Reflexive Discourse Analysis: A Methodology for the Practice of Reflexivity

What does it take to implement reflexivity in practice? Despite demands for increased reflexivity in and beyond International Relations (IR), the relative absence of methodological and pedagogical material developed specifically for the implementation of reflexivity contrasts with the abundant literature focusing on other research practices such as research design, data collection or data analysis. As a result, scholars and students interested in implementing reflexivity in a rigorous manner that goes beyond the mere acknowledgement of one’s situatedness, often find themselves at a loss. This article aims to supplement philosophical and ethical conceptions of reflexivity by developing a methodology for the practice of reflexivity.

Reflexivity has experienced a rapid process of legitimation in IR. Having been consigned to the margins of the field early on (Keohane, 1988: 173), reflexivity has more recently been recognised as one of the discipline’s four ‘philosophical ontologies’ (Jackson, 2011) and has become the focus of numerous publications (Amoureux and Steele, 2016). Since the 1980s, scholars have developed a wide range of interpretations and definitions of reflexivity: as a historical event or product of social trajectories (Alejandro, 2016; Beck et al., 1994), a buzzword driving academic debates (Tickner, 2013), or a process of observing one’s positionality (Sjolander and Cox, 1994). ‘From constructivists to realists to feminists and post-structuralists’ (Amoureux and Steele, 2016: 1), reflexivity has shown itself capable of offering new routes for theorisation and empirical innovation on topics as varied as ‘states, sovereignty, political identity, security’ (Peterson, 1992: 1). Peterson (1992) has highlighted, for example, how the absence of reflexivity might lead to the reproduction of unconsciously gendered representations. A consensus seems to have emerged around the idea that reflexivity serves two purposes: an epistemological and a socio-political one. The former deals with the assessment of biases that limit the validity of knowledge claims. The latter refers to the awareness that knowledge and discourse co-produce the socio-political world they refer to. To put it simply, reflexivity enables scholars to do a ‘good job’ and to account for the role of social sciences ‘in shaping social reality and hierarchies’ (Leander, 2002: 602–604).

However, and despite the success of reflexivity as a research programme, the literature focusing on how to implement reflexivity in practice is scarce and scattered. The few works that have formalised ways to ‘do reflexivity’ have only done so in connection with specific topics. This specificity makes it difficult for readers to apply these insights to other domains. Consider, for example, the body of work that advances a reflexive approach to ethics (Amoureux, 2015; Lynch, 2008), which includes Ackerly and True (2008: 694) who develop a research ethic grid for ‘scholars who wish to engage in feminist-informed inquiry’. As underlined by Leander, this relative absence and the associated ‘problems linked to’ the practice of reflexivity runs the risk of scholars ‘rejecting sociological reflexivity out-of-hand’ rather than providing them with ‘a reason to explore the issue further’ (Leander, 2002: 604). These concerns join the broader literature arguing for the need to develop methodologies that are adapted to the diversity of IR theoretical frameworks and account for the premise within critical theories that knowledge and discourse produced by scholars are not neutral (Ackerly et al., 2006; Aradau and Huysmans, 2014; Brincat, 2012).

This article aims to address this gap by developing a methodology for the practice of reflexivity. To do so, I suggest that we use our own discourses as an empirical entry point, based on the common definition of discourse as ‘language in context’. Indeed, if the discourses we produce both reflect the socio-political order we have internalised via socialisation, and represent the implicit medium through which we unconsciously legitimise, naturalise and normalise this socio-political order, then discourse represents a key site for the practice of reflexivity. My definition of reflexivity encompasses the aforementioned definitions and can be summarised as follows: reflexivity is the practice of making conscious and explicit our practices, beliefs and dispositions.

Reflexive discourse analysis (RDA), the methodology I present in this article, aims to reflexively assess and transform our socio-discursive engagement with the world, so as to render it consistent with our intentional socio-political objectives and to minimise our unconscious participation in the reproduction of the socio-political order. RDA comprises three components:

- an empirical component, assessing through discourse analysis (DA) the implicit socio-political dimensions of the discourses we produce;

- a transformative component, aiming toward the self-resocialisation of the sociological dispositions associated with the discursive issues identified; and

- a reconstructive component, creating a ‘compass discourse’ (either synthesising a socio-political issue that we aim to challenge or an alternative to it) to guide RDA’s empirical and transformative work. RDA is completed for a specific text when we have achieved ‘discursive consistency’ — i.e. when the implicit dimensions of the text match the intentional socio-political objectives we have set up for it via our compass discourse, after iteratively repeating the empirical and transformative stages.

I first introduce the theoretical rationale behind RDA and show how using discourse as an entry point to reflexivity helps to address the reflexive challenges underlined in the literature. I then illustrate different aspects of the approach through an engagement with two thinkers that inspired me in undertaking this initiative: Auguste Comte and Etienne de La Boétie. Drawing on these historical examples, I explain in the last section how I used RDA to produce this very article, emphasising the need for an expansion of methodology’s scope to encompass research practices adapted to reflexivity and critical theories.

Introducing reflexive discourse analysis

From a theoretical perspective, discourses simultaneously make the world understandable by ordering it and organise the world socio-politically through this ordering capacity. Discourses are also the main medium of socialisation, the lifelong process of acquiring the dispositions that enable individuals to participate in society. Discourses not only socialise us via their explicit dimensions (for example, a formal explanation of what ‘having good manners’ means) but also via the implicit dimensions they inertly carry. As discursive agents, we unconsciously inherit and socialise others into these implicit elements of discourses, which are both a necessary condition for communication (such as the shared knowledge that lies within the consensual definitions of the words we use) and a key mechanism of the invisible reproduction of the socio-political order (as assumptions and biases are naturalised within this taken-for-granted shared knowledge).

Given the discursive nature of academic knowledge, discourses are at the core of our professional activity. They are the instrument through which we communicate research to colleagues, students and broader society. Moreover, our professional specialisation – the production, cultivation and transmission of discourse and knowledge – makes us active agents in the process of socialisation.

If, based on these premises, scholars have highlighted the interest of thinking the implications of the invisible socio-political role of discourse in tandem with questions of reflexivity, they have yet to demonstrate how to make methodological use of this relationship.

Discourse is empirically assessable material. I suggest repurposing the methods used to analyse this material – often referred to under the umbrella term discourse analysis – in order to implement reflexivity in practice. On the one hand, we use discourse as an empirical entry point to implement reflexivity. On the other hand, we use reflexivity to shed light on the otherwise invisible dimensions of our own discourses. In order to show how this can be achieved through RDA, this section lays down the theoretical rationale behind developing a discourse-based approach to reflexivity and introduces the different components of this approach.

A theoretical framework to address the recursive challenge of reflexivity

The idea that knowledge holds the potential to be either oppressive or emancipatory is a theoretical-ontological premise shared among the otherwise diverse body of scholarship commonly identified as ‘critical’. This literature aims to harness the potential emancipatory power of knowledge, but the nature of these socio-political ambitions indeed vary. Emancipatory aspirations can structure collective research agendas (as in the case of anti-racist or feminist scholarship) or be more specifically bounded within discrete pieces of work (for instance, in regards to the promotion of ‘companionship’ (Austin et al., 2019) or ‘joy’ (Penttinen, 2013) in IR). More broadly, academic engagement with normative ideals goes well beyond the ‘critical’ milieu and value-led research is traditionally accepted in IR, as illustrated by the number of works advocating for ‘peaceful’ or ‘democratic’ values in world politics.

‘Good’ intentions, however, are not enough. The aforementioned distinction between explicit (often intentional) and implicit (often unconscious) dimensions of discourse raises questions about the methodological-epistemological endeavours undertaken by scholars working within such traditions: How do we ensure that our discourses actually have the potential to produce rather than counter-produce their intended socio-political effects? How do we know what we think we know about the socio-political role of the discourse and knowledge we produce?

In the absence of practical answers to these concerns, scholars promoting reflexivity in IR have been criticised for not applying to themselves the normative stances they administer to others (Holden, 2002; Jackson, 2011: 168–169). While denouncing the harmful socio-political effects of ‘IR’ discourses, ‘critical’ scholars have been accused of excluding the discourses they produce from the scope of their scrutiny (Bilgin, 2009; Hobson, 2007). Despite the denunciation of the lack of reflexivity in IR, they have not demonstrated how to reflexively account for their own practices of knowledge production in accordance with their theoretical framework and emancipatory intentions. Moreover, it has now been shown in different domains that research explicitly challenging socio-political phenomena may have unintentional harmful effects (e.g. the surrender to global capitalism (Kapoor, 2017) or the reproduction of Eurocentrism and colonised thinking (Alejandro, 2018; Allen, 2017; Duzgun, 2020)). Contrary to what our identification with criticality might lead us to believe, the adoption of a ‘critical’ framework does not guarantee that the knowledge we produce is emancipatory (Anderl and Wallmeier, 2018; Latour, 2004).

While these observations do not imply that this pitfall can be found in every piece of IR scholarship, they do provide an empirical grounding to illustrate how scholars can unconsciously stumble into this issue. They also highlight the long acknowledged methodo-epistemological recursive challenge experienced by social scientists (Bourdieu, 1992; Giddens, 1987)¹: we are socialised into discourses and dispositions produced by the socio-political order we aim to challenge, a socio-political order that we may, therefore, reproduce unconsciously while aiming to do the contrary. The recursivity of our situation as scholars – and, more precisely, the fact that the dispositional tools we use to produce knowledge about the world are themselves produced by this world – both evinces the vital necessity of implementing reflexivity in practice and poses the main challenge in doing so. How can we use reflexivity to ensure that the discourses we produce are not only empirically valid and theoretically coherent but also potentially perform the intentional objectives behind our research project?

Two main reflexive traditions have been developed in IR, which partially address this problematic. The first tradition takes IR – as a discipline and a social field – as the object of its reflexive enquiry (see for example D’Aoust, 2012). In this type of work, the researcher is often not included in the reflexive analysis. The second tradition focuses on the self as the locus of reflexivity (see for example Dauphinee, 2015), without necessarily including the contextual sociological and political collective structures in the analysis. Following Hamati-Ataya, I agree that ‘each of these focuses offers invaluable insights for IR scholarship’ (2012: 627). However, by taking either IR’s collective structure or its individual agents as the object of reflexivity, these traditions fall prey to the ‘collective’ vs ‘individual’ and ‘structure’ vs ‘agent’ dichotomies that Bourdieu identifies as key obstacles for the production of knowledge capable of reflexively effecting structural transformation (Bourdieu, 1987). Our analytical framework needs to articulate agent and structure together, in a way that enables us to identify where and how we may recursively be (re)producing the structure, for us to be able to reflexively use this structure-agent nexus as a reflexive site for structural change.

To conceptualise such articulation, Bourdieu develops the concept of habitus. For Bourdieu, habitus names the set of dispositions that agents acquire through socialisation. Dispositions (for example linguistic and identification categories) represent both the generative and constraining dimensions of structure within the self (they enable individuals to do social things and play a social role but only within a specific framework). This conceptual move invites us to rethink the conditions of structural change, which not only result from practices agents enact onto the ‘outside’ world (for example through social movements) but also those that agents enact onto themselves, via reflexive practices: ‘Sociology provides with an extraordinary autonomy, especially when we use it not as a weapon against others or as a defensive tool but as a weapon against oneself, a tool of vigilance’ (Bourdieu, 1987: 38). However, neither Bourdieu nor the IR scholars engaging with the Bourdieusian reflexive tradition (Eagleton-Pierce, 2011; Hamati-Ataya, 2013b) have offered guidance in how to turn this conceptual exploration into a practice that one can implement onto oneself.

To address this gap, I suggest integrating elements of discourse theory into the Bourdieusian interpretation of reflexivity, as discourse theory provides a bridge between theory and practice thanks to its methodological arm: DA. More particularly, I hope that conceptualising the relationship between discourse and dispositions will help students and researchers who often struggle to imagine what reflexivity ‘looks like’ subjectively. To be able to ‘do reflexivity’ as a research practice, we must turn theoretical debates about reflexivity into roadmaps for its practical implementation. The following points summarise the key tenets of the theoretical bridge I offer:

- Unconsciously inherited dispositions represent the embodied structure of the socio-political order within the self;

- Our capacity to produce social change through reflexivity depends on our capacity to make these dispositions conscious so that they can be transformed;

- Dispositions are the socio-cognitive matrix generating and constraining discourse production which enable social agents to produce a varied but limited series of discourses;- Discourses therefore represent the symptomatic manifestation of scholars’ structural dispositional embeddedness;

- As discourses reflect dispositions, they enable us to empirically identify and target the dispositions that are incompatible with our socio-political objectives.

This framework has practical implications regarding the way we experience ourselves and our work. Following the subversion of the traditional role of the researcher precipitated by participatory action research, I argue for the need to problematise the now more commonly circulated image of the researcher as ‘an observer who is situated in a context and who is looking at the world with a specific point of view’ (Leander, 2002: 602). I invite researchers to conceive themselves as social agents engaged in the world via their discursive activity; social agents that are structured and can potentially structure this world through this engagement. Against disciplinary developments that have defined discourse as outside the scope of IR’s practical turn², I therefore argue that discourses are social practices and that this understanding is key for their amenability to reflexive intervention.³ Within the constraints of this article, I will focus more specifically on the referential and relational aspects of our socio-discursive engagement: What and who do we speak about and cite? What are the social configurations and modes of relating that we potentially normalise, legitimise or challenge through our discourses? These reflexive questions join an ongoing conversation about citational politics, shadow citational networks, and the conditional access to the myriad of simultaneous and asynchronous ‘situations of utterances’ in which we participate via reading and publication.

I say ‘potentially’ to emphasise that researchers’ capacity to intentionally produce socio-discursive effects is dependent on the socio-political conditions of reception of their discourses and is therefore partly independent from researchers’ will. This point is illustrated through the notion of ‘performativity’, the capacity of discourses to manifest into being what they refer to if the contextual dimensions are met (which can include readers’ perception of the authority – or lack of thereof – of the writer, the match between readers’ internalised writing norms and the format of the text, etc.). Nonetheless, the way we socio-discursively engage with the world through the explicit and implicit dimensions of discourse are elements that researchers can reflexively account for, and a core dimension of what is socio-politically produced if these conditions are met. For this reason, while emphasising the inherent limits associated with the methodological quest to reflexively conquer the discursive dimensions of our scholarly engagement, I have constructed a framework that lays the foundation for a discursively grounded approach to reflexivity. I developed RDA, as a methodology based on discourse, to help researchers recursively experience themselves as socio-discursively engaged agents of the socio-political world they study, and to give them tools to reflexively transform this world by focusing on empirically assessable dimensions of this engagement.

RDA: an iterative process based on discourses

Reflexive discourse analysis comprises three components: (1) the definition of a compass discourse, (2) the discourse analysis of the discourse we produce, (3) the self-resocialisation of the dispositions associated with the discursive elements we aim to transform. Below I present what each component entails, how they relate to each other and how they address unresolved challenges currently identified within the literature on reflexivity. The details of RDA’s results and implementation are explored in the sub-section ‘Implementing RDA to produce this article’.

The definition of a compass discourse

As scholars, we want to ensure that the texts we write have the potential to produce rather than counter-produce the changes we would like to see in the world. The first component of RDA – the creation of a ‘compass discourse’ – aims at defining a discourse that can bound and empirically guide scholars’ reflexive work. The compass discourse is an empirical yardstick through which we assess, via DA, whether the implicit dimensions of our discourses are consistent with our intentional objectives.

Two types of compass discourses can be used to conduct RDA. On the one hand, ‘problem-oriented’ compass discourses define the socio-political issue one aims to tackle in a way that makes it assessable through DA. By ‘assessable’, I mean that it is possible to empirically investigate via DA whether this compass discourse is implicitly present in the text under scrutiny. In a previous project, for example, I used Eurocentrism as a compass discourse to account for the implicit Eurocentric biases of the literature I had been socialised into and to assess whether the discourse I was producing was itself Eurocentric (Alejandro, 2018). I defined Eurocentrism as comprising three dimensions discursively assessable via DA: the denial of ‘non-Western’ agency; the representation of the ‘West’ as the teleological core of human change; and the universalisation of ‘Western’ practices, values and norms. I was able to assess these dimensions discursively through the analysis of tenses, active sentence construction and active vocabulary, binary pairs and processes of passivisation, specification, genericisation, normalisation and essentialisation.

On the other hand, ‘alternative-oriented’ compass discourses go one step further and address ‘the danger of reconstructionist refusal’ in which ‘scholars are most likely simply to default to, or retreat back into’ producing the same type of issues they challenge by not constructing any alternatives (Hobson, 2007: 104). This type of compass discourse requires not only the identification of the core socio-political issue we aim to challenge, but also the creation of a discursive alternative to this issue. The ‘compass discourse’, thus defined, is a tool we use as a praxeological beacon in an attempt to perform our imagined utopia via the discourses we produce.

In this article, I use RDA to address socio-cognitive discriminations, an issue that has been driving my academic engagement. Socio-cognitive discriminations, at once, prevent the circulation and diversification of knowledge and legitimise out-group discriminations. They manifest in the double delegitimisation of knowledge produced by those identified as ‘others’ and delegitimisation of ‘others’ based on the perception that their knowledge is not legitimate. As an alternative to this discriminatory phenomenon, I have constructed a compass discourse based on my interpretation of love (more details in the sub-section ‘Constructing my compass-discourse’). I define love as a dispositif led by transformative individual experiences and characterised by unifying social forces that reconfigure social organisation through the establishment of non-discriminative relationships, thereby overcoming group divisions. This definition is the compass discourse that has guided my reflexive work throughout the various iterations of this article.

Criteria for a functional compass discourse include: (1) clarity and internal coherence; (2) alignment with intentions, theoretical framework and argument (here, scholars may find themselves developing several compass discourses if they aim to address different socio-political issues within a single piece of work); (3) assessability via DA (see p.7 for criteria unpacking of what I mean by ‘assessable’); and, in the case of alternative-oriented compass discourse, (4) mutual exclusivity with the phenomena one aims to oppose. Noteworthy is that alternative-oriented compass discourses need not be realistic. On the contrary, I would encourage scholars to be utopian so that the alternative they create is as clear as possible.

The meaning of a compass discourse does not need to be the product of a consensual agreement (the compass discourse is mainly a methodological tool that researchers use for themselves, not the object of a conceptual debate). Indeed, scholars aiming to tackle the same issue might develop different and even conflicting alternative compass discourses. For example, despite pursuing the similar objective of social change as Laclau and Mouffe, my definition and use of love as a compass discourse clashes with their popular emphasis on ‘the acceptance of the ineradicability of antagonism’ as a condition for social change (Laclau & Mouffe, 2014: xvii). Finally, it is important to underscore that the intentional objectives underpinning the compass discourse we chose do not need to be shared by academia overall. Like any other method, RDA does not hold a universal moral compass. As such, RDA can also be used by scholars that hold derogatory or discriminatory positions and want to reflexively achieve discursive consistency.

Using discourse analysis on our own discourses

How do we know what we think we know about the potential effects of our discourses? How do scholars assess the assumptions about their positionality? These questions join the concerns raised by scholars regarding the lack of empirical research supporting reflexive claims (Hamati-Ataya, 2013a) and the risks involved in ‘ignoring the dynamism of positionality in practice’ (Soedirgo and Glas, 2020: 1). I suggest that we use DA as a practical answer to these questions. More specifically, the second component of RDA consists in conducting DA on the texts we produce, such as drafts and previous publications. In order to assess whether the implicit dimensions of our discourses match our intentional socio-political objectives, we use our compass discourse as a benchmark.

Discourse analysis is an umbrella term for a wide range of methodologies developed across the humanities and social sciences. These flexible methods provide an empirical grounding for the study of the invisible dimensions of the social-political order, making them popular within critical scholarship, including in IR since the 1990s (Dunn and Neumann, 2016; Hansen, 2006).

Rather than develop a new DA method, I suggest that we use the rich methodological literature already available and repurpose DA toward an analysis of our own discourses, employing the same critical enquiry as we would for more traditional textual material (like political speeches or news media). Indeed, DA enables us to generate knowledge that can be applied to various stages of reflexive analysis, which include exposing the ‘implicit, underlying, taken-for-granted or concealed’ dimensions of discourse (Harvey, 2020) and empirically assessing social agents’ positionality (Nero, 2015). DA encompasses different methods that address different disciplinary and theoretical needs – such as Foucauldian discourse analysis, discursive psychology, critical discourse analysis – as well as more ‘toolbox-oriented’ approaches (Gee, 2011). As such, it represents a set of methodologies easily adaptable to the diversity of reflexive and theoretical objectives.

A high level of expertise in DA is not required to produce meaningful results via RDA, nor does it substitute the need to reflexively assess our own discourses.⁴ Following my compass discourse defined above, in this paper, I used RDA to assess whether I have been unconsciously delegitimising those I have been socialised to perceive as out-groups and the knowledge they produce. I conducted DA on previously written texts, looking for example for the use of pronouns defining group boundaries (such as an inclusive ‘we’ vs. exclusionary ‘them’) and the collocated verbs, adjectives and lexical fields attached to these groups, in order to identify how these social groups were implicitly framed, represented and polarised in my writing.

Self-resocialisation

What do we do once we have identified discrepancies between the implicit dimensions of our texts and the intentional objectives highlighted by our compass discourse? An awareness of the existence of biases does not automatically result in the transformation of the dispositions linked to these biases. RDA aims to tackle this confusion, described by Hamati-Ataya as at the root of the reflexive programme’s failure in IR: ‘the equation of “reflexive theory” with “critical” and “emancipatory theory” and the consequent confusion of ethical/normative issues with strictly epistemic/theoretical ones’ (Hamati-Ataya, 2013a: 669).

Alongside other IR scholars (Choi et al., 2019), I suggest thinking of methods as transformative tools. Mere word substitution is a band-aid in the same way that political correctness can co-exist alongside the discriminative behaviours it aims to soothe (Bonilla-Silva and Forman, 2000). The third component of RDA aims to transform the dispositional matrix that generates the discursive elements we aim to substitute. RDA’s definition of the researcher as a social agent invites us to use reflexivity not only as a tool to expand the academic options available to us (Hansen, 2015: 17), but also to problematise and expand the dispositional options that constrain us structurally and determine the modalities of our referential and relational forms of engagement. RDA therefore implements in practice Bourdieu’s definition of reflexivity as a process that ‘opens possibilities for rational action, aiming at undoing or redoing what history has done’ (Bourdieu, 1993: 1399)⁵ . Based on the empirical results from the analysis of our own discourses, RDA methodologically supplements the study of academics’ habitus to purposefully transform our socio-cognitive dispositions through their ‘reappropriation’ (Bourdieu, 1987: 26).

The concept of ‘resocialisation’ encapsulates this practice. ‘Resocialisation’ was first developed to describe the effect of ‘total institutions’ (such as asylums and prisons) on individuals’ socialisation (Goffman, 1957). The idea of ‘self-resocialisation’, understood as the deliberate practice of unlearning and relearning elements of our socialisation, better matches the needs of reflexive scholarship. Indeed, if discourses reproduce the social and political order through socially acquired and embodied dispositions, one way of breaking this cycle is to directly account for and transform these dispositions through self-resocialisation.

How knowledge about the social world can be harnessed as know-how for self-transformation, however, remains an unresolved issue in social sciences. Practices of self-transformation therefore represent a methodological and pedagogical blind spot despite their centrality in other fields, such as therapy, self-help or spirituality. This is especially surprising given critical theories’ interest in the conditions of social change. Moreover, Foucault after Nietzsche warns us that ‘technologies of the self’ are not necessarily emancipatory as they can exacerbate the internalisation of forms of control, while simultaneously obscuring their dangerousness (Martin et al., 1988).

Further work is needed, therefore, to popularise practices of self-transformation for methodological purposes. That being said, RDA only requires us to locate the specific case-by-case dispositions associated with the discursive inconsistencies identified. Put differently, if, while conducting DA on my previous texts, I identify discursive mechanisms misaligned with my compass discourse, my next step should be to identify the dispositions reflected in these linguistic elements. For example, both my compass discourses of ‘love’ and ‘Eurocentrism’ point to the nexus between processes of identification and knowledge circulation, which led me to investigate dispositions related to ethnocentrism. Such dispositions entail in-group vs. out-group categorisation and representations as well as value systems that legitimise certain forms of knowledge at the expense of others.

New methodological opportunities arise once we acknowledge that discourses are empirically accountable manifestations of inherited dispositions, that methods designed to make explicit the implicit dimensions of discourse exist, and that socialisation is a continual process that we can consciously influence. RDA is an iterative process based on the comparative assessment of different discourses. By comparative, I refer to the comparison between the compass discourse and the discourses we produce, and between the different discourses we produce at different points in time. By iterative, I mean that the process is to be repeated until discursive consistency is achieved. We cannot fully apprehend the effects of self-resocialisation while we are undertaking it and we need to assess whether our efforts to transform our dispositions worked as intended by engaging in several phases of discourse analysis. The goal is then to iteratively transform these dispositions until the implicit dimensions of a text match its intentional objectives.

This methodology for the practice of reflexivity can be used in combination with any kind of empirical and theoretical work. Like other methodological practices, the process itself does not need to be shared with the reader but can if it bolsters our argument. As the ambition of this article is methodological, I will share in the following section the practices that inspired me to develop this methodology before unpacking how I used it to produce this paper in the final section ‘This article as a recursive performance’.

Illustrating the approach with two historical cases

My formulation of RDA is indebted to two texts: Positive Polity by Auguste Comte (1975 [(1851]) and Discourse on Voluntary Servitude by Etienne de La Boétie (1549). These texts do not develop a ‘methodology’ per se. Both of them, however, identify love as an alternative to the socio-political order they aim to challenge. Each illustrates one aspect of RDA. Comte’s text is a rare account of self-resocialisation in social sciences, as he recounts how he transformed himself, as a knowledge producer, to create a better society. Discourse on Voluntary Servitude exemplifies how a text can achieve discursive consistency with an implicit value system that legitimises an alternative to what the explicit argument criticises.

Auguste Comte’s experience of self-resocialisation

Over a century ago, the French philosopher Auguste Comte identified the opportunities and challenges engendered by the co-production of scientific discourses and the social and political order. Famous for being the founder of positivism, few people know the reflexive work at the foundation of Comte’s prolific production. Born at the turn of the 19th century, he experienced the social upheavals and political crises spanning through France and Europe during those turbulent times. Comte believed in the capacity of science to improve the human condition, but scientific knowledge had not yet been applied to the social-political problems of Western Europe. On the one hand, natural scientists had developed a scientific approach to knowledge that focused exclusively on the natural world; on the other, French ideologues produced ideas to influence the social world but did so within the essayist tradition. Comte established positivism as an epistemological framework that aimed to bridge the gap between the natural scientists and ideologues, resulting in a new science: a science about the social.

Concomitant with the birth of positivism and social sciences was the recognition of the recursive condition of social scientists and the challenges attached to producing discourses directed towards social change. A new way of thinking the world was necessary but this endeavour raised a practical problem that Comte endeavoured to tackle as part of his positivist method. The problem he identified goes as follows. Facts and knowledge are phenomena produced by a relation between the objects and the subjects of perception, and generated by different dimensions of experience: the body (through the senses), the sentiments, and the mind: ‘In a word, every phenomenon supposes a spectator: since the word phenomenon implies a definite relation between an object and a subject’ (Comte, 1975: 356). Knowledge and discourses of knowledge are thus a product of socialisation. Accordingly, how can one produce discourses capable of guiding society toward a better alternative when one is the product of this society?

As a result, one needs to transform oneself in order to produce knowledge that can transform the world. The transformation of an individual’s dispositions is the key to the production of alternative discourses that construct new social and political configurations. Comte perceived this individual transformation as the condition of possibility for social and political transformation. He decided to take himself as the object of his experiment in order to become the catalyst and prototype of a new society in ‘the dedication of a life so specially renewed to the direct object of Western reorganisation’ (Comte, 1975: xiv). The solution Comte proposes is a praxis of self-resocialisation comprising ‘two successive undertakings, the one mental, the other social in purpose’ (Comte, 1975: x).⁶

The first, so-called objective stage is the most famous, and it seems that positivism is often reduced to a superficial and embryonic vision of it. This phase aims at unlearning one’s dispositions in order to empty the mind and identity of its content and enable new frameworks of perception and interpretation. Rather than an objectivist switch easy to turn on, Comte explains how this ‘laborious training’ plunged him into years of psychiatric instability and depression (Comte, 1975: 14). After 12 years of ‘objective’ work, Comte believed to have achieved a sufficient level of ‘cerebral disponibility’⁷ and the capacity to perceive the ‘involuntary shortcomings’ of his socialisation (Comte, 1975: x).

The objective phase, however, is not an end in itself. From there, he considered himself ready to embark on the second stage of his journey, the conscious reconstruction of himself as a social subject, in which he equipped himself with a new ‘affective culture’⁸ by means of a self-resocialisation based on love. Comte defines ‘universal love’ as positivism’s ‘principal and final characteristic’ (Comte, 1975: xxxvi). The second stage of the process consists in the embodiment of dispositions that produce relationships more aligned with the type of society he aims to create. Comte identifies this stage as the subjective phase, which took him six years to achieve. The feelings rather than the mind define the nature of positivism, which in the end, according to Comte, is social.

He describes the transition from the first to the second phase in the following terms:

Comte aspired to reorganise the way history, his family and society had taught him to relate to the world. He used the three relationships he considered most important in his life as the starting point of his reconstruction; turning his relationships with his mother, with the woman he loved and with his maid whom he regarded as his daughter into ideal types of his modes of relating with people he experienced as of superior, equal and inferior status. Positive Polity is thus dedicated to the person he considered most important in the construction of his epistemological project, the love of his life:

Comte established positivism as a spiritual practice, with the intent of initiating himself into becoming the prototype of this transformation based on harmony and love.

The discourse on voluntary servitude as an example of RDA’s output

Considering its time-consuming cost and apparent socio-psychological challenges, is it worth engaging in RDA or is it a distraction from research proper? RDA can be used in both a bounded and an endless way: bounded, if we use it for a methodological purpose, in a specific research project; endless, if, instead of aiming for discursive consistency in the text, we aim for discursive consistency in the self, using each piece of text as an opportunity to reflexively transform the self in a journey of coming to embody our own imagined utopia. Comte committed his life to this boundless reflexive experiment. It is, however, not what RDA is designed for, even if it could be used for this purpose.

The Discourse on Voluntary Servitude illustrates a case of discursive consistency, an example of how the final stage of a text on which RDA has been conducted may look after a bounded implementation of the method. Written around 1549 by Etienne de La Boétie, The Discourse is one of the first texts that theorises domination as a social configuration reproduced by the participation of political subjects in a system they may criticise but still follow. This theoretical argument is now relatively common within critical scholarship. The innovative contribution of this text lies elsewhere: La Boétie’s text implicitly legitimises an alternative, mutually exclusive, value system to the one supporting the tyrannical system he explicitly opposes.

The text does not merely state alternative values. Instead, these are embedded within its implicit dimensions in a way that normalises and legitimises an alternative to what we can expect readers to perceive as valuable. This discursive move opens a socio-cognitive space that readers can use to decenter themselves from the value system supporting the tyrannical order into which they have been socialised. The recursive process occurring in this text, which also appears in other texts, inspired me to use DA not only as a practice for knowledge production but also as a practice for knowledge communication. It helped me become more sensitive to the discursive mechanisms present in my writing. Here, I summarise the three main dimensions of The Discourse that allow this text to achieve discursive consistency.

First, the text develops an analytical framework that highlights the agency of political subjects in the process of domination to support the argument that political subjects reproduce domination they are subjected to. Domination thus understood relies on the combination of different factors:

- Habits acquired through socialisation that reproduce the tyrannical socio-political system through naturalisation (La Boétie, 1549: 48);

- Interests that incentivise subjects to engage with the tyrannical pyramidal order in the hope of benefiting from and advancing within it (La Boétie, 1549: 48);

- A value system that depicts the position of the tyrant as privileged, prevents subjects from questioning their habits, and entices participation in the competitive pyramidal order. Despite complaining about tyranny, subjects of tyranny tend to envy the position of the tyrant and thus pursue a position closer to his without questioning the actual sacrifices this pursuit requires. The value system therefore contributes to the continuous oiling of the reproduction of the system. As long as it exists, subjects of tyranny will be willing to replace those positioned above them in the pyramidal system whenever possible (La Boétie, 2002: 14–15).

Second, La Boétie discursively constructs his text as a situation of recursivity in which he and the readers are included in the political order he analyses. This discursive construction transforms the different dimensions highlighted by the analytical framework into political sites where readers are called to implement their reflexivity in order to free themselves from the yoke of servitude. Different mechanisms enable this recursive effect. Some of them are implicit (such as the use of the first and second person) while others are explicit (as when La Boétie invites the readers to reflexively analyse their socio-political engagement in order to resist the hierarchical order presided over by the tyrant) (La Boétie, 1549: 5, 48). In addition, the introductory paragraph links the critical theoretical framework and the praxeological ambition of the text by inviting readers to reflect on their own practices (namely, the relation between criticising and acting) and to question the tensions that may exist between discourses, values, truth and agency.

Third, the text legitimises a value system that supports an alternative socio-political order and delegitimises the value system linked to the socio-political order it opposes. It does so by exhalting love and friendship. La Boétie shows how love/friendship are not only superior values to the ones supporting the system he denounces, but also incompatible with the hierarchical relations necessary to the reproduction of the system. He shows how the dispositions required to be dominant – as epitomised by the status of the tyrant – are incompatible with those needed to develop the most valuable type of relationships: loving friendships.

La Boétie discursively constructs an alternative relationship between the readers and the tyrant, and an alternative vision the readers can have of themselves. Instead of implicitly valuing the powerful position of a man who is criticised on the basis that he ‘can do what he wants’, La Boétie shows that the tyrant can only do what he wants in a limited range of possibilities, which exclude the conditions of a good life: peace of mind (because he lives in a state of paranoia where everybody wants to take his place), friendship (because his position requires him to develop abusive and hierarchical relationships and dispositions) and the company of good people (because their dispositions are incompatible with the tyrant’s value system). Behind the apparent privilege surrounding dominant positions lies real obstacles to what life has to offer. These obstacles may not even be perceived by those in dominant positions themselves if their socialisation does not allow them to imagine an alternative to their situation. However, for those who can, there is more reason to pity than to envy the tyrant. At the same time that La Boétie denounces tyranny, he discursively legitimises an alternative value system, thus implicitly performing its analytical framework through a potential resocialisation of the reader.

In conclusion, each example illustrates an aspect of RDA: the experience of self-resocialisation narrated by the researcher experiencing it and a text achieving discursive consistency between the implicit dimensions of discourse and its explicit objectives. These cases inspire us to engage in reflexive endeavours and illustrate some of RDA’s practicalities. Still, they remain vague, quite disconnected from the daily academic practices of the 21st century and do not offer clear guidance on how to engage such processes in practice. In the final section, I address these missing links by unpacking how I used RDA to produce this article.

This article as a recursive performance

Inspired by these authors, I developed RDA as a methodology that would enable me to put my reflexive ambitions into practice. In fact, I used this methodology to create this methodology. This last section shows how I have written this article neither as a normative-programmatic stance, nor an acknowledgement of what others have done before, but as a performative exercise in which I implement the methodology while enunciating it. I sought to design a methodology that would not only be a means towards transformation but also a discourse that demonstrates the transformation produced by this methodology. This article thus constitutes a recursive performance.

I first provide an overview of the conceptual-theoretical underpinnings of my compass discourse. I then unpack how I used RDA to produce this article and what the effects of this method on my socio-discursive engagement were. I conclude by discussing how RDA challenges the traditional scope of methodology in and beyond IR.

Constructing my compass discourse

The implementation of the empirical and transformative components of RDA (DA and self-resocialisation) depends on the appropriate construction of a compass discourse. To illustrate RDA in this article, I have focused on an issue at the core of my academic engagement: socio-cognitive hierarchies which prevent the circulation and diversification of knowledge while legitimising out-group discriminations. Inspired by the works of Comte and La Boétie, I explored the idea of love as a socio-political alternative. The widespread use of this word outside academia and the relative neglect of it in social sciences call for an effort of disambiguation to pinpoint a definition I can use as an operational compass discourse.

Despite a growing IR literature on feelings and emotions, love remains under-investigated in the discipline, both as an ontological object and framework of analysis.⁹ In social sciences more broadly, scholars have discussed the ambiguous nature of love, both as a site of power asymmetries (García-Andrade et al., 2018; Gregoratto, 2018) and a mode of resistance and political transformation (Boltanski, 1990; Negri and Hardt, 2005). Research more closely related to my objectives has investigated Paulo Freire’s concept of ‘radical love’ as a strategy to research and teach differently (Douglas and Nganga, 2015; Jolivette, 2015).

I am interested in another approach to love, one that contributes to this tradition but draws upon different sources. Despite its radicalism, this approach has gone unnoticed due to its implicit naturalisation within the use of another category. When Lévi-Strauss introduced structuralism in social sciences, he posited humanity as the unit of analysis of his newly established structural anthropology (Levi-Strauss, 1996). Humanity is conceived as a social group that functions as a whole and shares a common trajectory despite the superficial divisive perceptions its members can hold of each other. Indeed, Lévi-Strauss argues that, contrary to the structuralist unitary understanding of humanity, humans have in fact tended to confine their perception of what constitutes humankind ‘to the borders of the tribe, the linguistic group, or even, in some instances, to the village. . .’ (Levi-Strauss, 1952: 12).

Lévi-Strauss’ ontological-theoretical choice secularises and legitimises in social sciences the Christian concept of humanity, originally developed in the Middle Ages as a cognitive tool to implement the New Testament’s core sociological innovation, namely, that individuals can overcome discrimination and hierarchisation of those they perceive as ‘others’ through the practice of unconditional love. The New Testament is a collection of texts that describes the journey of a Jewish man (Jesus) who gained access to a knowledge (all humans are equal in creation and deserve equal love) and experience (the perception of this equal value and the experience of a love independent of the social characteristics of its recipients) in direct contradiction with the ethnocentrism of the group to which he belonged. The text describes the struggles he endured as he tried to embody this knowledge/experience through a transformation of his dispositions. On the one hand, discourses of unconditional love are shown to be politically subversive as they challenge the existing hierarchies of the time. On the other hand, the journey from knowledge to embodied knowledge is described as a constant challenge, as the text shows how Jesus repetitively falls back into the unconscious discriminatory patterns he had been socialised into before being reminded of his own teaching by others (including the victims of his prejudices) and reflexively acknowledging and correcting his behaviour.

Building on these insights, I define love as a dispositif led by transformative individual experiences, and characterised by unifying social forces that reconfigure social organisation through the establishment of non-discriminative relationships, thereby overcoming group divisions. I thus adopt an emancipatory interpretation of love as a basis for my compass discourse by focusing on a bounded and radical definition of love, without denying that other practices, also defined as ‘love’, entail discriminative and hierarchical dimensions.

Love as a compass discourse is a utopian move, but this compass discourse is not a ‘Kumbaya speech’. This discourse is a practical solution that addresses issues that have been empirically demonstrated in sociology and anthropology of science and knowledge. The idea that science is a social field like no other (organised around Mertonian principles, for example) has long been problematised (Vinck, 2007). Processes of socialisation that determine who we unconsciously experience as in-groups and out-groups and what we perceive as legitimate and illegitimate knowledge also occur in academia. They establish hierarchies both specific to academia and contribute to non-academic social and political struggles (Bourdieu, 1984). Socialisation comprises the acquisition of knowledge perceived to be legitimate within the socialising groups and the subsequent legitimation of the individual manifesting this identity and knowledge. As knowledge is linked to identity, any epistemic breakthrough potentially results in a sociological break up. Producing a knowledge that departs from the norms legitimised by the in-groups challenges the relation individuals producing such knowledge have with their group of belonging. Based on these theoretical and empirical elements, a compass discourse such as love — which holistically encompasses feelings and value systems — supports the socio-political (and not only cognitive) aims of RDA as well as the multi-dimensional aspects of the discriminations I aim to tackle in this specific article.

Implementing RDA to produce this article

Having defined my compass discourse, I undertook the second stage of RDA, the discourse analysis of my own discourses. The objective was to assess whether the texts I had produced so far aligned with my compass discourse or, on the contrary, implicitly reproduced the socio-cognitive discriminations I aimed to challenge. The results of this enquiry revealed discursive inconsistency, as manifested by the discursive mechanisms highlighted on p. 9. Indeed, my discourses systematically legitimised those I identified as my academic in-group – so-called ‘critical’ scholarship – and delegitimised those I identified as the attendant out-group – so-called ‘non-critical’ or ‘mainstream’ scholarship. This led me to unwittingly scapegoat the latter and deny its potential epistemic contributions to my work.

Based on this acknowledgement, I turned to RDA’s third stage – self-resocialisation. The discursive mechanisms identified highlighted how the identification categories I inherited from my socialisation into ‘criticality’ dichotomously organised my perception of academic knowledge and agents according to the binary ‘critical’ vs ‘mainstream’. Expanding my analysis to the literature in which I was trained and that I was using as an analytical framework helped me pinpoint the polarised connotative associations attached to these essentialising categories. In the literature, ‘mainstream’ scholarship is identified as solely responsible for all of IR’s ills (Hamati-Ataya, 2013a; Hobson, 2014; Knutsen, 2014; Van der Ree, 2014; Vasilaki, 2012) and negatively framed to mirror the positive self-defined characteristics of ‘critical’ scholarship. This use is consistent across academia, as ‘mainstream’ scholarship is generally opposed to what is ‘new’ (Mikkola, 2017) and ‘heterodox’ (Spash, 2012). More specific to IR is the conflation of ‘mainstream’ and ‘positivist’ scholarship as a result of the post-positivist movement that swept the discipline in the 1990s (Lapid, 1989).

RDA enabled me to take perspective on these dispositions in which I had been socialised by the literature. This method gave me tools to reflexively account for my referential and relational socio-discursive engagement and expanded the range of options I perceived as legitimately available to me. Using RDA enabled me to make practical reflexive choices more in line with my compass discourse. It enabled me to produce less discriminative discourses.

I did not pick Comte, La Boétie, Lévi-Strauss and Jesus randomly. I purposefully looked for solutions to the ‘critical’ problem of practical reflexivity in the literature I had been socialised to perceive as uninteresting, backward and dangerous – or to use the common epithet: ‘mainstream’. These authors symbolise some of the categories my identification with criticality had socialised me to dismiss: positivism, the old-fashioned, structuralism and Christianity. I was socialised as an IR student at a time when one could not be ‘critical’ without being ‘post-structuralist’ and ‘post-positivist’, even though I had never been directly exposed to structuralism or positivism, nor read the original texts associated with these movements. Calls for diversity in IR sometimes emphasise the interest of using religion as a way to expand IR theoretical horizons but this seems to only apply to supposedly ‘non-Western’ religions – as if IR theory was already infused with Christian ideas and as if Christianity was inherently ‘white’ and ‘Western’. In a constant ‘neophilic’ appeal (Neal, 2019) for the ‘post’, the ‘next new thing’ and ‘taking the next step’, I was generally discouraged to think that radicalism could be found in the past or before the word ‘critical’ made its way into social sciences. Choosing to include such references rather than exclude them, choosing to present them in a positive rather than negative light, choosing not to merely state their heuristic value, but rather legitimise them implicitly by using them and integrating them to RDA’s architecture and my compass discourse, all represent choices I made available to myself through the dispositional transformation produced by RDA.

In accordance with my compass discourse, the discourse produced in this article also enables the construction of social bridges that overcome group divisions between traditionally polarised groups: ‘critical’ vs ‘mainstream’. RDA, as a knowledge and practice built upon knowledge produced by both groups, is a relational knot. It problematises the unapologetically antagonising essentialisation of these groups. It dissolved the naturalising power this categorisation had on me and enabled me to question the persuasion that these groups were inherently motivated by opposing objectives and interests. Using RDA throughout this project pushed me to reflexively account for the origin and effects of my identification and challenged the perception I had of these bodies of scholarship, which resulted in the production of alternative discourses. If this already constitutes a contribution in itself, I hoped to go one step further by supporting readers in reflexively problematising their own potential socialisation into criticality. I tried to write this article in a way that makes them experience the process I myself went through.

Recursive problems require recursive solutions, which traditional writing models can hardly achieve. I consciously avoided presenting all my results and the recursive unpacking of this article as ‘information’ at the beginning of the article in order to give readers the time and space to experience the potential resistances produced by their own dispositions towards the referential and relational options I mobilise in this article. As mentioned above, conditions of performativity are external to the text and epistemic breakthroughs lead to sociological breakups. Accordingly, I anticipate that some readers will drop the reading of this article before the ‘recursive twist’ is revealed, due to the referential and relational choices I have made as a result of me departing from my ‘critical’ socialisation. RDA does not enable scholars to control how RDA-transformed discourses will be perceived, and the effects of the discourses we produce via RDA might not be the same as intended. Other methods need to be used to assess the conditions of reception of discourses. For example, socialised in French secularist social sciences, I was expecting ‘reflexive Jesus’ to be the most likely breaking point of the relationship between me and my readers – a relationship that is a necessary condition for RDA to potentially contribute to the resocialisation of readers via RDA-produced discourses. To my surprise, the reference to love, in a social scientific article, is what has generated the most passionate resistances so far.

Towards expanding the scope of methodology

Methodology is sometimes perceived as a decontextualised set of techniques one uncritically ‘applies’ without taking into account the social and political localisation intrinsic to all knowledge production. Against this perception, this article builds on authors underlining the need to ‘take methodology back’ to critical theories (Hansen, 2006: xvi), following long-standing methodological traditions that develop methodological approaches for social change (Feyerabend, 1975). Accordingly, I argue that a redefinition of the status and practice of methodology is needed to overcome the challenges raised by reflexivity and implement methodologies aimed at turning reflexivity into a research practice beyond meta-epistemological statements.

I argue that methodology needs to be ‘detechnicised’, resociologised and repoliticised. The difference between framing scientific activities as a set of practices (through which researchers interact with other subjects and engage in relationships that produce social configurations and institutions) rather than as a set of techniques (through which researchers account for and transform objects in the world) is at the core of Habermas’ work on the relation between science and ideology (Habermas, 1971). Habermas argues that it is within the technicisation of science – resulting from the denial of science as a set of practices and resulting in the perception that scientific work is socially neutral – that the ideological naturalising power of science lies. Reflexivity – and RDA as a methodology to implement reflexivity – only fully makes sense if we experience our work as a set of social practices in the Habermasian sense. Scholars have already successfully put on the agenda the need to make more evident the political underpinnings of everyday professional activities, such as curricula development and reading lists (Phull et al., 2019; Tuhiwai Smith, 1999). I argue that this shift from technicisation to de-technicisation needs to be systematised and expanded to all research practices, grounding IR’s practical turn methodologically.

Thus understood, methodology extends beyond data collection and interpretation to encompass all individual and collective activities of reflection, formulation and debate about any research practice. In other words, methodology can be defined as the dimension of research practice that deals with the way we do things, the overall framework (or discourse) we develop to structure and disclose the underpinnings of the choices we make in order to optimise our capacity to achieve our (cognitive and socio-political) academic objectives. Not engaging with the methodological dimensions of our work — that is to say, choosing not to make an effort to render transparent the way we conduct research — is, therefore, a political choice that favours the reproduction of the socio-political order happening through us.

Conclusion

How can we implement reflexivity as a methodology that can be taught alongside other research practices and can be tailored to the diverse objectives of scholars? The aim of this article is to turn reflexivity into a distinctive practice of research that goes beyond the traditional meta-discourse about the situatedness of the researcher.

Towards this end, I developed a methodology of reflexivity based on discourse that I call reflexive discourse analysis. RDA guides scholars in reflexively accounting for and transforming their socio-discursive engagement. It does so by guiding them in assessing the implicit dimensions of their discourses and ensuring these are consistent with their intentional objectives. RDA is an advanced second-order methodology that comprises three components: (1) the construction of a compass discourse that synthesises a socio-political issue we aim to challenge or its alternative; (2) the discourse analysis of the discourse we produce to empirically assess the discursive consistency between them and our compass discourse; and (3) the transformation of the dispositions associated with the discursive inconstancies we have identified. To illustrate the method, I recursively used RDA to produce this very article, and show how it enabled me to produce less discriminative discourses that were more in line with my compass discourse based on love. In so doing, I aimed to demonstrate the difference between statements of intentions about reflexivity and empirically grounded methodological practices of reflexivity.

RDA offers a contribution to practical discussions about methodologies for reflexivity. However, as is the case with all methodologies, it has its limitations. As I first developed RDA for my own practice in the absence of adapted resources I could use, RDA draws on pre-requisite knowledge (discourse analysis and technologies of the self), which are unevenly available in academia. Further work is required to make reflexivity as a research practice accessible to a broader readership, and more readily usable for scholars equipped with different skill sets. Another limitation is that RDA does not shield us from the psycho-emotional unease that results from the continual updating of our reflexive goals. Rather, it painfully sheds light on the limits of our reflexive knowledge. The acknowledgement of one’s own inconsistencies, while an objective reflexive accomplishment, often leads to shame and guilt (over the involuntary effects our past practices) and anxiety (regarding all our practices that are yet to be reflexively examined). For example, if my compass discourse based on love enabled me to produce less discriminatory discourses regarding my ‘critical’-related ethnocentrism, many other iterations of RDA would have been necessary to enable this article to match my ideal socio-discursive engagement taking, for example, into account dimensions such as ageism, ableism, gender or race. I try to balance these paralysing thoughts with an optimistic analogy: as the highest marginal value of physical activity lies in the move from sedentarity to exercising twice a week, I believe the highest marginal value of reflexivity lies in the move from reflexive speculation to reflexive action. Practising reflexivity empirically where we can, one compass discourse at a time, is better than nothing.

In addition to addressing the question of how to practice reflexivity in and beyond IR, this article contributes to revisiting what it means to be critical and rethinking criticality in research and writing. Based on a sociological understanding of research practice, this article provides a playful and hyper-modernist demonstration that the acknowledgement of the co-production of discourse, knowledge and socio-political order does not imply the demise of empirical work nor does it prevent the development of methods for operationalising reflexivity. Rather, RDA takes seriously the social theories of discourse underpinning critical theories in order to internalise them reflexively in our research practices. If IR (and academia more broadly) is an institution that produces discourses of authority for different audiences, it is also a professional sector that hires thousands of citizens across the world who, maybe more so that other individuals, can influence other people’s practices precisely due to the audience-oriented nature of their job. What are the modes of relating we are socialising the next generation into, and do these modes of relating meet the needs of the world’s most pressing issues? Can professionals trained in social theory use this training to produce a world different than the one they have inherited when they are themselves torn by thin-skinned parochialism and dehumanising quarrels? RDA is a tool I developed to extend the scope of critique so as to critique ‘critical immunity’ itself, and therefore to contribute to extending the scope of methodology to match the needs of critical theorists and criticality. As I showed in this article, using reflexivity as an emancipatory practice with targeted socio-political effects is possible, albeit difficult. It requires radical, dedicated commitment and efforts that extend beyond what is usually defined as academic work.

Acknowledgments

I would like to thank Felix Anderl, Paul Beaumont, Gokhan Ciflikli, Alex Stoffel, the two anonymous reviewers as well as the participants of the workshops ‘Performing Utopia: Discourses of Love, Compassion and More in International Relations’ and ‘Post-Critical IR: Activism, Critique, Change’ for their valuable feedback.

Funding

The author received no financial support for the research, authorship, and/or publication of this article.

Notes

1. Used in mathematics, linguistics or informatics, the concept of ‘recursivity’ enables us to think about the existence of an element included within itself (e.g. a recursive programme that requires an algorithm that uses itself to reach its results). As social scientists, the recursivity of our condition deals with the fact that we are both subjects (as discourses are the medium through which we analyse) and objects of the academic discourses we produce (as we are social agents belonging to the world we analyse).
2. As underlined for example by Epstein (2013).
3. In Foucault’s words, ‘discourses are not about objects; they do not identify objects, they constitute them and in the practice of doing so conceal their own invention’. Foucault therefore invites us to treat discourses not only ‘as a group of signs (signifying elements referring to contents of representations) but as practices that systematically form the object of which they speak’ (Foucault, 1972: 49).
4. See for instance Billig’s discourse analysis of discourse analysts’ discourse, which highlights how these scholars are themselves not immune from falling prey to the recursive challenge presented above (2008: 783)
5. Cited in Leander (2002).
6. These follow a preliminary stage in which he relates that he had to enter a continuous meditation of 80 hours after he realised that the immense task of experimenting and formalising positivism would happen through him (Comte, 1975: x).
7. Author’s translation.8.Author’s translation.9.For an exception see (Hartnett, 2018).

References

Ackerly, B, True, J (2008) Reflexivity in practice: power and ethics in feminist research on international relations. International Studies Review 10(4): 693–707.

Ackerly, B, Stern, M, True, J (eds) (2006) Feminist Methodologies for International Relations. Cambridge: Cambridge University Press.

Alejandro, A (2016) Walking the reflexive talk. Available at: http://www.e-ir.info/2016/09/17/walking-the-reflexive-talk/ (accessed 7 October 2020).

Alejandro, A (2018) Western Dominance in International Relations? The Internationalisation of IR in Brazil and India. London; New York: Routledge.

Allen, A (2017) The End of Progress – Decolonizing the Normative Foundations of Critical Theory. New York: Columbia University Press.

Amoureux, JL (2015) A Practice of Ethics for Global Politics: A Practice of Ethics for Global Politics. New York: Routledge.

Amoureux, JL, Steele, BJ (eds) (2016) Reflexivity and International Relations – Positionality, Critique, and Practice. London: Routledge.

Anderl, F, Wallmeier, P (2018) Modi der Kritik des internationalen Regierens. Ein Plädoyer für immanente Kritik. Zeitschrift für Internationale Beziehungen 25(1): 65–89.

Aradau, C, Huysmans, J (2014) Critical methods in international relations: the politics of techniques, devices and acts. European Journal of International Relations 20(3): 596–619.

Austin, JL, Bellanova, R, Kaufmann, M (2019) Doing and mediating critique: an invitation to practice companionship. Security Dialogue 50(1): 3–19.

Beck, U, Giddens, A, Lash, S (1994) Reflexive Modernization. Stanford: Stanford University Press.

Bilgin, P (2009) The international political “Sociology of a Not So International Discipline”. International Political Sociology 3(3): 338–342.

Billig, M (2008) The language of critical discourse analysis: the case of nominalization. Discourse & Society 19(6): 783–800.

Boltanski, L (1990) L’amour et La Justice Comme Compétences: Trois Essais de Sociologie de l’action. Paris: Editions Métailié.

Bonilla-Silva, E, Forman, TA (2000) ‘I am not a racist but. . .’: mapping White college students’s racial ideology in the USA. Discourse and Society 11(1): 50–85.

Bourdieu, P (1984) Homo Academicus. Paris (ed.). Editions de Minuit.

Bourdieu, P (1987) Choses Dites. Paris: Editions de Minuit.

Bourdieu, P (1992) Double Bind et Conversion. In: Réponses: Pour Une Anthropologie Réflexive. Paris: Le Seuil, 217–224.

Bourdieu, P (1993) Comprendre. In: Bourdieu, P (ed.), La Misère Du Monde. Paris: Le Seuil, 1389–1447.

Brincat, S (2012) On the methods of critical theory: advancing the project of emancipation beyond the early Frankfurt school. International Relations 26(2): 218–245.

Choi, S, Selmeczi, A, Strausz, E (eds) (2019) Critical Methods for the Study of World Politics: Creativity and Transformation. London: Routledge.

Comte, A (1975) Positive Polity. London: Spottiswoode and Co.

D’Aoust, A-M (2012) Introduction to the sociology/ies of international relations. Journal of International Relations and Development 15(1): 90–97.

Dauphinee, E (2015) Narrative engagement and the creative practices of international relations. In: Lamoureux JL and Steele BJ (eds), Reflexivity and International Relations. London: Routledge, 44–60.

Douglas, T, Nganga, C (2015) What’s radical love got to do with it: navigating identity, pedagogy, and positionality in pre-service education. International Journal of Critical Pedagogy 6(1): 58–82.

Dunn, KC, Neumann, IB (2016) Undertaking Discourse Analysis for Social Research. Ann Arbor: University of Michigan Press.

Duzgun, E (2020) Against eurocentric anti-eurocentrism: international relations, historical sociology and political marxism. Journal of International Relations and Development 23(2): 285–307.

Eagleton-Pierce, M (2011) Advancing a reflexive international relations. Millennium – Journal of International Studies 39(3): 805–823.

Epstein, C (2013) Constructivism or the eternal return of universals in international relations. Why returning to language is vital to prolonging the owl’s flight. European Journal of International Relations 19(3): 499–519.

Feyerabend, P (1975) Against Method. London: Verso Books.

Foucault, M (1972). The Archaelogy of Knowledge. London: Routledge.

García-Andrade, A, Gunnarsson, L, Jónasdóttir, AG (2018) Feminism and the Power of Love: Interdisciplinary Interventions. London; New York: Routledge.

Gee, JP (2011) How to Do Discourse Analysis: A Toolkit. London: Routledge.

Giddens, A (1987) Social Theory and Modern Sociology. Cambridge, Mass: Policy Press.

Goffman, E (1957) Characteristics of total institutions. In: Symposium on Preventive and Social Psychiatry. Washington DC: The National Academies Press.

Gregoratto, F (2018) The ambiguity of love: beauvoir, Honneth and Arendt on the relation between recognition, power and violence. Critical Horizons 19(1): 18–34.

Habermas, J (1971) Technology and science as ideology. In: Toward a Rational Society. London: Heinemann.

Hamati-Ataya, I (2012). IR theory as international practice/agency: a clinical-cynical Bourdieusian perspective. Millennium: Journal of International Studies 40(3): 625–646.

Hamati-Ataya, I (2013a) Reflectivity, reflexivity, reflexivism: IR’s ‘reflexive turn’ – and beyond. European Journal of International Relations 19(4): 669–694.

Hamati-Ataya, I (2013b) Transcending objectivism, subjectivism, and the knowledge in-between: the subject in/of ‘Strong reflexivity’. Review of International Studies 40(01): 153–175.

Hansen, L (2006) Security as Practice: Discourse Analysis and the Bosnian War. London: Routledge.

Hansen, L (2015) Ontologies, epistemologies, methodologies. In: Laura S (ed.), Gender Matters in Global Politics: A Feminist Introduction to International Relations. London: Routledge, 14–25.

Hartnett, L (2018) Love in a Time of Empire: An Engagement with the Political Thought of Tolstoy, Tagore and Camus. PhD Thesis, London School of Economics, UK.

Harvey, L (2020) Discourse analysis. Social Research Glossary. Quality Research International. Available at: https://www.qualityresearchinternational.com/socialresearch/discourseanalysis.htm (accessed 5 November 2020).

Hobson, JM (2007) Reconstructing international relations through world history: oriental globalisation and the global dialogic conception of inter-civilisational relations. International Politics 44(4): 414–430.

Hobson, JM (2014) Re-viewing the eurocentric conception of world politics. Millennium – Journal of International Studies 42(3): 485–514.

Holden, G (2002) Who contextualizes the contextualizers? Disciplinary history and the discourse about IR discourse. Review of International Studies 28(02): 253–270.

Jackson, PT (2011) The Conduct of Inquiry in International Relations: Philosophy of Science and Its Implications for the Study of World Politics. New York: Routledge.

Jolivette, A (2015) Research justice: radical love as a strategy for social transformation. In: Jolivette, A (ed.), Research Justice – Methodologies for Social Change. Bristol: Bristol University Press.

Kapoor, I (2017) Cold critique, faint passion, bleak future: post-development’s surrender to global capitalism. Third World Quarterly 38(12): 2664–2683.

Keohane, R (1988). International institutions: two approaches. International Studies Quaterly 32(4): 379–396.

Knutsen, TL (2014) Western approaches. Millennium – Journal of International Studies 42(2): 448–455.

La Boétie, E de (1549) Le Discours de La Servitude Volontaire. Université du Québec à Chicoutimi: Les classiques des Sciences Sociales. Available at: http://classiques.uqac.ca/classiques/la_boetie_etienne_de/discours_de_la_servitude/discours_servitude.html (accessed 7 October 2020).

Laclau, E, Mouffe, C (2014) Hegemony and Socialist Strategy: Towards a Radical Democratic Politics. London: Verso.

Lapid, Y (1989) The third debate: on the prospects of international theory in a post-positivist era. International Studies Quarterly 33(3): 235–254.

Latour, B (2004) Why has critique run out of steam? From matters of fact to matters of concern. Critical Inquiry 30(2): 225–248.

Leander, A (2002) Do we really need reflexivity in IPE? Bourdieu’s two reasons for answering affirmatively. Review of International Political Economy 9(4): 601–609.

Levi-Strauss, C (1952) Race and History. Paris: UNESCO.

Lévi-Strauss, C (1996) Anthropologie Structurale Deux. Paris: Plon.

Lynch, C (2008) Reflexivity in research on civil society: constructivist perspectives. International Studies Review 10(4): 708–721.

Martin, LH, Huck, G, Hutton, PH (eds) (1988) Technologies of the Self: A Seminar with Michel Foucault. Amherst: University of Massachussets Press.

Mikkola, M (2017) On the apparent antagonism between feminist and mainstream metaphysics. Philosophical Studies 174(10): 2435–2448.

Neal, A (2019) ‘Neophilia’ in Horizon Scan: critical security studies for the next 50 years. Security Dialogue 50(4): 9–37.

Negri, M, Hardt, A (2005) Multitude: War and Democracy in the Age of Empire. London and New York: Penguin.

Nero, S (2015) Language, identity, and insider/outsider positionality in Caribbean Creole English research. Applied Linguistics Review 6(3): 341–368.

Penttinen, E (2013) Joy And International Relations: A New Methodology. London & New York: Routledge.

Peterson, VS (1992) Gendered States: Feminist (Re)Visions of International Relations Theory. Boulder: Lynne Rienner Publishers.

Phull, K, Ciflikli, G, Meibauer, G (2019) Gender and bias in the international relations curriculum: insights from reading lists. European Journal of International Relations 25(2): 383–407.

Sjolander, CT, Cox, WS (1994) Critical reflections on international relations. In: Sjolander CT and Cox WS (eds), Beyond Positivism: Critical Reflections on International Relations. Boulder: Lynne Rienner Publishers.

Soedirgo, J, Glas, A (2020) Towards active reflexivity: positionality and practice in the production of knowledge. PS: Political Science & Politics 53(3): 527–531.

Spash, CL (2012) New foundations for ecological economics. Ecological Economics 77(C): 36–47.

Tickner, AB (2013) Core, periphery and (neo)imperialist international relations. European Journal of International Relations 19(3): 627–646.

Tuhiwai Smith, L (1999) Decolonizing Methodologies: Research and Indigenous Peoples. London and New York: Zed Books.

Van der Ree, G (2014) Saving the discipline: plurality, social capital, and the sociology of IR theorizing. International Political Sociology 8(2): 218–233.

Vasilaki, R (2012) Provincialising IR? Deadlocks and prospects in post-western IR theory. Millennium – Journal of International Studies 41(1): 3–22.

Vinck, D (2007) Sciences et Société – Sociologie Du Travail Scientifique. Paris: Armand Colin.

Author biography

Audrey Alejandro is Assistant Professor in Qualitative Text Analysis at the London School of Economics and Political Science, working at the crossroads of International Relations, international political sociology and political science. Her work focuses on discourse and knowledge, the role they play in society and world politics, how scholars can apprehend their effects and what it takes to produce discourses and knowledge that helps to create a world consistent with our socio-political objectives and values. She is author of Western Dominance in International Relations? The Internationalisation of IR in Brazil and India (Routledge, 2018), co-convenor of the network ‘DoingIPS’, chair of the British International Studies Association working group ‘IR as a social science’, and co-editor of the book series Trends in European IR Theory (Palgrave).

Contact

Audrey Alejandro, London School of Economics and Political Science, Houghton Street, London WC2A 2AE, UK. Email: [email protected]

Citation

Alejandro A. Reflexive discourse analysis: A methodology for the practice of reflexivity. European Journal of International Relations. 2021;27(1):150-174. doi:10.1177/1354066120969789

Copyright License

• Title: “Reflexive discourse analysis: A methodology for the practice of reflexivity”
• Author: “Audrey Alejandro”
• Source: “https://journals.sagepub.com/”
• License: “CC BY-NC 4.0”