Abstract
Defining and bounding a case are especially important components in designing a case study (Rule & John, 2015, p. 8; VanWynsberghe & Khan, 2007, p. 90; Yin, 2018, p. 27). Discussions concerning case study as a methodology and the relationship between theory and case study research have been featured in the literature for some time now (Rule & John, 2015, p. 3; Yazan, 2016, p. 134). Yet, the specific process of identifying or constructing a case and the contribution of the literature and theory in this context seem neglected. This article conveys some insights on the dilemmas related to the role of the literature and theory in defining and bounding a case.
Throughout the article, I draw upon my experiences as an early career PhD-student. In my PhD project, the general research question is “What interests prevail in Norwegian Svalbard politics and why?” The overall theme is power and interests in Norwegian Svalbard politics. I am interested in investigating what power in Norwegian Svalbard politics is, and how it works. On the surface, Norwegian Svalbard politics may seem rather straightforward, as the overarching policy objectives of the firm enforcement of sovereignty, maintenance of peace and stability, and the preservation of wilderness and viable Norwegian settlements have been consistent since the mid-1980s (Norwegian Ministry of Justice and Public Security, 2016, p. 5). However, conflicts of interest among these expressed objectives and other interests of both state and non-state actors occur. This tension only increases with new possibilities for development and challenges in relation to the globalization of the Arctic, with climate change, new stakeholders, and resource exploitation as key elements. To investigate this puzzle, my PhD project takes the form of a qualitative case study of three ongoing political processes in Norwegian Svalbard politics. In the course of designing my case study, I encountered some dilemmas around defining and bounding the case and in figuring out what role the literature and theory would play in these decisions. This article aims to present insights on such dilemmas based on my experiences.
Briefly put, case studies are detailed investigations of specific phenomena (Andersson, 2020, p. 2; George & Bennett, 2005, p. 5). Designing a case study involves making decisions on the objectives and structure of the study, on how it can be carried out, and on how implications for theory can be drawn (George & Bennett, 2005; Rashid, 2019, p. 2). The first and most important step in designing a case study is to define and bound the case (Rule & John, 2015, p. 8; VanWynsberghe & Khan, 2007, p. 90; Yin, 2018, p. 27).
Defining a case involves identifying the entity to study in order to answer a general research question (George & Bennett, 2005, p. 77; Rule & John, 2015, p. 4; VanWynsberghe & Khan, 2007, p. 85; Vaughan, 1992, p. 176; Walton, 1992, pp. 121–122; Yin, 2018, p. 31). To bound a case is to specify the spatial, temporal, and other explicit features of this entity (Yin, 2018, p. 31). It is to separate the phenomenon being studied from the context. In these processes, the literature and theory can play important roles in providing guidance for what to look for and what the phenomenon under study looks like. However, the role of the literature and theory in defining and bounding a case is somewhat of a paradox. The article is divided into three parts: first, I consider how to define a case; second, I discuss how to bound the case; and third, I dig deeper into the dilemmas of using the literature and theory in these processes. I conclude with final reflections on what I have learned in trying to define and bound a case in the study of power and interests in Norwegian Svalbard politics.
Defining a Case
Defining a case implies identifying the entity we need to study in order to say something about a more general phenomenon that our research question concerns (George & Bennett, 2005, p. 77; Rule & John, 2015, p. 4; VanWynsberghe & Khan, 2007, p. 85; Vaughan, 1992, p. 176; Walton, 1992, pp. 121–122; Yin, 2018, p. 31). This first step in case study research is what Ragin (1992a, p. 221) called the “first casing,” where we direct the study toward a specific category of social phenomena. Defining a case involves decisions on the entities to be studied in order to answer the research question. However, this process may not be as straightforward as one may think.
In classic case studies, the case entities are usually individuals, organizations, or communities (Becker, 1970, p. 75; Yin, 2018, p. 28). The research question “what interests prevail in Norwegian Svalbard politics and why,” however, pointed at a political entity as the case. The problem then centered on identifying which political entity.
Two alternatives I debated were policies or politics. To decide which was more appropriate for my study, I revisited the objective of my study and compared it with the definitions of relevant terms. In the Cambridge dictionary (s.a.) “policy” is defined as “a plan of action or a set of rules agreed by a business, a political group or a government, saying what they will do in a particular situation,” whereas “politics” refers to “the activities of the government or people who try to influence the way a country is governed.” What I referred to in my research proposal to show the existence of conflict of interests in Norwegian Svalbard policies/politics was the actual, written policy objectives, passed repeatedly in the Svalbard white papers. If I considered this the entity to study, my case could be defined as “policies.” On the other hand, it was not so much these objectives in themselves that I was interested in investigating, but rather the processes whereby some interests are prioritized at the expense of others. I figured that “politics” was at least a more appropriate term than “policies” in describing what I was actually going to study. Yet, this term did not seem precise enough either, as I wanted to be able to detect “the activities of the government or people” and more structural aspects of power. This led me to the decision to define my cases as political processes.
Although studying political processes has a long tradition in political science and other social sciences, one caution vis-à-vis such cases is that it can be difficult to identify their starting and ending points (Yin, 2018, p. 29). When the entity under study is a process, it is not necessarily easy to distinguish the context from the case. Thus, more clarifications are necessary for the case study to begin. This is where bounding comes in.
Bounding a Case
As Yin (2018, p. 31) wrote: “once you have defined your case, other clarifications – sometimes called
In my study, I realized that there are several ways in which political processes in Norwegian Svalbard politics could be bounded. First, I considered the spatial feature, or scope. At first, this seemed rather given: Norwegian Svalbard politics surely must deal with political decisions concerning Svalbard. However, Svalbard politics is a rather broad term, encompassing several possible conflicts of interests, happening at several political levels, and spanning over 100 years of history and a large geographical body.
To untangle this term, I revisited the objective of my study and the research question. To illuminate the question of what interests prevail and why, I would have to know something about which interests are at play. If, for instance, security, commercial, and environmental interests seemed relevant interests with the potential for conflict, they should be present in the political processes chosen as cases. Further, the puzzle of consistent policy objectives versus conflicts of interests mainly concerns the national political level, as the objectives are being expressed in the Svalbard white papers issued by the Norwegian Ministry of Justice and Public Security (2016). This meant that it made sense to focus on political processes at the national level to frame the case. Thus, the scope of the cases, “political processes in Norwegian Svalbard politics”, were bounded to encompass political processes at the national level where relevant interests are present. Of course, though, if actors at the local or international level tried to influence these processes, they would have to be given some attention as well.
Next, I considered how to bound the case in time. It is possible to do case studies both on historical and contemporary matters (George & Bennett, 2005, pp. 17–18; Yin, 2018, p. 12). I knew that political processes of Norwegian Svalbard politics in the past had been richly covered. I was also aware of several interesting – both ongoing and recently concluded - political processes that had not been investigated profoundly. Based on this, I chose to bound my cases to contemporary political processes. My idea was that historical events would still be given some attention in the study as I thought it likely that they would have some impact on the present political processes. Instead of being independent study entities, however, such historical events would be treated as the historical background and as part of the context.
In terms of geographical scope, I decided to include only the land areas and exclude the maritime ones. In Svalbard politics, the maritime area is subject to other laws, which is also the subject of an ongoing international debate. To include both Norwegian politics concerning the land and maritime areas and to ensure similarity between the cases would thus be challenging. The decision to geographically bound the case without reference to the maritime zones aligned well with the definition of Svalbard provided by The Svalbard Treaty (1920), where the maritime areas beyond the territorial sea are excluded. Thus, I figured, I could use the Svalbard Treaty’s definition of Svalbard to bound the case. However, this decision also came with its own challenges. As Harper (1992, p. 142) argued, there can be a mismatch between bureaucratically-derived and experienced boundaries. Grydehøj (2020, p. 267) claimed that the dominant scholarly discourse on Svalbard, rooted in the legal understanding, tends to overlook the meaning of Svalbard for those people who actually live in and visit it. In relation to my study, this could mean that the conflicts of interests I was going to investigate could possibly supersede the geographically boundary of the case derived from the Svalbard Treaty.
Through these decision-making processes I was made aware that like defining the case, bounding was not a straightforward affair. In several respects, it was rather difficult to make a clear-cut division between the case and context. I could not help but wonder whether there could be such a thing as bounding too much. Could it, in any way, harm the study if the limits for what I was studying were too clearly defined? This eerie question led me into a rabbit hole of dilemmas of using the literature and theory.
The Dilemmas of Using the Literature and Theory
To dig deeper into these dilemmas, let me first draw a distinction between the literature and theory, or rather facts and theory. With the literature, here, I mean a factual description of the topic of study, in this case, Svalbard. By theory, I refer to the ways in which to understand such facts. Whereas both are meaning-systems, we expect facts and theories to function referentially and conceptually, respectively, vis-à-vis the real world (Reed, 2011, p. 19). Although we need some level of interpretation to understand that Svalbard is an Arctic archipelago, this fact has a concrete, real-world expression, that is, the group of rocks between 10° and 35° longitude east and 74° and 81° north. The same is true for, let’s say, the fact that several states contest the Norwegian position that the Svalbard Treaty does not apply in the waters of Svalbard beyond the territorial sea. Theory provides “a new way to view” such facts (Reed, 2011, p. 21). In international relations, an often employed theoretical perspective to explain Svalbard issues is realism or geopolitics, where international politics is understood as rivalry among states for geographical space and resources (Kaltenborn et al., 2020, p. 26; Pedersen, 2020, p. 279). In relation to the above-mentioned example, a realist theoretical explanation can be to claim that different states’ positions on the geographical scope of the Svalbard Treaty is because of the competition over resources and influence in these strategically important waters.
In defining and bounding the case, both the literature and theory can be useful. There are however both pros and cons with this. Let us discuss the literature first. Yin (2018, p. 27) argued that a researcher should start with the literature in order to narrow down the field of interest and identify questions for further research. This is what one wants while initiating the research process through a literature review. According to George and Bennett (2005, p. 71), “situating” one’s research this way is “key to identifying the contributions the new research makes.” This “soaking and poking,” as the argument goes, will lead the researcher to invaluable understandings of the basic outlines of the case. For me, it could, for instance, be helpful investigating what constitutes key interests in Norwegian Svalbard politics, according to the literature. Such a review could direct me toward choosing relevant cases. If, for example, the literature highlights security interests as especially important, it would guide me to choose cases that encompass security interests.
There is a paradox of using the literature as a starting point; while it tells us where to look, it can also keep us from seeing, thus biasing the research process and suppressing potentially surprising findings (Garvey & Jones, 2021, p. 1; Rule & John, 2015, p. 4; Vaughan, 1992, p. 195). While using the literature to narrow down the research interest and questions, important and interesting aspects that are not yet covered can escape one’s attention. This can be considered a rationale for choosing cases and doing fieldwork before reading up on the literature. This way, the process of defining and bounding cases will be conducted without the constraints of the literature, which may be believed to promote innovation and more relevant research. As Becker (1970, pp. 79–80) pointed out, “the researcher may not be sure what problem is most deserving of study.” Following the previous used example, security interests are not necessarily the most important feature in Norwegian Svalbard politics just because it is the most prominent theme in the literature. Thus, a heavy emphasis on the literature can lead me away from more interesting aspects.
Even though there is an ongoing discussion on the role of the literature, researchers can, according to Reed (2011, p. 18), achieve a strong consensus at the level of bare social facts. It is not very controversial to say that in order to investigate the interests that prevail in Norwegian Svalbard politics, it is necessary to have basic knowledge of how to understand interests and what Norwegian Svalbard politics is. The conflict sharpens when it comes to the discussion on the role of theory.
There are certainly those who promote the advantages of using theory as a starting point. Important advantages include that it helps direct the attention in data collection and analysis, and that it avoids sidestepping problematic relations between ideas and evidence (Garvey & Jones, 2021, p. 1; George & Bennett, 2005, p. 89; Ragin, 1992a, p. 219; Rule & John, 2015, p. 4; Yin, 2018, p. 27). In relation to my research question, some theoretically informed propositions on what power is and how it works in relation to interests could thus be helpful. Behavioralism is a well-known theoretical perspective on power in political science. As Lukes (2005, p. 5) explained, this perspective understands power as “relative to several, separate, single issues and bound to the local context of its exercise”. Applying a behavioralist perspective on power would direct the focus onto the decision-making process in the political arena and with political actors and lobbying groups serving as relevant actors (Lukes, 2005, p. 5). This would provide clear boundaries for the levels, actors, and timespans that should be included in the case.
However, the argument that the literature keeps us from seeing also applies to using theory as a starting point. A related argument is that theoretically informed propositions are not able to embrace the complexity and diversity of the empirical world (Becker, 1970, p. 76; Ragin, 1992a, pp. 219–220). In this view, theory can be limiting and should therefore not be imposed until the later stages of research. As Ragin (1992b, p. 6) noted, his colleague Becker strongly advocated that it is counterproductive to begin research with a clear understanding of the case and its contents. Instead, the final realization of the nature of the case may be the most important part of the research process, and is not arrived at until the project is complete. An extreme version of this critique is found within the new materialism camp. Here, it is proposed that one should move from any prior construction of the case to letting the cases with their potential definitions and boundaries emerge as the study unfolds (Andersson et al., 2020, pp. 2–8). Following the earlier example of applying a behavioralist theory in bounding the case, I would overlook several important aspects. For instance, this theory does not involve the importance of the agenda-setting decisions that come up before the actual decision-making process, the actors excluded in the formal political process, and the mechanisms that influence, shape, or determine interests in the first place. Thus, applying theory could prevent one from discovering important aspects of which interests prevail and why.
Against this, Yin (2018, p. 34) argued that moving into the field without any predilections at all is simply impossible: “you cannot start as a true tabula rasa”. No one can free themselves from existing knowledge. Having once read behaviorist theory on what power is and how it works in relation to interests, for instance, I cannot block these ideas when I conduct research on what interests prevail in Norwegian Svalbard politics and why. What I can do, however, is to be explicit about my theoretical orientation and make use of it. As Vaughan (1922, p. 196) argued, explicitly using theory as a starting point can help avoid bias, as it makes both the researcher and readers more attentive to it.
Several strategies can be employed to counteract bias, such as explaining the case as fully as possible, formulating tentative hypotheses, collegial exchange, using insiders and outsiders, writing memos, and drawing comparisons with existing documented cases (Becker, 1970, pp. 77–78; Garvey & Jones, 2021, p. 5; Vaughan, 1992, p. 197). It is also important to address biases related to privilege and equity. Some practical advice on reflecting on a researcher’s identity in relation to that of their research participants include consultations with the participants, formalized collaborations or partnerships with the communities under study, discussing ethical standards from the participants’ perspective, clear communication, active listening, making the research process transparent, writing reflective notes, using a social identity map, and integrating time for researcher reflection and debriefing (CohenMiller et al., 2022, pp. 5–8; Jacobson & Mustafa, 2019; McDougall & Henderson-Brooks, 2021, pp. 3–8; Olmos-Vega et al., 2022, pp. 5–6; Schneider, 2018, p. 153; Thambinathan & Kinsella, 2021, pp. 4–5).
Not having theory as a starting point makes it difficult, if not impossible, to focus one’s attention and make sense of the empirical world (Ragin, 1992a; Yin, 2018, pp. 217–218). When one has the opportunity to consider anything interesting, it is difficult to know where to draw the line. There is however also a dilemma in choosing between guidance and flexibility. While having a theoretical starting point offers guidance, it can also hinder flexibility. This disadvantage comes into play if one clings to the view that the scientific value of the research endeavor would be dramatically altered if one allows for major changes in this design throughout the process (George & Bennett, 2005, pp. 22–24; Yazan, 2016, p. 140). This would certainly be a problem if one shares the perspective of, say, Becker (1970, pp. 79–80) or Stake (Yazan, 2016, pp. 140–141. If one agrees that it is impossible to know what problem is most deserving of study at the outset of a research project, one may rather be inclined to abandon any theoretical assumptions from the beginning. One may further favor the flexibility to implement major changes to the research design at any point in the research process.
As George and Bennett (2005, p. 130) noted, however, “observation is theory-laden, but it is not theory-determined. Evidence can surprise us and force us to revise our theories and explanations.” Having theory as a starting point does not mean that we should stick with it if the evidence tells us not to. Theory can inform the construction of a case in a loose fashion, but if the data contradict the theory, their use should be revisited. As Rule and John (2015, p. 8) pointed out, “theory informs case study but is in turn informed by the case or cases in a way that can further develop the theory”. This way, we can stumble upon “‘the loose ends’, the stuff we neither expect nor can explain, that pushes us toward theoretical breakthroughs” (Vaughan, 1992, pp. 175–176). The definition and boundary of a case can be treated as a “sensitizing concept” in that it “gives the user a general sense of reference and guidance in approaching empirical instances,” with the possibility of being tested, improved, and refined throughout the study (Blumer, 1954, pp. 7–8). Research then becomes a recursive process. Although the case is preliminarily defined and bounded, the definitions and boundaries may change for the better as the study unfolds. It is thus possible to use theory as a starting point in defining and bounding the case.
However, the choice of theory should be well-informed. Further, as Collins and Stockton (2018, p. 9) noted, “choosing to center a particular theory comes with the threat of becoming myopic”. The risk of overreliance on theory is evident, and needs to be approached with a sense of openness vis-à-vis conflicting explanations as the study unfolds. For instance, there are numerous concepts of power that can be helpful in unraveling the interests that prevail in Norwegian Svalbard politics and why they do so, instead of merely sticking to the behavioralists’ view. It is necessary to consider several and conflicting theoretical explanations of the phenomena under study in order to avoid myopic views, not at least when it comes to a highly theorized topic like power.
As I read up on theory and history, I saw how the boundaries of my cases had changed. First, having a master’s degree in political science and international relations made it difficult for me to simply leave this theoretical knowledge behind. Having this academic background, however, meant that my knowledge was centered on realist and behavioralist assumptions on what power is. Following this, my case boundaries would limit relevant interests to those related to geopolitics and relevant actors to those within the formal political arena. Second, based on my knowledge of Svalbard history and law, to temporarily bound the case to contemporary processes and geographically to land area and territorial sea made perfect sense. However, as I dug into other theories on power I realized that these strict boundaries were all in the danger of being too narrow. Was it a good idea, then, to just ignore environmental interest, for instance, or actors that were working to influence the agenda outside the formal political arena? Convinced that this would leave me with a too naïve an idea of what power is in Norwegian Svalbard politics, I decided to widen my case boundaries.
Nevertheless, as I entered the field, I could not help but wonder whether my case boundaries were still too narrow. Where exactly was I to distinguish the case from the context while trying to define where my processes started and ended? My informants wanted to start the history of the process I was investigating by going longer back in time than I had expected. I also found that a group of important political actors were cruise expedition tourism operators, which made me question whether I could really confine my geographical boundaries to the land and territorial seas alone. In the moment of writing, I still have not decided on where exactly I should place the temporal and geographical boundaries of my cases and whether they should be as fixed as I first imagined. I nevertheless think that having well-informed theory as a starting point guided me on what may have otherwise felt like an ocean of important findings Figure 1. Illustration of how the boundaries changed as I read up on theory and history. The second boundary is dashed as it is not yet settled.
Some Final Reflections
In this article I have discussed the dilemmas I encountered in defining and bounding the case in the study of what interests prevail in Norwegian Svalbard politics and the reasons for their prevalence. I soon discovered that this was not as straightforward as I first assumed, as there were several possibilities in deciding on the entity I was going to study and how it could be bounded. I did, however, find that revisiting the research question and study object while carefully examining them in relation to the definitions of possibly relevant terms was helpful as it led me to a preliminary definition of my case as a political process. This also helped me bound the case to some extent, but these decisions were also based on pragmatic considerations and my own research interests.
The process of constructing the case made me question whether there was such a thing as bounding too much. I dug deeper into the dilemmas of using the literature and theory in defining and bounding a case. Despite the paradox of using the literature and theory as a starting point, I find it possible to find a middle ground between flexibility and guidance. In the previous discussion, I made a clear distinction between the literature and theory based on Reed’s differentiation between facts and theory. In this sense, the literature and theory do fulfill different purposes: whereas the literature tells us how reality looks, theory tells us what to look for. This also points toward the fruits of using both the literature and theory in combination. To complicate the dilemmas of using the literature and theory further, however, the terms “theory” and “literature” are often used in different ways. This can make it difficult to know when to rely on a claim as basic knowledge or as a lens with which to view the world. To accommodate this challenge, I have come to distinguish between “how it looks” and “what to look for”-claims rather than to trust blindly that something that poses to be a fact is indeed one. This is further underscored when we take into account that researchers, the very authors of the literature and theory, are themselves a part of the power game around defining and bounding not only specific cases and concepts, but the social world at large.
Finally, social knowledge claims are often a mix of theory and facts (Reed, 2011, p. 23). Power, I think, is a very good example of this, being an “essentially contested concept” (Lukes, 2005, p. 11). As the discussion above makes it clear, this does indeed complicate the process of defining and bounding a case. My attempt to discuss how to define and bound a case in the study of interests and power in Norwegian Svalbard politics made me realize that there are many dilemmas involved in using the literature and theory in the process. While it tells us where to look, it can also keep us from seeing. While it provides guidance, it can hinder flexibility. Although there may not be a clear solutions to these dilemmas, I have discovered that treating preliminary definitions and boundaries as sensitizing concepts can allow me to find “the stuff” that pushes the study toward more interesting findings and theoretical innovations.
In this article, I have argued that it is possible to find a middle ground between flexibility and guidance when it comes to assessing the role of theory in defining and bounding a case. The question is, of course, how we can find this middle ground. The methodology textbooks do not provide the researcher any clear-cut answers. There will always be some voices arguing for the use of theory and some against it. Like most textbooks, I do not offer any final conclusions on what the role of the literature and theory should be in defining and bounding a case. I do however hope that the insights conveyed in this paper will encourage others to reflect further. This is where the answer to solving the dilemmas of using theory and the literature in designing a case study lies: not in methodology textbooks or journal articles, but in a researcher’s reflection, vis-à-vis their research.
Footnotes
Acknowledgments
) for English language editing is also appropriate.Declaration of Conflicting Interests
Funding
ORCID iD