TY - JOUR
AU - Mahé,, Anne-Laure
AB - Abstract This article examines writing as the last link in the epistemology-theory-methodology alignment. Although political scientists dedicate a great deal of their time to writing, conversations on this topic remain scarce within international relations and political science overall. Notably absent are analyses of the actual writing choices scholars make and what these mean for the knowledge they produce. This article uses the tools of literary analysis to take a closer look at the mechanics of three published academic articles in the fields of international relations and comparative politics. It focuses on how qualitative interviews are written, demonstrates how authors can conceal or reveal the dialogical dimension, and examines how they deal with the conundrum of the representation of research participants. This kind of reflexive analysis reveals the epistemological foundations of a given research article and can be used to identify instances of misrepresentation and misalignment. As such, it is an important tool for the improvement of academic writing. Resumen En este artículo, se examina la redacción como el último eslabón en la alineación epistemología-teoría-metodología. Aunque los científicos políticos dedican gran parte de su tiempo a escribir, las conversaciones sobre este tema siguen siendo escasas en el marco de las relaciones internacionales y la ciencia política en general. Cabe destacar la notable ausencia de análisis de las elecciones de redacción reales que realizan los académicos y lo que estas significan para el conocimiento que producen. En este artículo, se utilizan las herramientas del análisis literario para observar más de cerca la mecánica de tres artículos académicos publicados en los campos de las relaciones internacionales y la política comparativa. Se hace hincapié en cómo se redactan las entrevistas cualitativas, se demuestra cómo los autores pueden ocultar o revelar la dimensión dialógica, y se examina cómo abordan el enigma de la representación de los participantes de la investigación. Este tipo de análisis reflexivo revela los fundamentos epistemológicos de un artículo de investigación determinado y puede utilizarse para identificar casos de tergiversación y desalineación. En tal sentido, es una herramienta importante para la mejora de la redacción académica. Extrait Cet article examine l'écriture comme dernier maillon de l'alignement épistémologie-théorie-méthodologie. Bien que les politologues consacrent une grande partie de leur temps à l'écriture, les discussions sur ce sujet restent rares au sein des relations internationales et de la science politique en général. Les analyses sur les choix réels d'écriture que font les chercheurs et leur signification pour les connaissances qu'ils produisent sont particulièrement absentes. Cet article utilise les outils de l'analyse littéraire pour examiner de plus près les mécanismes de trois articles universitaires publiés dans les domaines des relations internationales et de la politique comparée. Il se concentre sur la manière dont les entretiens qualitatifs sont rédigés, montre comment les auteurs peuvent dissimuler ou révéler la dimension dialogique et examine comment ils traitent la problématique de la représentation des participants à la recherche. Ce type d'analyse réflexive révèle les fondements épistémologiques d'un article de recherche donné et peut être utilisé pour identifier les cas de fausse représentation et de désalignement. En tant que tel, c'est un outil important pour l'amélioration des travaux académiques. writing, methodology, epistemology, literary analysis, reflexivity When asked “how is your writing,” political scientists tend to answer by talking about the places they find (un)conducive to writing, the tools they use, and whether they are getting anything done at the moment. They are unlikely to discuss their use of adverbs, turns of phrase, or preferred typography. Indeed, while writing is an activity they engage in on a daily basis, reflections on its technical aspects remain marginal in the discipline's literature. In general, conversations about writing mostly happen online, where specialized websites regularly publish columns on the topic, and they center on a few major ideas: that writing is difficult, that academics write terribly, and that we misguidedly think of it as the last step in the research process.1 General method textbooks in political science, both quantitative and qualitative, are surprisingly silent on the matter.2 When they are not, they often propose guidelines for structuring a manuscript (see for instance Dunleavy 2003), which amounts to stating—and enforcing—disciplinary conventions. Volumes targeting specific methods such as fieldwork and interviewing sometimes include chapters on writing (Schatz 2009b; Mosley 2013a; Kapiszewski, MacLean, and Read 2015), but they are often more concerned with which information should be included in the final report, rather than how it should be incorporated in the research (Bleich and Pekkanen 2013).3 Writing generally is not included in debates about epistemology or methodology. It is indeed absent from recent conversations on those topics in international relations, which have been renewed by the development of interpretive and discursive approaches and a growing interest for ethnography and critical approaches (Aradau and Huysmans 2014; Jackson 2016). Yet, writing raises issues that are quite similar to those raised by the concept of methods, which Aradau and Huysmans (2014) argue have been misleadingly conceptualized as technical devices that are disconnected from the broader question of the philosophy of science. Given the focus of most textbooks on establishing conventions for good academic writing, the same criticism applies to conceptualizations of writing. Just as methods, writing is composed of “material devices that enact worlds” (Aradau and Huysmans 2014, 604). Current epistemological and methodological debates in international relations, therefore, do not consider the fact that writing is the very process through which we treat and provide evidence and that, consequently, it could be placed at the center of reflections on the logic of scientific inquiry. Writing is, I argue, the missing link connecting epistemology, theory, and methodology. This means that analyzing our writing choices is a way to identify and investigate our epistemological underpinnings. It especially enables us to highlight whether those dimensions are aligned in a given work and what are the impacts of misalignments and mismatches on the knowledge we produce. For authors such as Jackson (2016) and Hall (2003), misalignments are problematic because it means that our methodologies are not capturing what we think they are, which makes it impossible to properly assess the scientific claims presented in a given research project. The idea of this alignment is of course subject to debate, with major authors arguing that epistemological divides between methodologies should not be overstated and that conflating methodological differences with epistemological ones can be misleading (Becker 1996; Keating and Della Porta 2010). In general, researchers have nuanced and pragmatic approaches. They use the tools most adequate to investigating their questions of interest: it might well be “bricolage,” rather than systematicity and logicality, that characterizes the conduct of scientific inquiry (Aradau and Huysmans 2014, 606). The tendency in political science to let writing take a back seat to other methodological issues is therefore highly problematic because it deprives us of an opportunity to reflect on how the decisions we make at the level of sentence structure, word choice, and even punctuation affect the knowledge we produce. Here, I focus on how interview materials are written in academic articles to demonstrate how our “ways of writing” are intrinsically connected to our “ways of knowing” (Moses and Knutsen 2012). Using the tools of literary analysis to go in-depth into the mechanics of texts produced by political scientists working in the fields of international relations and comparative politics, I empirically investigate whether researchers who use similar tools to collect data make similar writing choices. This approach is inspired by debates on the craft of writing that have taken place in other disciplines such as sociology and anthropology, though it also differs in various ways. Sociologists have reflected on what sociological writing is and how it differs from literature since Bourdieu (Passeron 1991; Bourdieu 2015), paying particular attention to the social conditions of the production and reception of the “sociological discourse.” They have produced major contributions on how to write “well” (Becker 2008; Wolcott 2009),4 reflections on the writing process (Eliasoph 2011), investigations into the ethical questions that entail writing choices (Beliard and Eideliman 2008), and innovative work on different modes of writing (Nocerino 2016). However, few authors have provided in-depth investigations into the nuts and bolts of particular texts beyond providing illustrative examples. This is something anthropologists have been much more involved in, especially in the 1980s when the “crisis of representation” shook the discipline. Critical authors then mercilessly deconstructed the conventions of anthropological writing, demonstrating how it had been plagued by all of the “original sins” of the discipline: exoticism, colonialism, neglect, and instrumentalization and objectification of the natives’ voices and meanings (Clifford and Marcus 1986). Authors such as Van Maanen (2011) and Crapanzano (1986) have shown how authority and objectivity is built through formal and stylistic elements, highlighting that even if “natives” are given a voice, this is often merely an artificial construct. This article follows their example and moves as close as possible to the text itself, in this case three scientific articles written by political scientists and published in peer-reviewed political journals. The analysis investigates how qualitative interview materials are written in those texts. This narrow focus is motivated first by the notion that qualitative researchers, especially those practicing ethnography and qualitative interviews, are more concerned with writing than their quantitative peers because their process of collecting data places them in lengthy face-to-face interactions much more frequently. Consequently, conundrums related to the representation of the “others” and their discourses are bound to appear with more prevalence. Additionally, qualitative evidence is textual in nature; it is not submitted to the quantification process specific to quantitative research,5 and transcribing interviews and scribbling field notes already entails making writing choices (Emerson, Fretz, and Shaw 2011). Second, while there is a long-standing trend in international relations since the 1980s of increasing use of quantitative methods (Nahmias-Wolinsky and Sprinz 2004, 6), qualitative approaches and interviews have gained popularity with the so-called “practice turn” that emphasizes “the need for scholars to conduct empirical research within the sites where practices are performed” (Cornut 2015). Beyond international relations, qualitative interviewing is a data-collection method that “occup[ies] a dominant place in qualitative social science research” (Whitehead and Baldry 2017, 135), making this article relevant for a broader community of scholars. After briefly defining literary analysis in the next section, I present the corpus of texts I analyze in further detail in a second section. The third and fourth sections provide the analysis itself, which first looks at the treatment of the dialogical dimension of interviews and then at the characters involved. I then conclude with some remarks on the usefulness of this type of analysis and the impact of the structure of academia on our ways of writing. What Is Literary Analysis and What Is It Good For? Literary analysis is the process of carefully examining a piece of literature in order to better understand its inner workings and to identify its underlying meaning(s). It is about identifying issues and proposing hypotheses about why a text is written one way and not another (Rapaport 2011, 2). As such, it is also an act of interpretation, though the use of the singular here is misleading: there are multiple interpretations of the same text depending on social groups, readership, and historical period (Montgomery et al. 2007, 10). In this type of analysis, we look at the various components of the text—elements of language and form—to ascertain how they relate to each other and to a broader meaning or narrative. Literary analysis consequently proceeds from comprehension to interpretation (Rapaport 2011, 2), with the former being the step in which the critic gathers evidence to build her interpretation. Comprehension of a literary work therefore “include[s] the ability to identify points of view, major themes, and key allusions (references to historical occurrences, myths, or passages in other influential texts, for example, the Bible)” (Rapaport 2011, 2). There are different approaches to literary analysis. The one I have just described is akin to close reading, which is about paying attention to “even the most minor textual details in order to develop explanations for the question of why things are presented the way they are” (Rapaport 2011, 4). Other approaches might move away from the text itself. For instance, contextual analysis “involves the establishment of a context (or con-texts) within which to situate and determine the meanings of a work by drawing direct connections between elements within the work and elements within the context” (Rapaport 2011, 5). Critical approaches are based on systematically applying a theory to interpret a literary work, which is what Freud was doing when he used his theory of the Oedipus Complex to interpret Hamlet (Rapaport 2011, 7–9). Lastly, social criticism approaches the text from a sociological perspective, interpreting it in relation to the society of either the author or the readers. Another strategy is to analyze the representations of a given society within the text, as Marxist and feminist criticisms do. Of course, looking at the formal elements of the text—the comprehension stage—plays a major role in each of those approaches, so it is important to specify at what exactly we should be looking.6 There are countless literary terms filling many extensive glossaries and little consensus on how to categorize them. Generally speaking, there are two categories of literary devices: literary elements, which are necessary features that can be genre specific, and literary techniques, which are tools used to convey a specific meaning and usually give literary qualities to a text. Classic literary elements, also called narrative elements (Rapaport 2011, 66), include plot, characters, settings, point of view, and action. Each of these can be divided into subcategories: for instance, characters can be protagonists or antagonists, and conflict can be external or internal. Literary techniques refer, for instance, to the use of figurative language (metaphors, allusions, etc.), to various ways of evocating sound such as alliteration, dissonance or puns, and to the use of symbolism and motifs. This is where elements of style, such as vocabulary, syntax, and typography, come into play (Montgomery et al. 2007, 10). Analyzing style is about identifying how the writer accomplishes her aims. It might seem that these analytical approaches are foreign to the social sciences and that literary criticism can only apply to “literary” texts. After all, does it make sense to talk about characters in an academic work? I argue that it does, and indeed literary theory and the social sciences have long been in dialogue. Marxism, structuralism, and postmodernism are, for instance, common to both fields, and international relations scholars have borrowed theories and methods from literature to develop discourse analysis methodologies. Anthropology has also shown how literary analysis can be used to great effect, as exemplified by Crapanzano's (1986) criticism of Geertz’s famous account of the Balinese cockfight. He demonstrates that writing choices convey the idea that the author understands the “deep” meanings of a cultural practice better than the Balinese themselves, and the writing choices reproduce his authority, for instance when through puns, titles, subtitles, and simple declarations, the “anthropologist” and his “Balinese” are separated from one another. In the opening section of “Deep Play,” Geertz and his wife are cast, however conventionally, as individuals. The Balinese are not. They are generalized. (Crapanzano 1986, 70) Academics use literary devices as much as other kinds of writers, creative or nonfictional, even though some authors argue it should be avoided so as to produce a text where the researcher writes herself out and “let the findings speak for themselves” (Locke and Golden-Biddle 1997, 4). Overall, the idea that academic and literary writing are two distinct categories is quite pervasive and even a topic of reflexive investigation (Becker 2008; Bourdieu 2015). Yet, I insist that it should not be overemphasized, because this might hinder our ability to identify when such devices are used to conceal relationships of power and processes of textual manipulation. Like other types of writers, academics use rhetoric to persuade and convince their reader (Bryman 2012, 684): there is no such thing as a form of objective and purely utilitarian writing that would reflect objective knowledge. Literary analysis, because it was not specifically designed for the analysis of political science, helps produce an estrangement from the texts that enables us to deconstruct this point of view and highlight the conundrums related to the task of representing “others.” Analyzing our writing choices is the first step in the process of becoming aware of the effects of our own writing, enabling us to make better choices as necessary. The Corpus This article attempts to fill the gap in the epistemology-theory-methodology alignment by conducting a close-reading of three academic articles. This approach is fundamentally empirical, as it starts with the text and moves on to interpretation. As my aim is not to make generalizable claims about how political science is written, and as literary analysis is not concerned with such goals, the texts forming the corpus do not purport to be representative of common practices in the discipline; such a project would need more extensive qualitative and quantitative investigation. The corpus was therefore constituted according to the logic of convenience sampling, choosing texts by virtue of their accessibility (Bryman 2012, 201). The articles were published by colleagues I personally know, which enabled me firstly to make sure that authors agreed to have their writing closely examined in a critical perspective, something not everyone might be comfortable with, and secondly to start a conversation with them on this topic. This is important because, in fine, the main goal of this article is to foster dialogue and reflexivity. Therefore, I sent the authors a first version of the article and they gave me feedback and precisions on their writing choices in informal online conversations. Some of their reflections are included in the analysis. Of course, the fact that all three authors belonged to Canadian institutions at the time of their articles’ publication and are junior scholars might impact on their writing. It means that they have been exposed to specific ways of practicing social sciences and as a consequence to specific ways of writing. Furthermore, it is likely that less established scholars are more likely to follow conventions and to demonstrate less freedom in their writing. Nonetheless, while it is the product of convenience sampling, the constitution of the corpus was nonetheless guided by specific criteria. I chose articles for practical reasons: textual analysis is a lengthy process that results in a lengthy analysis because the corpus needs to be quoted and looked at in details. It was therefore more manageable to choose a limited number of shorter texts as opposed to books. This also reflects evolutions in IR and more broadly social sciences: the publication of peer-reviewed articles has become an increasingly important means to diffuse knowledge and it is likely that now most scholarly writing takes this shape. Furthermore, I realized while conducting the analysis that this limited corpus provided me with enough evidence to sustain my argument, so in a way I reached a point of data saturation. Second, the articles needed to be published in peer-reviewed journals because that means that they are in line with current writing conventions in political science. Third, the three articles belong to the field of international relations, broadly defined—a field in which interviews have specificities: they are often conducted with elites in institutional settings, addressing topics that involve a degree of secrecy, and imply the use of foreign languages (Alles, Guilbaud, and Lagrange 2018).7 Lastly and importantly, they needed to follow different theoretical and methodological approach while still using qualitative interviews as a data collection method since I develop an argument about the differences in writing across epistemological positions. Following Moses and Knutsen (2012), I consider epistemological traditions as a continuum bookended by two ideal-types: positivism and interpretivism.8 Positivism is based on a foundationalist ontology purporting that “the world exists independently of our knowledge of it” (Marsh and Furlong 2002, 22) and can be investigated by scientific methods. As Marsh and Furlong (2002, 22) state, “we can establish regular relationships between social phenomena, using theory to generate hypotheses [that] can be tested by direct observation.” The observer can be objective and the knowledge produced is not mediated by her interpretations. Positivist researchers aim to identify patterns and to make causal statements and consequently see the scientific project as aimed “at the general (nomothetic) at the expense of the particular (idiographic)” (Moses and Knutsen 2012, 9). In this perspective, experimental and statistical methods are to be prioritized, followed by comparisons and case studies. At the opposite end of the spectrum, the interpretivist tradition “rejects the notion that the world exists independently of our knowledge if it. Rather, they contend that the world is socially or discursively constructed” (Marsh and Furlong 2002, 26). No social phenomenon exists independently of our interpretation of it, and therefore researchers focus on the production of interpretations and meanings. This also means that there is no such thing as objective knowledge. Methodologically, such an approach gives precedence to qualitative methods in order to “help us establish how people understand their world” (Marsh and Furlong 2002, 27). Generalization is not a concern; emphasis is placed on the specificity of contexts, though limited and analytical generalization is possible. Bearing in mind that the epistemological positions I outlined are ideal-types, the first article, entitled “A Question of Trust: Military Defection during Regime Crises in Benin and Togo,” published by Morency-Laflamme (2018) in Democratization, seems located closest to the positivist end of the epistemological continuum as the author makes causal claims and seems to share the positivist emphasis on falsifiability and hypothesis-testing, writing that he draws “on a most similar case comparison to assess the validity of the hypotheses stated above to control for alternative hypotheses” (469). Morency-Laflamme examines the role of trust between military officers and opposition forces in fostering mass defections of military personnel during prodemocracy uprisings. The article demonstrates that loyalist stacking creates a core of military personnel with a strong stake in regime preservation, while counterbalancing allows for the possibility of a military-opposition alliance. The findings are derived from a most similar cases comparison of civil-military relations during mass uprisings in Benin (1989–1990) and Togo (1990–1993). Most of the demonstration is based on secondary sources, but interviews are used twice in the text.9 The author explains that he selected the cases purposefully to test his hypotheses, which reflects a deductive approach to conducting research. The second article, from Martel (2017), sits at the opposite end of the continuum, firmly located within an interpretivist tradition. “From Ambiguity to Contestation: Discourse(s) of Non-traditional Security in the ASEAN Community,” published in the Pacific Review, analyzes the uses of the concept of “non-traditional security” within the Association of Southeast Asian Nations (ASEAN). While it is often seen as an uncontroversial term, referring to supposedly self-evident issues, the article demonstrates that there are divergent, and potentially contradictory, interpretations of nontraditional security meaning and implication in ASEAN and the wider region. The author uses an interpretative approach “that focuses on the role of discourse in the social construction of reality” (Martel 2017, 552) and proceeds to a discursive analysis of a corpus that includes “official ASEAN documents, policy-oriented publications from think-tanks, press articles, and communiqués produced by regional civil society networks engaging in the ASEAN process” (Martel 2017, 552). Transcripts from interviews conducted in the field by the author complete this corpus. The third article, “A Tradition Coopted: Participatory Development and Authoritarian Rule in Sudan,” is one of my own (Mahé 2018), giving me the opportunity to exert reflexivity on my own writing. Published in 2018 in the Canadian Journal of Political Science, it represents a middle ground between the other two. I analyze how participatory development is used in the authoritarian context of Sudan to renegotiate power relations at the local level. The method is a case study of development projects implemented in a Sudanese province, with data collected during fieldwork through observations, interviews, and the collection of official documentation. The article demonstrates that there is a disconnect between the discourse surrounding participatory devices, which establishes a horizontal relationship between citizens and the local government, and the actual practices that strengthen the latter's power. My emphasis on a difference between practices and discourses echoes a third, more recent epistemological tradition called “realism” (Marsh and Furlong 2002) or “scientific realism” (Moses and Knutsen 2012), which acknowledges that, while “the world exists independently of our knowledge of it” (Moses and Knutsen 2012, 30), “there are deep structures that cannot be observed and what can be observed may offer a false picture of those phenomena/structures” (Marsh and Furlong 2002, 30). Consequently, “there is often a dichotomy between reality and appearance” (Marsh and Furlong 2002, 31), so what actors say should not be taken at face value. Methodologically, this posture, that acknowledges that the world exists and is socially constructed by reflexive agents, finds equal utility in quantitative and qualitative methods. I argue that analyzing how these authors write interview data will reveal different ways of tackling and using interview data, which are in turn related to different epistemological foundations. This is in line with Mosley's argument that positivist and interpretivist scholars do not use their interviews in the same ways. For scholars working from a positivist tradition, interviews are “a means of generating objective knowledge, either to generate or test falsifiable hypotheses” (Mosley 2013b, 10). Interviews are used to identify causal processes and test theories. Furthermore, “although positivist scholars are sensitive to the existence of ‘interviewer effects’—in that their individual characteristics, and how these are perceived by their interviewees, may influence the information that is provided—their focus is more on interview data as a product, often collected over a relatively short period of time” (Mosley 2013b, 10). On the other hand, scholars coming from an interpretivist position will use interviews for theory development and “highlight the need for attention not only to information itself, but also to how, and by whom, the information is generated and gathered” (Mosley 2013b, 10). This does not mean that they never use interviews to test falsifiable claims, but they consider interviews more as a process than a product, asking “how interviewees themselves make sense of the world, and why the interview data take the form that they do” (Mosley 2013b, 10). Data collected through similar tools are analyzed differently depending on the type of knowledge the researcher intends to produce and consequently written differently. Our ways of writing ineluctably show our epistemological positioning. In the following sections, I delve into the literary analysis itself, focusing first on what happens to the dialogical dimension of interviews in the selected articles and second on how the characters involved in the dialogue are represented. Interviews as Dialogue Interviewing research participants is a peculiar and often uncomfortable exercise. As an artificially crafted situation that involves at least two people, it is often a moment of working misunderstanding, as the interviewee's interpretation of what is going on and what the conventions of this situation are might significantly differ from the interviewer's expectations. These misunderstandings, as well as the context of the interview, partially determines its content, yet depictions of the interview situation rarely find their way into academic manuscripts. Thus, not only the basic setting, but even the dialogical nature of the interview, and therefore of the data collected, are sometimes hidden through syntactical, grammatical, and typographical choices. In my article, there are moments where I clearly acknowledge the interview situation: During my second interview with Kedar, a former high-level official and a volunteer for the Nafir's Committee in Khartoum, as I mentioned participation, he interrupted to ask, “Is it participation or partnership?” He then explained, “The government does not have enough money, in such a big country, for infrastructure, basic services. But institutions, companies have the money. If you consider them as genuine partners it will make them happy, they will accept the idea (interview with Kedar, Khartoum).” (Mahé 2018, 240) Here, the presence of the interviewer is obvious. The direct quote is contextualized by referring to the fact that it was a second interview and shows that the statement came in response to one of my remarks, where I offhandedly spoke of participation and inadvertently elicited the reaction from the interviewee. These writing choices are motivated by the relevance of the spontaneous nature of his reaction, and of the fact that he interrupted the question, to the argument I am making, where I am questioning the use of participation to fund development projects. Were this contextual information left out, the reader might interpret the quote as the answer to a direct question, such as “is this participation or partnership?” Yet, here I did not solicit a reflection on the distinction between participation and partnership; it came from the interviewee himself. This means it is important to him and he wanted to tell me about it. In this example, the reader is made to envision the encounter, but the dialogical situation can also be shown without explicit reference to the context of the exchange. See for instance this excerpt from Martel's (2017) article: According to a former ASEAN Secretary-General, [NTS refers to] what is topical at [the] time or what is at that moment . . . on the radar scope. So people say [NTS]: Oh, what does that mean? You ask everybody in the ASEAN membership you get ten different answers. There was no effort made to converge the understanding or definition or anything like that. But it is a good . . . line to say: OK, now we agree on [NTS]. What is [NTS]? ABC or everything as long as you fully agree on something, . . . we go with it.10 (Martel 2017, 553) In addition to the footnote that explains where the quote is from, the reader can also make an inference as the author mentions earlier in the methodological section of the article, that she conducted interviews with ASEAN diplomats, members of Track 2 networks such as the Council for Security Cooperation in the Asia-Pacific (CSCAP) and the ASEAN-Institutes for Strategic and International Studies (ASEAN-ISIS), and representatives of civil society organizations associated with the Solidarity for Asian People's Advocacy (SAPA). (Martel 2017, 552) The choice of a lengthy direct quote, signaled by the indentation of the paragraph, shows that there are voices other than the author's in this text. Yet, the interviewer's presence is not acknowledged in the main text in this excerpt. The use of the pronoun “you” could be interpreted as referring to her, however, it could just as well be the impersonal, generic “you” that colloquially replaces “one.” In contrast to these writing strategies, Morency-Laflamme's (2018) article, as shown in the following excerpt, chooses not to show the situation of the interview. This choice tends to conflate the interviewee's speech with the author's voice. Lacking links to military personnel, antiregime leaders miscalculated their standing and “thought that officers would massively defect and join their movement. This was a mistake as President Eyadéma had consolidated his control on the military even before the conference.”11 The absence of social ties between military personnel and the opposition allowed President Eyadéma and his closest associates to portray the antiregime force's demands as an attempt of southern elites to expel northerners from power.12 (Morency-Laflamme 2018, 471) The main text does not refer to the interviewer, even though a quote must necessarily be the answer to one of his questions and is a discourse that was produced for him to hear. The footnotes state only that the quotes are taken from interviews, without indicating whether the interviews were conducted by the author, though it seems more logical to state when this is not the case than the other way around. Martel (2017, 522), however, did reference the interview in her footnote and wrote “interview by the author.” Morency-Laflamme's style of narration also helps deemphasize the interview situation. Though he uses direct speech in the first sentence shown above, signaled by quotation marks, in the next sentence he uses indirect speech. For the reader, the footnote is the sole indication that the content of the sentence is taken from an interview. The switch to indirect speech means that the distinction between the author's and the interviewee's voices is more tenuous, whereas the quotation marks made it visible. In the second sentence, there is continuity between the two voices. The author is using a style of narration that resembles free indirect speech, a mix of direct and indirect speech that makes it challenging to clearly separate what the narrator says from the character's discourse (Montgomery et al. 2007, 278). The shift from the one to the other is usually signaled by changing pronouns or tenses, but also through punctuation, such as the use of exclamation points, which are usually reserved for direct speech. Morency-Laflamme does not use these grammatical or typographical signifiers. Furthermore, the continuity between voices is enhanced by a citation style that does not differentiate in-text between interviews and bibliographical references. In order to know the source of the information contained in the sentence, the reader must check the endnotes. These writing choices, which reveal or attenuate the dialogical nature of the interview, are not made in a vacuum: they are linked to the methodologies and theories the authors use and to their epistemological groundings. First, in my article, the data were collected “through interviews, observations, and official documentation during fieldwork in Khartoum and North Kordofan” (Mahé 2018, 238). While this information in itself does not tell us much about the epistemological bases of the analysis, the fact that I include reflexive statements about the limitations of the data13 is indicative of fieldwork conducted with an ethnographic perspective as defined by Schatz (2009a). Moving away from classic anthropological definitions that emphasize the length of the stay and the intensity of the immersion, Schatz defines political ethnography as “a sensibility that goes beyond face-to-face contact. It is an approach that cares—with the possible emotional engagement that implies—to glean the meanings that the people under study attribute to their social and political reality” (Schatz 2009a, 5). This definition accommodates the diversity of ethnographic practices in political science and acknowledges the recent evolutions of the method. This is an inherently constructivist approach, which fits with the Foucauldian perspective I am using to analyze the data (Mahé 2018, 238). Other signifiers of a constructivist approach are the acknowledgment that data are coproduced—“this presents a limitation of the data, as the discourse with the interviewees was, in some ways, coproduced with the translator” (Mahé 2018, 238)—and the insistence on the contested nature of the information provided by the interviews—“I take care in the paper to indicate every time there are contradictory accounts or interpretations of a piece of information”(Mahé 2018, 239). In addition, as demonstrated earlier, I clearly show the interview situation and the dialogical nature of the data, which corresponds to the writing conventions of ethnography. As a method, ethnography is “based on the close-up, on-the-ground observation of people and institutions in real time and space, in which the investigator embeds herself near (or within) the phenomenon so as to detect how and why agents on the scene act, think, and feel the way they do” (Wacquant 2003, 5). It therefore emphasizes emic discourses and points of view, the ones the participants hold, as distinct from the etic discourses of the researcher's interpretation. Writing up research based on ethnographic fieldwork means first showing the contests, contradictions, and uncertainties that characterize meaning-making processes and, second, separating as clearly as possible the emic from the etic, though in theory and practice it might be challenging to do so. This is a necessity derived from the criteria of scientific rigor of such qualitative research, which are distinct from those of quantitative and more positivist research (Olivier de Sardan 2008). Instead of validity and reliability, it is about credibility, (i.e., the degree to which the account represents the perspectives of the research's participants and is believable to them) (Lincoln and Guba 1985; Creswell and Miller 2000). Olivier de Sardan's principle of double adequacy is a useful tool for assessing this type of research; there should be correspondence between the empirical data and the “real world” and correspondence between the empirical data and the research's argument (Olivier de Sardan 2008). To enable the reader to assess this, fieldworkers need to show great transparency at every step of their argument and method. A second criteria of rigor, transparency is the foundation of the “ethnographic pact” between the reader and the researcher. This is the idea that what the author is talking about really happened, the statements she reports were truly spoken to her, and she is not depicting a fictional reality (Olivier de Sardan 2008, 28). Credibility and transparency are built in to specific writing conventions, among them, providing depictions and dialogues that elicit a reality effect and writing reflexively on the data. While such conventions have been criticized by some anthropologists since the “crisis of representation” of the 1980s (Clifford and Marcus 1986), they remain staples of ethnographic writing. This explains why the second interview with Kedar in Mahé (2018) was written as it was in the main text. By making myself visible in the exchange, I not only highlight something I believe is relevant to the interpretation and theoretical argument, it also reminds the reader that I was there and I heard this speech. Thus, I reactivate the ethnographic pact and provide credibility and transparency. And, this is not the only function of this narrative device; it is also a well-established mode of claiming authority, a writing device of persuasion that aims to make the message more convincing (Crapanzano 1986, 52). In addition, I emphasize in the methodology section of the article that, given the authoritarian context of Sudan, information provided by interviewees should be carefully assessed. In this perspective, it is crucial that the reader knows where the data comes from, so she can assess for herself their credibility and the credibility of the author's interpretation, but also that the authorial voice and the voices of the interviewees remain clearly distinct. Martel's writing reflects similar, though slightly different, concerns. Her approach also calls for transparency, since she focuses on analyzing discourses and intends to provide an in-depth analysis of texts produced by various actors and highlight their variations. This entails the necessity of clearly identifying who is talking and from where. Indeed, this is the core of her demonstration: she shows that interpretations of nontraditional security differ according to the actors observed and the “tracks”—official diplomacy, unofficial diplomacy, and alternative diplomacy (Martel 2017, 551)—of the ASEAN diplomatic process to which they belong. Martel therefore systematically states who is talking, writing for instance “in the words of a former director of the ARF unit” (2017, 558), or “for an ASEAN diplomat, the development of ASEAN cooperation” (2017, 560). The reader always knows exactly where quotes come from, and furthermore, the author makes abundant use of direct quotation, as evidenced by the abundance of quotation marks in the article, which are used even for single words. This signals that statements are the exact words of interviewees, not paraphrases. Yet, while transparency is key, this is not necessarily the same transparency as in my work. For instance, it does not extend to providing details on the setup of the interview or making the interviewer visible. This derives from her interpretive approach which focuses on “the role of discourse in the social construction of reality” (Martel 2017, 552).14 Nonetheless, as Clifford (1983, 139) explains, “quotations are always staged by the quoter,” so while direct quotes may appear to contribute to transparency, with their clear distinction between various voices, the writer and the reader must remain careful. Morency-Laflamme's (2018) writing choices reflect a different perspective from the other author. They demonstrate that he considers the interview to be a tool to collect factual data, echoing a more positivist perspective that “seeks to record facts to mirror an external reality” (Van Puyvelde 2018, 3) and therefore “tends to consider interviews as sources of witness accounts” (Van Puyvelde 2018, 3). Three characteristics of the text fuel this interpretation. First, Morency-Laflamme does not refer to interviews as part of his methodology. The dedicated section of the article only explains that this is a most similar case comparison and argues for the relevance of the chosen cases. The reader only discovers that interviews were conducted if she reads the footnotes. Second, the fact that interviews are quoted the same way that other secondary sources are and that there is no formula such as “he said” or “according to” in the article when Morency-Laflamme uses interviews shows that interviews are considered to be sources of factual, objective information to be used to test a hypothesis. This of course does not mean that the author fails to question the credibility and subjectivity of his sources, but that the reader is left to trust that he has solved this issue previously while conducting and analyzing the interview. It should be noted that, in any case, this is not necessarily a major issue for the validity of Morency-Laflamme's arguments, since the interview material does not appear to be the central and sole foundation of the argument. Information from secondary sources play a more decisive role, and they provide evidence that make the interviews’ content credible. In other words, even if the interviews were taken out of the article, the argument would still stand. Authors therefore use different ways of writing that confer a different epistemological status to the data: either they reflect an empirical reality that exists independently of the observer, or they only give us access to people's subjective understanding of the empirical world. However, in practice, those different conceptions are more of a continuum than two mutually exclusive categories if we consider how authors might navigate from one way of writing to another. Indeed, even from a constructivist perspective, I also often use interviews as sources of factual information, just as Morency-Laflamme does. For instance, to explain a Sudanese tradition called nafîr, I write, “This is enabled by the inclusiveness of the tradition: “It is people from extended families and neighbours, regardless of ethnic groups, social classes . . . Merchants participate, they close their shops” (interview with Asad, official at the University of Kordofan, El Obeid)” (Mahé 2018, 240). This statement is not questioned and appears to be taken at face value, though of course the ethnographic pact set up earlier in the article provides the reader with the means to reflect on its credibility. Authors therefore use interview materials for different purposes depending on what they want to do in various parts of their texts. Identifying the Characters If interviews are dialogues, then they necessarily involve at least two characters: the interviewer and the interviewee. In addition to choosing whether to actually represent those characters on the page, the author must first decide how she identifies them. This issue is related to questions of anonymization, which have received much more attention than other aspects of writing, but it also goes beyond a focus on naming. Regarding this issue, the authors of the corpus have made different choices. Here is how I refer to the participants: “Nafîr now exists in urban areas and, for instance, many roads in Khartoum have been built by relying on it often in partnership with the government (interview with Ghazi, professor at the University of Khartoum, Khartoum)” (Mahé 2018, 239). “It is a word that people know, they recognize it” (interview with Alima, former volunteer, Khartoum)” (Mahé 2018, 239). “Members were selected rather than elected (interview with Yasin, civil servant, El Obeid)” (Mahé 2018, 244). The identification strategy is always the same: quote (interview with NAME, OCCUPATION, PLACE). In addition, there is an appendix online that provides a table with the names and occupations of the interviewees, as well as the date and place of the interviews. Sometimes, research participants disappear as individuals, their words attributed to a broad group identified as “interviewees” or “informants” and therefore left without speakers: “Calls for participation were published in the newspapers. According to some interviewees, there was even a list of names” (Mahé 2018, 244). “My informants sometimes talked about ‘meetings’’ (Mahé 2018, 245). Martel uses similar formulations: “[m]oreover, while many of the activists interviewed see the grouping's interest in NTS as mostly beneficial” (2017, 560). However, while she groups different interviewees together, she categorizes them in a defined subgroup, “the activists.” These writing strategies can be explained by the concern for readability: they avoid an exponential increase in the use of citations, which interrupts reading flow when they are “in-text.” To avoid this, authors also need to consider which quotes are purely illustrative, contributing to the reality effect, and which are used for interpretative purposes. Additionally, journals place word limits on articles. This creates a strong incentive for authors to make economic use of space and rely on such simplifications, especially for material that is considered secondary. There are also usually elements of speech that many interviewees share, which can be interpreted as a sort of common knowledge or discourse within the group, even if the author has not interviewed all of its members. In those cases, citing all of the interviewees who mentioned the piece of information can grow tiresome for both author and reader. Yet, when we quote academic work but refer broadly to a stream of research, writing, for instance, “[p]revious research concluded that factionalized armies are more likely to abandon an incumbent regime during a regime crisis” (Morency-Laflamme 2018, 468), we are expected to refer specifically to the main authors that compose this “previous research.” There is no reason that similar standards of scientific rigor and integrity should not apply to quoting interviews. Yet, if these ways of writing are at odds with the requirements of transparency, they can nevertheless guarantee anonymity and confidentiality. In this case, it is important for the author to clearly state it. Unfortunately, anonymization strategies are rarely explicit in journal articles and often remain a black box (Nespor 2000). I summarily write that “[a]ll data were anonymized for consistency and as a precautionary principle” (Mahé 2018, 238). Martel does not mention anonymity, but she appears to be implementing some form of it. She references most of her participants not by name, but by their official position or the group they belong to (the diplomats, the activists…). There are exceptions: “[f]or Rizal Sukma, Executive Director of the Centre for Strategic and International Studies (Jakart)” (Martel 2017, 554). The reader can infer here that while most participants have asked to be kept anonymous, others have allowed their names to be cited.15 Martel uses the same template to identify her participants; sometimes she makes the identification in the main text, and other times only in a footnote. Morency-Laflamme also refers to his participants solely by their positions, but in his case, the positions are ones held formerly, at the historical moments he is interested in. His footnotes read as follows: “[i]nterview, National Conference representative, Lomé [Togo], November 2013”; “[i]nterview, Ex-RPT dignitary, Lomé [Togo], November 2013.” Emphasizing the interviewees’ former official positions makes sense because it indicates why these individuals were chosen to be interviewed in the first place. But is this enough? Indeed, their current positions might also be interesting to mention, because they may color and constrain the interviewees’ discourses. What the representative of a national conference held in 1991 remembers about it decades later to a foreign researcher will be different depending on whether he is today in the opposition, or in the ruling party, or retired from politics. For instance, here I refer to Kedar's former position: “[d]uring my second interview with Kedar, a former high-level official and a volunteer for the Nafir's Committee in Khartoum” (Mahé 2018, 240). This detail serves a purpose: it signals that this person has been close to power and therefore might not be disposed to speak critically about it. By giving this information, I draw attention to the limits of Kedar's testimony and place the reader in a position of trying to assess his reliability, although anonymization means I know a little bit more than the reader does. As Vainio (2013, 690) states, when we anonymize our participants, we conceal dimensions that might have impacted how we received, and treated, their discourses. As Vainio (2013, 690) states: Social psychological research on social identity, stereotypes, and anonymity reveals that information about people's identity shapes our interpretations of their motivations for what they say and do (e.g., Fiske 2009). Accordingly, we must assume that researchers, too, are affected by information about their research participants’ gender, ethnic background, educational level and so on. Reflecting on this issue and avoiding biases is something the researcher should do at the time of analysis, deciding which information about the participants is theoretically relevant, and which is not and should not be included. But as readers, we can never fully know what the author knows and is not telling us. We cannot assess how this unexpressed information affects the interpretation proposed in the article. A notable difference between the anonymization strategies used in the three texts is my use of made-up names. Though the names come from the same regional area as the interviewees actual names, this raises the issue of what information is being inadvertently conveyed. For instance, some names could be socially marked or dated in either the Sudanese or the reader's context, giving potentially false information on the informants. In addition, it could be argued that naming informants is a problematic act of symbolic violence that deprives the interviewee of agency and autonomy. Feminist scholars see nonanonymization as a way to give voice and representation to members of subaltern groups (Vainio 2013, 688). For some participants, being named in research is a way to access a degree of notoriety and social and symbolic recognition (Beliard and Eideliman 2008, 134). It is also important to consider who wants to be anonymous and who does not, because it is usually the dominant members of a given society who want to be named, as they are less likely to suffer consequences. Accordingly, there is a risk of systematically acknowledging their discourses and neglecting those of the most vulnerable. At the same time, there is a growing awareness that consent to participation in a given research is dynamic (Beliard and Eideliman 2008, 135): participants do not necessarily agree to be named once and for all, because contexts can change, and research that was deemed harmless can suddenly become sensitive, especially for researchers working in volatile settings such as authoritarian states and conflict-ridden societies. Taking this fluidity into account, the researcher may then decide to anonymize people who previously agreed to be named. The content of the research itself is subject to change as the analysis progresses, and information provided may be used to build a different argument than the one presented to participants at the time of interview without any intent to deceive on behalf of the researcher. There are, therefore, major ethical debates to consider when deciding whether to anonymize and how, especially because the various strategies can have unexpected side effects. For instance, deciding not to name interviewees and to refer to them only by their position can also mean concealing their gender, such as when Martel (2017, 555) cites “a former director at the ASEAN Secretariat,” though of course she could add the adjective “female” or “male.” It is interesting to note that this choice does not present itself the same way in other languages. For instance, in French, most nouns are gendered, so mentioning the interviewee's position actually provides two pieces of information instead of only one. In this case, Martel would have to choose between the masculine directeur, the feminine directrice, or a form of nongendered writing, for instance, directeur • trice. The question of how to anonymize is therefore not only connected to ethical or theoretical issues, but also to the language we write in, which for many scholars is not necessarily English. A whole set of social information disappears when we either erase or change names,16 and while it may be irrelevant to our arguments, we can wonder whether new patterns would emerge in Martel's work if she indicated her participants’ genders.17 However, the anonymization strategy is not only a matter of choosing between no names and made-up names. There are many other alternatives. Instead of calling my interviewee “Kedar,” I could have given him a last name, or called him Kedar L., K. Lastname, K.L, or simply K. In addition to asking whether our implicit biases regarding social classes, ethnicity, gender, and so on influence whether we refer to someone by his or her full name, we need to ask what the impact of those alternatives would be on the reader. Using a first name alone creates a degree of familiarity and closeness—at least for readers raised in societies where being on a first-name basis is a sign of intimacy—that is coherent with an ethnographic approach, where empathy18 plays a central role. On the opposite end of the spectrum, using initials only produces an effect of depersonalization. In Kafka's The Trial, Joseph K's incomplete last name reflects the shallowness and limitations of the character and distinguishes him from the other individuals that populate his surreal world, none of whom have amputated patronyms. By giving names to my participants, I attempt to personify them, to convey to the reader the idea that they are real people, not theoretical abstractions, who are impacted everyday by the authoritarian regime they live in. Compared to the strategies fiction writers use to make their characters come alive on the page, this is a highly simplistic and limited process of characterization. Defined as “the representation of persons in narrative and dramatic works” (Baldick 2008, 37), characterization “includes direct methods like the attribution of qualities in description or commentary and indirect (or “dramatic”) methods inviting readers to infer qualities from characters’ actions” (Baldick 2008, 37). As Stein (1995) explains, characterizing is not about simply giving a name or an occupation, but also involves using the unusual characteristics or specific details of appearance that also give a sense of personality. Finding the right traits to characterize is challenging, and quick characterization can easily veer into mindless stereotyping19 that “produces a stock character who, at best, comes across as less than fully human and, at worst, appears as a negative stereotype” (Emerson et al. 2011, 70). In most academic writing, providing too much detail or sophistication in a character description can be deemed irrelevant, but Stein uses examples from historians to demonstrate the possibility of successfully and subtly characterizing in academic nonfiction. Characterization is especially important in approaches that use fieldwork and ethnography, but all research tells a “story” and therefore has characters, plot, and setting. As Van Maanen explains, “common-denominator people do not make good stories. To say ‘the police sometimes kill people for mistaken reasons’ is, flatly, not a story. To say ‘Officer Allen shot Officer Roberts while both men were on a drug stakeout’ is a story. The narrator of a tale cannot converse in a story with ‘types’ of people” (2011, 105). The trick and challenge for good characterization is to avoid clichés and generalizations and to present characters that are complex and who appear as unique and distinct from others. In literature, good characterization precludes generalizations, understood here as the use of words or expressions that could be applied to any character, in other words, overused clichés. For qualitative researchers working within a more positivist perspective that aims to uncover the “regularities of nature” (Moses and Knutsen 2012, 48), there is tension between characterization and generalization understood in its classical form as the abstracting process through which “from single observations we try to draw inferences to more general formulations to be extended to future situations” (Mayring 2007). “Good” characterization might actually contribute to the impression that the findings are based on highly context-specific and anecdotal data and therefore hardly fit for generalization. Those researchers can overcome this issue by relying on understanding of generalization that are more adapted to qualitative research, such analytic (Yin 2009) or theoretical (Mitchell 1983) generalization, but anonymization is also a tool they can use. While some strategies help flesh out lively characters, others contribute to generalization by turning individuals “into usable examples or illustrations of generalizing theoretical categories, such as social classes, ethnic groups, genders, institutions, or other theoretical constructs” (Vainio 2013, 690). Of course, one could argue that good characterization is a tool for generalization, because it pushes readers to identify with the characters, recognizing themselves in them. Most importantly, anonymization strategies are not the only elements of a text creating this tension. In her analysis of Goethe's depiction of the Roman Carnival in the Italian Journey, Crapanzano draws attention to Goethe's use of tenses: “he writes in the ‘present’ tense—a tenseless tense if you will, which serves at once to give a feeling of timeless flow and to permit generalization” (1986, 65). Approaches that focus on the ideographic, the particular, therefore favor different writing styles than ones focusing on identifying the patterns and regularities of the social world. Lastly, it is important to note that interviews sometimes include a third participant who is often left unidentified: the translator. This appears not to be an issue Martel and Morency-Laflamme faced, since they do not mention translators and therefore probably did not use them, but I state that “most interviews were conducted with a translator” (Mahé 2018, 238). I further acknowledge his or her presence in the appendix, where I provide a list of the interviews quoted in the article and marks with an asterisk those that were translated. However, when I refer to those interviews in the main text, this intermediary disappears. The appendix says that the interview with Yusni was translated, but here is how it is written in the main text: “The businessmen promised to build 100 new classrooms; they then went to visit merchants, who had not been present at the meeting, asking each for a sum, which they gave voluntarily (interview with Yusni, local businessman, El Obeid)” (Mahé 2018, 246). What appears as the interviewee's voice is therefore a construct, a composite of two voices presented as one. Usually, the translated material is written in indirect speech as a means to alleviate the fact that the researcher is putting words they did not pronounce into the mouths of the interviewees, but there is one instance in the article where I use direct quotes for material from a translated interview. This creates confusion: is this a quote from the translator—a likely interpretation if the interview is marked as translated in the appendix—or does the switch to direct speech signal that the interviewee uttered the sentence in a language familiar to me? Besides, there are more creative and efficient ways of writing translators back in, an act that is necessary to research transparency and ethics. New typographical signs or different fonts can be used to signal their presence within the main text, both when writing indirect and direct speech. A simple solution would be to add the asterisk that I use to indicate translation in the appendix directly in the main text, writing, for instance: “(interview with Yusni,* local businessman, El Obeid).” Of course, this would only work with in-text citation style and would be less useful for referencing in footnotes or endnotes, since the idea is for the reader to know as she is reading that it is a translation. Here, we can see once again how various citation systems might contribute to, or impede, making distinctions between voices. Another challenge arises when material has been translated more than once. For instance, I might do an interview with the help of a translator who translates from language A to language B, which is my native language. I may then write an article in language C, translating the data from language B to C to do so. Reporting those multiple translations in the article is, I argue, another step toward increased methodological and analytical transparency, in addition to providing interesting information on the linguistic practices of social scientists. Conclusion Writing matters and is a major component in every stage of a research project. The lack of attention it has received from political scientists is consequently both surprising and detrimental, because thinking about writing equals reflecting on processes of knowledge production. By looking at how interviews are used in three articles, I have shown that writing choices reflect different conceptions of the data and the sources themselves as they are connected to epistemological positions. Authors can nonetheless navigate from one way of writing to another at different moments in a text, but they should always question what makes a piece of data credible for some uses and not for others. Applying reflexivity to writing is therefore an entry point into broader questions about research design and social science philosophy. It is critical for helping us understand the impact of our writing choices and can lead us to make better ones if needed, something I hope this attempt at textual analysis shows. Academic writing is neither a private activity nor a space of total freedom. As Bourdieu (2015) argues, it is a social activity whose content is dictated in part by the structures in which we are embedded. Academia's current organization, with the major role played by peer-reviewed journal publications in professional recognition and career advancement and the related exponential growth in the number of articles published, dictates some of our writing choices. Junior scientists are often advised to look at previously published articles in the journals they aim for as models for their own texts (Bryman 2012, 704), which helps entrench conventions that might not be suitable for all epistemological and theoretical approaches. We use the fonts and the citation styles required by journals but do not necessarily investigate their impact on the transparency of our analyses in a systematic manner. Since guidelines for quoting from interviews and other primary sources are rarely provided, we are left to devise our own solutions. While this means that a degree of flexibility exists, in practice those solutions are often found by looking into the methods of other published authors using similar materials. This can lead to the erroneous perception that those are set academic conventions that should be followed and, if we are not careful, can lead to mindless copying, especially for scholars who are less confident about their own writing. Following existing models can lead to writing that is too formulaic influenced by the preferences of specific research traditions and overall not fitted to the specific nature of the data at hand (Corden and Sainsbury 2006, 10). Reflexivity is essential to preserve flexibility and creativity. The latter is especially under stress in a context of increased competition in the academic job market and the pressure to “publish or perish.” This discourages scholars, especially recent graduates, from challenging established conventions. It leaves academics with little time to reflect on the normative values conveyed by the writing conventions they adhere to and the effects of these conventions on their readers. To conclude, while I argue that close reading is a tool well worth using, it is essential to keep questioning the structural constraints of our profession if we want to truly understand why and how we write. Acknowledgments Thanks to Stéphanie Martel and Julien Morency-Laflamme for agreeing to have their articles put under the microscope, to the anonymous reviewers for their useful suggestions, to Jean-Philippe Michaud for his help on literary analysis, and to Frédéric Mérand, Romain Busnel, and Agathe Lelièvre for their comments on early versions of the article. Footnotes 1 The notion that “writing is thinking” seems however to be widely shared, though the more deductive approaches might still be averse to it. Nonetheless, the misconception that writing is the last step in a process is often institutionally entrenched in the way graduate programs and funding schemes are designed and is repeated in the discourse of students, who generally plan to dedicate the final years or months of their project to writing. 2 See among others Marsh and Furlong (2002); Box-Steffensmeier, Brady, and Collier (2008); as well as King, Keohane, and Verba (1994); Brady and Collier (2010), though they are not textbooks strictly speaking. 3 This is an underlying question that sometimes comes into focus. Mosley (2013b, 6) tackles it directly, but briefly, when she mentions how authors refer to their interviewees by different words: participants, interlocutors, or informants. 4 The notion that social scientists are terrible writers, abusers of jargon, and abstruse formulations is widespread. Yet, in a rare empirical investigation of this question, Biber and Gray (2010) find that academic writers do not use “elaborated” and complex structures. They nonetheless use a more “compressed” style that make their texts less explicit in meaning, which poses a challenge to novice readers. 5 This is not to say that this process is what separates quantitative and qualitative. The criteria of distinction are a debated issue: while for Bryman (2012, 36) qualitative method “emphasizes words rather than quantification in the collection and analysis of data,” Gerring (2011, 362) focuses on the number of dataset observations, while other authors imply that the difference lies in the epistemological (Jackson, Drummond, and Camara 2007). 6 I provide here a brief and incomplete overview since there are countless literary devices and stylistic elements to be analyzed. The interested reader will find more complete information in one of the many glossaries on the topic, such as Baldick (2008). 7 This issue is not only about the presence of interpreters but also, for instance, about the impacts of the reliance on “international English” by either the interviewee, the interviewer, or both (Alles, Guilbaud, and Lagrange 2018, 113). 8 I borrow the typology of Marsh and Furlong (2002) rather than Moses and Knutsen (2012), who differentiate naturalism and constructivism, because those terms are more commonly used in the literature on the topic. In addition, the concept of constructivism also refers to a precise sociological approach, which could cause confusion. 9 The interviews were realized during fieldwork in Benin and Togo, information not included in the article that I learned from personal interactions with the author. 10 Interview by the author. 11 Interview, National Conference Representative, Lomé [Togo], November 2013. 12 Interview, Ex-RPT dignitary, Lomé [Togo], November 2013. 13 I note for instance that “apart from one exception, all interviewees were men and most of them part of the local socioeconomic elite. As such the voices of underprivileged and marginalized groups are less observable in the data” (Mahé 2018, 238). 14 That the interview is a discourse specifically produced for the interviewer—and even coproduced—presents theoretical and methodological challenges for practitioners of discourse analysis (Cruickshank 2012). Such data are often seen as secondary, something Martel (2017, 552) seems to acknowledge when she writes that the transcripts of interviews “complete” a corpus composed mainly of official documentation. 15 The author told me in an informal conversation in April 2018 that she mentions the names of some participants because they are well-known actors that other researchers working on ASEAN will know. It therefore adds a dimension to the discourse to name them. 16 I do not delve into this topic since it has been widely discussed elsewhere, for instance by Weber (2008) and Nespor (2000). 17 It is interesting to note here that the issue of concealing gender is also relevant when it comes to references where it is connected to citation systems. The APA style, for instance, only indicates the initial of the author's first name. 18 There is empathy between the researcher and research participants, but the ethnographer also aims to generate empathy on behalf of the reader. Of course, the first form might be fictitious, crafted in writing through the use of first names, when the reality in the field was very different. 19 Emerson et al. (2011, 70) give the following examples: “a retarded person,” “a homeless person,” and “a black/white/Asian.” References Alles Delphine , Guilbaud Auriane , Lagrange Delphine . 2018 . “ Interviews in International Relations .” In Resources and Applied Methods in International Relations , edited by Devin Guillaume , 109 – 22 . Basingstoke : Palgrave Macmillan . Google Preview WorldCat COPAC Aradau Claudia , Huysmans Jef . 2014 . “ Critical Methods in International Relations: The Politics of Techniques, Devices, and Acts .” European Journal of International Relations 20 : 596 – 619 . Google Scholar Crossref Search ADS WorldCat Baldick Chris . 2008 . The Oxford Dictionary of Literary Terms . 3rd ed. Oxford : Oxford University Press . Google Preview WorldCat COPAC Becker Howard S. 1996 . “ The Epistemology of Qualitative Research .” In Ethnography and Human Development: Context and Meaning in Social Inquiry , edited by Jessor Richard , Colby Anne , Shweder Richard A. , 53 – 72 . Chicago : University of Chicago Press . Google Preview WorldCat COPAC Becker Howard S. . 2008 . Writing for Social Scientists: How to Start and Finish Your Thesis, Book, or Article . Chicago : University of Chicago Press . Google Preview WorldCat COPAC Beliard Aude , Eideliman Jean-Sébastien . 2008 . “ Au-Delà De La Déontologie .” In Les Politiques de L'enquête , edited by Fassin Didier , Bensa Alban , 123 – 41 . Paris : La Découverte . Google Preview WorldCat COPAC Biber Douglas , Gray Bethany . 2010 . “ Challenging Stereotypes about Academic Writing: Complexity, Elaboration, Explicitness .” Journal of English for Academic Purposes 9 : 2 – 20 . Google Scholar Crossref Search ADS WorldCat Bleich Erik , Pekkanen Robert . 2013 . “ How to Report Interview Data .” In Interview Research in Political Science , edited by Mosley Layna , 84 – 105 . Ithaca : Cornell University Press . Google Preview WorldCat COPAC Bourdieu Pierre . 2015 . Sociologie Générale, Volume 1. Cours Au Collège de France (1981–1983) . Raisons d'agir/Le Seuil . Google Preview WorldCat COPAC Box-Steffensmeier Janet M. , Brady Henry E. , Collier David . 2008 . The Oxford Handbook of Political Methodology . Oxford : Oxford University Press . Google Preview WorldCat COPAC Brady Henry E. , Collier David . 2010 . Rethinking Social Inquiry: Diverse Tools, Shared Standards . Lanham, MD : Rowman & Littlefield Publishers . Google Preview WorldCat COPAC Bryman Alan . 2012 . Social Research Methods . 4th ed. New York : Oxford University Press . Google Preview WorldCat COPAC Clifford James . 1983 . “ On Ethnographic Authority .” Representations 1 : 118 – 46 . Google Scholar Crossref Search ADS WorldCat Clifford James , Marcus George E. , eds. 1986 . Writing Culture: The Poetics and Politics of Ethnography . Berkeley : University of California Press . Google Preview WorldCat COPAC Corden Anne , Sainsbury Roy . 2006 . Using Verbatim Quotations in Reporting Qualitative Social Research: Researchers’ Views . York : University of York . Google Preview WorldCat COPAC Cornut Jérémie . 2015 . “ The Practice Turn in International Relations Theory .” In The International Studies Encyclopedia, International Studies Association Online , edited by Denemark Robert A. , Marlin-Bennett Renee . Blackwell Publishing , Oxford Reference Online . Google Preview WorldCat COPAC Crapanzano Vincent . 1986 . “ The Masking of Subversion in Ethnographic Description .” In Writing Culture: The Poetics and Politics of Ethnography , edited by Clifford James , Marcus George E. , 51 – 76 . Berkeley : University of California Press . Google Preview WorldCat COPAC Creswell John W. , Miller Dana L. . 2000 . “ Determining Validity in Qualitative Inquiry .” Theory Into Practice 39 ( 3 ): 124 – 30 . Google Scholar Crossref Search ADS WorldCat Cruickshank Jørn . 2012 . “ The Role of Qualitative Interviews in Discourse Theory .” Critical Approaches to Discourse Analysis across Disciplines 6 : 38 – 52 . WorldCat Dunleavy Patrick . 2003 . Authoring a PhD: How to Plan, Draft, Write, and Finish a Doctoral Thesis or Dissertation . Basingstoke : Palgrave Macmillan . Google Preview WorldCat COPAC Eliasoph Nina . 2011 . Making Volunteers: Civic Life after Welfare's End . Princeton : Princeton University Press . Google Preview WorldCat COPAC Emerson Robert M. , Fretz Rachel I. , Shaw Linda L. . 2011 . Writing Ethnographic Fieldnotes . Chicago : University of Chicago Press . Google Preview WorldCat COPAC Fiske Susan T. 2011 . Social Beings: Core Motives in Social Psychology . New York : Wiley . Google Preview WorldCat COPAC Gerring John . 2011 . Social Science Methodology: A Unified Framework . Cambridge : Cambridge University Press . Google Preview WorldCat COPAC Hall Peter A. 2003 . “ Aligning Ontology and Methodology in Comparative Research .” In Comparative Historical Analysis in the Social Sciences , edited by Mahoney James , Rueschemeyer Dietrich , 374 – 404 . Cambridge : Cambridge University Press . Google Preview WorldCat COPAC Jackson Patrick Thaddeus . 2016 . The Conduct of Inquiry in International Relations: Philosophy of Science and Its Implications for the Study of World Politics . Abingdon : Routledge . Google Preview WorldCat COPAC Jackson Ronald L. , Drummond Darlene K. , Camara Sakile . 2007 . “ What Is Qualitative Research? ” Qualitative Research Reports in Communication 8 : 21 – 28 . Google Scholar Crossref Search ADS WorldCat Kapiszewski Diana , MacLean Lauren M. , Read Benjamin L. . 2015 . Field Research in Political Science: Practices and Principles . Cambridge : Cambridge University Press . Google Preview WorldCat COPAC Keating Michael , Porta Donatella Della . 2010 . “ In Defence of Pluralism in the Social Sciences .” European Political Science 9 : S111 – 120 . Google Scholar Crossref Search ADS WorldCat King Gary , Keohane Robert O. , Verba Sidney . 1994 . Designing Social Inquiry: Scientific Inference in Qualitative Research . Princeton : Princeton University Press . Google Preview WorldCat COPAC Lincoln Yvonna S. , Guba Egon G. . 1985 . Naturalistic Inquiry . Beverly Hills, California : Sage . Google Preview WorldCat COPAC Locke Karen , Golden-Biddle Karen . 1997 . “ Constructing Opportunities for Contribution: Structuring Intertextual Coherence and ‘Problematizing’ in Organizational Studies .” Academy of Management Journal 40 : 1023 – 62 . WorldCat Mahé Anne-Laure . 2018 . “ A Tradition Co-Opted: Participatory Development and Authoritarian Rule in Sudan .” Canadian Journal of Political Science 51 : 233 – 52 . Google Scholar Crossref Search ADS WorldCat Marsh David , Furlong Paul . 2002 . “ A Skin Not a Sweater: Ontology and Epistemology in Political Science .” In Theories and Methods in Political Science , edited by Marsh David , Stoker Gerry , 17 – 41 . Basingstoke : Palgrave Macmillan . Google Preview WorldCat COPAC Martel Stéphanie . 2017 . “ From Ambiguity to Contestation: Discourse(s) of Non-Traditional Security in the ASEAN Community .” Pacific Review 30 : 549 – 65 . Google Scholar Crossref Search ADS WorldCat Mayring Philipp . 2007 . “ On Generalization in Qualitatively Oriented Research .” Forum: Qualitative Social Research 8 . WorldCat Mitchell J. Clyde . 1983 . “ Case and Situation Analysis .” Sociological Review 31 : 187 – 211 . Google Scholar Crossref Search ADS WorldCat Montgomery Martin , Mills Sara , Durant Alan , Fabb Nigel , Furniss Tom . 2007 . Ways of Reading: Advanced Reading Skills for Students of English Literature . Abingdon : Routledge . Google Preview WorldCat COPAC Morency-Laflamme Julien . 2018 . “ A Question of Trust: Military Defection During Regime Crises in Benin and Togo .” Democratization 25 : 464 – 80 . Google Scholar Crossref Search ADS WorldCat Moses Jonathon , Knutsen Torbjørn . 2012 . Ways of Knowing: Competing Methodologies in Social and Political Research . Basingstoke : Palgrave Macmillan . Google Preview WorldCat COPAC Mosley Layna , ed. 2013a . Interview Research in Political Science . Ithaca : Cornell University Press . Google Preview WorldCat COPAC Mosley Layna . 2013b . “ Introduction. ‘Just Talk to People’? Interviews in Contemporary Political Science .” In Interview Research in Political Science , edited by Mosley Layna , 1 – 30 . Ithaca : Cornell University Press . Google Preview WorldCat COPAC Nahmias-Wolinsky Yael , Sprinz Detlef F. . 2004 . “ Introduction: Methodology in International Relations Research .” In Models, Numbers, and Cases: Methods for Studying International Relations , edited by Nahmias-Wolinsky Yael , Sprinz Detlef F. , 1 – 19 . Ann Arbor : University of Michigan Press . Google Preview WorldCat COPAC Nespor Jan . 2000 . “ Anonymity and Place in Qualitative Inquiry .” Qualitative Inquiry 6 : 546 – 69 . Google Scholar Crossref Search ADS WorldCat Nocerino Pierre . 2016 . “ Ce Que La Bande Dessinée Nous Apprend De L’écriture Sociologique .” Sociologie et Sociétés 48 : 169 – 93 . Google Scholar Crossref Search ADS WorldCat Olivier de Sardan Jean-Pierre . 2008 . La Rigueur du Qualitatif. Les Contraintes Empiriques de L'interprétation socio-Anthropologique . Louvain-La-Neuve, Belgium : Academia-Bruylant . Google Preview WorldCat COPAC Passeron Jean-Claude . 1991 . Le Raisonnement Sociologique: L'espace Non-Poppérien Du Raisonnement Naturel . Paris : Nathan . Google Preview WorldCat COPAC Rapaport Herman . 2011 . The Literary Theory Toolkit: A Compendium of Concepts and Methods . Hoboken, NJ : Wiley-Blackwell . Google Preview WorldCat COPAC Schatz Edward . 2009a . “ Ethnographic Immersion and the Study of Politics .” In Political Ethnography: What Immersion Contributes to the Study of Power , edited by Schatz Edward , 1 – 22 . Chicago : Chicago University Press . Google Preview WorldCat COPAC Schatz Edward . 2009b . Political Ethnography: What Immersion Contributes to the Study of Power . Chicago : University of Chicago Press . Google Preview WorldCat COPAC Stein Sol . 1995 . Stein on Writing: A Master Editor of Some of the Most Successful Writers of Our Century Shares His Craft Techniques and Strategies . Basingstoke : Palgrave Macmillan . Google Preview WorldCat COPAC Vainio Annukka . 2013 . “ Beyond Research Ethics: Anonymity As ‘Ontology,’ ‘Analysis,’ and ‘Independence.’ ” Qualitative Research 13 : 685 – 98 . Google Scholar Crossref Search ADS WorldCat Van Maanen John . 2011 . Tales of the Field: On Writing Ethnography . Chicago : University of Chicago Press . Google Preview WorldCat COPAC Van Puyvelde Damien . 2018 . “ Qualitative Research Interviews and the Study of National Security Intelligence .” International Studies Perspectives April. https://doi.org/10.1093/isp/eky001 . WorldCat Wacquant Loïc . 2003 . “ Ethnografeast: A Progress Report on the Practice and Promise of Ethnography .” Ethnography 4 : 5 – 14 . Google Scholar Crossref Search ADS WorldCat Weber Florence . 2008 . “ Publier Des Cas Ethnographiques: Analyse Sociologique, Réputation Et Image De Soi Des Enquêtés .” Genèses 1 : 140 – 50 . Google Scholar Crossref Search ADS WorldCat Whitehead Kevin A. , Baldry Kim . 2017 . “ Omni-Relevant and Contingent Membership Categories in Research Interview and Focus Group Openings .” Qualitative Research 18 : 135 – 52 . Google Scholar Crossref Search ADS WorldCat Wolcott Harry F. 2009 . Writing up Qualitative Research . 3rd ed. London : Sage . Google Preview WorldCat COPAC Yin Robert K. 2009 . Case Study Research: Design and Methods . 4th ed. Los Angeles : Sage . Google Preview WorldCat COPAC © The Author(s) (2019). Published by Oxford University Press on behalf of the International Studies Association. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - Aligning Epistemology and Writing: A Literary Analysis of Qualitative Research
JF - International Studies Perspectives
DO - 10.1093/isp/ekz004
DA - 2019-08-01
UR - https://www.deepdyve.com/lp/oxford-university-press/aligning-epistemology-and-writing-a-literary-analysis-of-qualitative-UmSVa4aH85
SP - 226
VL - 20
IS - 3
DP - DeepDyve
ER -