the unexpected and chance events occur which reveal the defining characteristics of the group’s view of the world (Geertz, 1973a: 412-413). This exposure and involvement in the group should occur over a very long period of time: Goffman states, “I think you should spend at least a year in the field” (Goffman, 1989: 130), and Okely describes a study in which researchers lived for the duration of an entire generation with their subjects in order to learn about attitudes towards inheritance (Okely, 1994: 23). This degree of interaction with a social group is intended to give the ethnographer a “deep familiarity” (Goffman, 1989: 130) with the conventions, norms and “piled-up structures of implication” that define the world as the subject sees it (Geertz, 1973: 7).

What the ethnographer does during the field work stage is not spelled out in any great detail in the literature and differs greatly depending upon the group to be studied. It can involve the collection of a great range of types of data, but the essential features of the research are that it must not be intrusive or entirely dependent upon the creation of artificial situations like interviews (although it can involve interviews); it must collect data about the ‘actual’ lives or work patterns of the subjects, so it must not rely solely on second hand accounts of what occurs in the group; and it must ignore no details, however trivial they may seem at the time (Okely, 1994; Emerson et al, 2001; Baker, 2006).

There is a “freshness cycle” involved in ethnographic research which means that the researcher will notice more on his first day in the field than on his second day and so on because what is strange early on in the research process will become expected and un-noteworthy as time progresses (Goffman, 1989: 130). Thus, it is essential that some record of these events is made to fix them in the mind of the ethnographer, and this record takes the form of ‘field notes’. Considerable debate exists as to the way field notes should be taken and as to what their content should be (see Emerson et al, 2001). What is widely agreed is that they should be as “thick” as possible (Geertz, 1973), with as much detail of “everything [the researcher happens to] hear, smell and see; even the colour of the carpets” (Okely, 1994: 23). Field notes are written whilst the researcher is in the midst of the ‘field’, and little attention is paid to style or even legibility, with notes often being “incomprehensible to onlookers” (Emerson et al, 2001: 356). But a record will then exist of what the researcher experienced as he or she experienced it – the notes will be a unique record of the ethnographer’s interpretation of other people’s interpretations of events as close to them happening as possible.

Once the ethnographer has completed their time in the field they will commonly transcribe their field notes. This is not a process that simply provides a bridge to other stages of the research, but rather a centrally important task that widely influences the rest of the project (Okely, 1994: 24). The process of writing up what was experienced allows themes to emerge, interpretations to be extended and entirely new understandings of what occurred to be incorporated. What allows ethnography to become the powerful tool it can be is the way in which the gathering of data, the analysis of the data and the writing up of this data becomes one interdependent process. Only by being disponible (Okely, 1994: 19) to new ideas and interpretations of the data can the ethnographer ensure that they are actually reporting “the structures of significance” as they see them (Geertz, 1973: 8). The layers of interpretation that the ethnographer adds to the interpretations of others are crucial to record, and emerge gradually as the data is examined. Okely recommends dividing up the data into overlapping categories, with some observations placed in many different categories if necessary (Okely, 1994: 23). As the ethnographer examines the data in the categories, themes that unite it or describe its differences will begin to emerge and these will form the criteria for selecting the theoretical literature to be reviewed – only once the data has been collected can a satisfactory analysis of relevant literature be attempted, if the ethnographer’s disponibility and complete openness to new ideas is to be preserved. This is a counter-intuitive element of ethnography, but as Okely sates, although the ethnographer may conduct extensive reviews of the literature before the field research begins, he or she must be willing to “eject hypotheses like so much ballast” if they are found to not be relevant to interpretations that appear during the field work (Okely, 1994: 19). Only through this rigorously ‘naked’ approach can researchers ensure the honesty and dependability that safeguard the value of ethnographies.

Geertz states that the line between “the mode of representation and the substantive content” of ethnographies is “undrawable” (Geertz, 1973: 16). This implies that the ‘write-up’ of ethnographic findings, the data gathering stages and the coding stages, are all inextricably linked. The ethnographer is concerned to turn his or her experiences into “an inspectable form … an account” (Geertz, 1973: 19). This account ‘fixes’ the ethnographers interpretations so that others can understand them. Although the ethnographic write-up should be understood to be principally the author’s interpretations of what he or she experienced, the ‘world’ they recreate must not be their’s in the sense that the world of Oliver Twist is ‘Dickensian’ (Clifford, 1986: 13). Rather, constant reference must be made to both the field work that occurred (thus locating the study in the world depicted), and the ethnographic method that drives the study (thus justifying the author’s right to make interpretive moves) (Clifford, 1986: 14). It must be emphasised that the research is still occurring as the write-up proceeds. As the author finds ways of expressing his or her findings, so his or her interpretations of the situation become explicated. In the write-up the ethnographer brings other theories and arguments from relevant literature into the analysis of the findings, and so this stage represents the confluence of the external theoretical ideas and internal interpretive findings. It is in the write-up that ethnographies plunge from their sheer beginnings into the deep waters only vaguely charted by other commentators (Geertz, 1973: 25).

Thus, an ethnography can be represented as starting with the field work and progressing through the note taking and transcription phases to the write-up. But this is an artificially linear construction of the integrated, interdependent and holistic process that gives ethnography its strength. The importance of the author’s interpretations guarantees that each stage of the development of these interpretations is equally important and reliant on every other stage. No one part of the process can make any real sense without the others.

2.3 Relevant Ethnographic Studies

This brief review serves as the background for one of the reasons for the use of ethnographic methods, and should not be viewed as the presentation of the theoretical literature used to analyse the findings, which is presented in the Theoretical Literature section later in the dissertation.

The context and complexities that constitute the jobs people perform is a source of interest to many authors because even though “work is a constant part of our lives” (Orr, 1996: 1), it has often been labelled “routine, unskilled… or even trivial” (Smith, 2001: 221). Barley argues that “occupational images” exist which distort our ideas of what it is that occupations actually do (Barley, 1996: 406-407). Only by embracing ethnographic methods can researchers dispel these images and arrive at the genuinely informative descriptions of what people do whilst at work.

Descriptions of the actual work of librarians when studied at a local level and described in ‘thick’ detail are not common. George et al state that “few studies have been done of libraries…using ethnographic methods” (George et al, 2003: 2) and of these few, the most influential and honestly ethnographic is Nardi and O’Day’s study of librarians as intelligent systems at the Apple Reseach Laboratories (Nardi and O’Day, 1996). They found that librarians to perform complex, multi-faceted roles that greatly assisted their users (Nardi and O’Day, 1996: 80). This generous view of the work of librarians is echoed in George et al, 2003, which pictures librarians a “key-stone species” in the knowledge ecology: without librarians all other parts of this ecology could not function. Plum, 1994, expresses this sentiment and states that librarians “create communities” using their “rituals of knowledge” to bind users together (Plum, 1994: 500). Klopfer, 2004, shows how commercial libraries in an Indian city have managed to occupy a niche, created by the short-comings of Indian public libraries, and are thus, greatly appreciated by users.

The small selection presented above represents a large percentage of all the ethnographic studies conducted in libraries, which in its self is a justification for more studies that use the techniques of ethnography to examine the actual work of librarians. All the studies above set out with the purpose of demonstrating the worth of librarians, at the expense of a more even-handed depiction of their work practices, a feature commented on by Sandstrom and Sandstrom, 1995: 175. So, an even-handed ethnography that maintains the author’s disponibility or openness to honest interpretations of the actual state of affairs is called for. This dissertation represents an attempt to satisfy this gap in the literature.

2.4 Additional reasons for the use ethnographic methods

The author noticed significant differences between CiCS and Library staff whilst working at the Information Commons, and from these observations grew this dissertation, and so from the project’s inception, the author’s interpretation of events was at its core. Ethnography is the only method to allow for the author’s interpretation of events to assume such importance, and so it presented its self as an attractive methodologic option.

As a member of the group to be studied, the author and the subject would be particularly closely bound. This raised significant ethical issues (discussed below), and also the epistemic demand that the crucial role of the author’s interpretations in the development of this project be embraced and scrutinised. ‘Participant observation’, a type of ethnographic study (Jorgensen, 1989: 8), allowed for this demand to be met, since like all ethnographic methods, it emphasises the role of the researcher at the centre of the study. Participant observation allows the researcher to completely adopt the weltanschauung, or world-view, of the subjects by literally being one of them (Baker, 2006: 179), and was therefore particularly suited to this study .

The other crucial aspect of this study which demands an ethnographic approach was its subject matter. The working patterns and the interactions of the library and CiCS people were what initially motivated the study, and ethnography is a powerful way of examining the very “local textures” that define people’s work-lives (Smith, 2001: 220). Using a method which took people out of their normal work context (such as focus groups or formal interviews) would divorce the subjects of this study from the context that gave their actions meaning and enabled the initial interpretations of them. In addition, the opening of the Information Commons was intended to signal a new era of cooperation and convergence between Sheffield University Library and CiCS staff. This study of the differences between these two groups could have resulted in a failure to gather useful or insightful data if the subjects of the study were put in situations where they were fully aware that their actions and answers to questions were being used to study how they differed. Only by retaining the workplace context could it be guaranteed that their actions were ‘natural’ and not conditioned or normalised by the organisational changes that had occurred. From an epistemic point of view, this study demands that the natural contexts of work in the IC be preserved, and so ethnography appeared the most fitting methodology for this study.

Thus, the lack of a substantial body of literature that ethnographically addresses the ‘occupational images’ that are widely held of people working in libraries, the fact that the author’s interpretation of events originally prompted the project, the author’s proximity to the staff of the IC, and the fact that the IC staff needed to be studied in their natural work environment, all motivated the use ethnographic methods.

2.5 Epistemic Issues

The central epistemic issue that is raised by this study of the staff of the IC concerns the fact that the author is a member of library staff. This could potentially influence attempts to record and interpret the events and occurrences that are witnessed. It could be that situations will be inadvertently interpreted favourably towards the library staff. Alternatively, the author may over-compensate for his historic library bias by interpreting situations favourably towards the CiCS people. There are two main responses to these possibilities. Firstly, ethnography emphasises the value of the ethnographer’s ‘closeness’ to the phenomena, and so the author’s proximity to the issues and his privileged position within the working environment of the IC should perhaps be viewed as an epistemic boon and not as a drawback to this study. And secondly, the central importance of the ethnographer’s interpretation of events to the value of ethnographies is crucial. Thus, if it is made clear that all the meaning drawn from the events witnessed is ascribed by the author, and that many of the layers of significance will have been added by the author, then it will be possible to produce research that is honest and valuable.

Another epistemic concern raised here that distinguishes this from other studies, is the fact that the author must make the conceptual move from “complete participant” to observer (Baker, 2006: 174). This is in contrast to the majority of studies, where the move is generally in the opposite direction, with the researcher beginning as a “complete observer” and only slowly gaining the insight of a participant. Tedlock (1991) argues that making the epistemic move taken in this study encourages a high level of honesty because as the unique position of the author is made more and more important, so the ability of the readers to triangulate the author’s findings using the work of others becomes more difficult (Tedlock, 1991: 77-78). Thus, a guiding and defining principle in this study must be the honesty that an ethnographic approach demands.

More generally, another epistemic issue raised by the fact of the author’s closeness to the subjects of this study, is that the “freshness cycle” (Goffman, 1989: 130) will be vastly truncated, or might possibly have prematurely expired. This means that many of the ‘unusual’ events that define the IC staff’s view of the world will not appear as such to the author because he is accustomed to them. Perhaps more damaging to the epistemic credibility of this study, it could be that the author is able to recognise unusual events among the CiCS people, but not among the library people. This could result in a study that focuses more on the CiCS people’s behaviour than it does on the Library people’s, whereas this dissertation intends to study both these groups.

The deeply interpretive nature of ethnography as a method and the author’s proximity to the events discussed means the sections in this dissertation in which the ethnographic findings are presented will be written from a first person perspective. This is because third person accounts can be seen as “deceptive” in the presentation of findings from qualitative research – the social interaction implicit in research is obliterated if the ‘I’ is not used according to some authors (Webb, 1992: 749). In addition, ethnographic accounts deal with “complex and ambiguous” subjective situations which require the place of the first person to be not only admitted to but explicitly referenced if they are to make sense (Tedlock, 1991: 78). The importance of ‘the subjective’ will not only be evident from the use of first person vernacular during the presentation of findings, but is also indicated by the structure of this dissertation, because the way this study is organised represents an attempt to honestly record the order that the research process were undertaken and therefore, the progression of the author’s personal engagement with these processes. The structure of this dissertation is outlined below. In keeping with traditional academic rigour and formality, references to the first person will be absent from the rest of this study.

These epistemic issues are discussed with reference to how well they were
resolved in this study, in the ‘Discussion of the Method’ section.
2.6 Ethical Issues

All research that involves human participants raises certain ethical issues about the treatment of those participants. It has been argued that ethnography is actually the most ethically sound of all qualitative research methods because it “does not have human subjects” (Jorgensen, 1989: 28, emphasis in original). Jorgensen states that since the subjects of all ethnographic research are studied under the ordinary conditions of their everyday lives with the minimum of intrusion from the researcher, the ethnographer has “no more or less of an ethical obligation to… [them]… than he or she would have under other everyday life circumstances” (Jorgensen, 1989: 28). However, it seems crude to suggest that participants in ethnographies are not affected at all by research that involves them as subjects, and so some potential impact upon the subjects of this study will be outlined and, the ways in which these impacts have been minimised will be discussed.

The actual process of extracting information from participants could potentially cause embarrassment, shame, anxiety, guilt or damage to self-esteem (Murphy and Dingwall, 2001: 340). However, these effects are felt largely by subjects involved in studies where they are put into unusual settings and asked to answer difficult or complicated questions. In this study, all findings will be drawn from observations conducted whilst the participants are working, from informal conversations between the author and the subjects, or from conversations overheard between the subjects and their colleagues. Thus, since their actions need not diverge from naturalistic behaviour, much of the pressure normally placed on the subjects of research will be avoided in this study.

Ethnography involves close contact with people, meaning that a relationship between the researcher and the subjects inevitable develops. This relationship can be abruptly ended when the research finishes, or it can often result in the subject feeling cheated and objectified by the researcher who could be seen as ‘using them for information’. In this dissertation, the author had worked with the subjects for over a year, and so had already developed a relationship of trust with them. In addition, when the field research was completed contact with the subject’s was not abruptly terminated, thus hopefully, avoiding any feelings of objectification being caused.

The publication of ethnographic research can also adversely affect participants. With the exposure of the views of peers, the existence of unrealised relationships, or the description of hidden social hierarchies, can come feelings of rejection and self-doubt. Studies that radically alter the status quo inside social groups are not considered to be methodologically sound ethnographies (Murphy and Dingwall, 2001: 341). To alleviate ethical concerns along these lines, this dissertation will not be consultable for five years after its completion, and even then, it will not be published.

This prior reflection upon and rigorous adherence to the ethical issues surrounding qualitative research is intended to ensure that none of the subjects in this study were adversely affected by this research. The extent to which this was realised is discussed in the Discussion of Method section. Below, the stages that were undertaken in this dissertation are explicated.

2.7 Research Procedure

As has been remarked upon before, the author’s interpretation of the differences between Library and CiCS staff in the Information Commons is what provoked this study, so an investigation of various methods which could incorporate these interpretations was embarked upon as the first stage in this dissertation. Once ethnography had been settled upon as the most appropriate method (for the reasons listed above), the manager of the IC was applied to for permission to study the staff whilst they worked. This permission was granted in May, so ‘entry’ into the IC was secured. Although it was impossible to formally inform all the people who were observed in the IC since many of them only worked there transiently and only formed fleeting parts of some observations, all the people who were observed closely and all people engaged in conversation by the author were explicitly informed as to the existence of the study.




The author began to take brief notes of any salient differences that were observed from shortly after formal permission had been granted for the study to go ahead. Copies of crib sheets were also collected, as were emails, notices and other objects that might have formed the basis for some comparative analysis.

In the first week of June (the last week of semester, and so the last opportunity to observe the IC staff at their busiest), the author was placed on the Library staff rota to work three full days in the IC (Tuesday-Thursday). The author worked exactly as the full time staff in the IC worked, and so served on the counter for more than half the day, shadowed the IC staff when they ‘roved’, and sat in the staff area and observed whilst the other IC staff performed their ‘off-counter’ duties.

The author did ‘work’ whilst doing the field research because it was a busy period, and the IC were glad of extra help, but simultaneously, what Emerson et al call “jotted notes” were taken (Emerson et al, 2001: 356). Very brief key words and phrases were recorded to serve as reminders of events that seemed significant at the time. These were written in a small, unobtrusive notebook so as not to disturb the flow of the discourse that was occurring at the same time.

At opportune moments (for example at lunch times, or when changes of personnel occurred at the counter), the author would retire to the student café attached to the IC to do the first write up of these notes. This initial write up was conducted away from the staff area so that the author would not to be distracted by other relevant events, and so as not to raise the profile of the study amongst the staff, or compromise the naturalistic attitudes that were required. Initially, the events recorded in the jotted notes were elaborated upon, and this process served to remind the author of countless other events that had occurred during the observation stage. After Geertz (1973) these events were recorded as “thickly” as possible, with as much of the author’s own understanding and interpretation of the events as possible included in the initial write up. This process proved very productive, and would provoke the author to go back to the IC staff with wholly altered ideas as to what counted as significant events to record. The author reflected these changing priorities by putting key words at the top of each page to refer to the subjects that were considered the most significant whilst the observations were made.

After three days of observation, large quantities of hand-written A5 pages had been produced, and these were typed up with any additional interpretations that occurred to the author included in this second stage of the write-up. Several copies of the typed up notes were printed off, with one observed event or sequence of interaction to an A4 sheet. These were then organised into thematic categories, with the same observation going into more than one category if necessary. By reading and re-reading the notes, eventually 5 broad and overlapping categories were decided upon: ‘expertise’, ‘job propriety’, ‘jurisdiction’, ‘elaboration’ and ‘de-skilling’.

At this stage the collection of relevant theoretical literature that might explain or be applicable to the findings began. Originally, literature that sought to explain the differences between library and computer staff was researched, but this was scant and much of it was largely to do with macro-level factors and so not relevant to this local study. The literature on jurisdictional claims proved more fertile, with Andrew Abbott’s The System of Professions (Abbott, 1988) appearing to have particular explanatory power in the context of this dissertation. The literature of expertise was also explored. This research enabled the author to return to the categories that the observations had provisionally been organised into, and discarded observations that did not fall into the two most fertile categories remaining after the survey of the theoretic literature: ‘expertise’ and ‘jurisdiction’ (the observations in the ‘job propriety’ category were found to be highly relevant to the literature to do with jurisdiction, so this and the jurisdiction category were collapsed into one).

The field notes were then written up again using spatial not thematic categories: observations from the whole of the IC; observations from the counter; and observations from the staff area. This was intended to present a more narratologically appealing Findings section, akin to the vignettes that Orr presents (for example, Orr, 1996: 15-23), or the scene setting used by Kunda (Kunda, 1992: 26-30). This write-up of the findings was then used to discuss and apply the theoretical literature to this study. The write-up stages of ethnography illustrate just how holistic the method is. The final form of the write-up of the Findings depended upon the Theoretical Literature because the latter section determines which of the findings will be most fertile for discussion, but what was relevant theoretical literature could not have been decided upon without a clear idea of the themes that were to be present in the Findings section.

The structure of this dissertation is an attempt to honestly reflect the stages of the research as they happened, despite the fact that attempts to represent the process of ethnography in a linear way somewhat ignore the method’s holistic nature. Thus, the method is what was initially researched and is what should be seen to ‘drive’ the rest of the study, so the Methodology section is placed first. The Findings are placed before the Theoretical Literature because they suggested the areas that were researched and were extant before the literature had been extensively explored. The Theoretical Literature needs to be presented before the Discussion of Findings because this latter section exists only in the light of the literature and was produced after the findings and the literature had been collected. The Discussion of Method section can clearly only appear after the use of these methods has taken place. The fact that this dissertation is structured like this also serves to constantly relate it to the methodological principles that ensure it is transparent and open, and therefore, valuable. The impact of the decision to separate the sections of this dissertation in this way is discussed in the Discussion of Method section.