The methodological diversity of OSS

Studying ontological (in)security, methodological questions matter. In this section, I provide an overview of the diversity of scientific methodologies in OSS. I thereby draw on the taxonomy of social scientific methodologies proposed by Jackson’s The Conduct of Inquiry in International Relations (see Table 1).Footnote ³² This taxonomy is based on the distinction between two fundamental wagers in philosophical ontology, neither of which can be definitely settled: the relationship between the knower and the known, and the relationship between knowledge and observation.Footnote ³³

Table 1. Four scientific methodologies, taken from Jackson (Reference Jackson2011, 37)

The first wager distinguishes the positions of dualism and monism.Footnote ³⁴ Mind-world dualism ‘maintains a separation between researcher and world such that research has to be directed toward properly crossing that gap.’Footnote ³⁵ Valid knowledge must thus ensure the ‘accurate correspondence between empirical and theoretical propositions … and the actual character of a mind-independent world.’Footnote ³⁶ In contrast, mind-world monism represents the idea that, as ‘“world” is endogenous to social practices of knowledge-production … speaking of “the world” as divorced from the activities of making sense of the world is literally nonsensical.’Footnote ³⁷ The second wager addresses the question of whether knowledge is limited to empirically observable phenomena or if it can transcend toward ‘in-principle unobservable objects.’Footnote ³⁸ The former position is labeled phenomenalism, the latter transfactualism.Footnote ³⁹ Taken together, the two wagers delineate four scientific methodologies: neopositivism, critical realism, analyticism, and reflexivity.Footnote ⁴⁰

For two reasons, I draw on Jackson’s taxonomy instead of more conventional differentiations, such as between positivism and non- or post-positivism,Footnote ⁴¹ positivism and interpretivism,Footnote ⁴² or explaining and understanding.Footnote ⁴³ On the one hand, obscuring the differences between contrasting approaches, these distinctions represent a ‘less-than-useful classificatory scheme that does not really clarify the issues at stake in the philosophical distinction.’Footnote ⁴⁴ Categorizing OSS through these conventional lenses thus brushes over fundamental differences in terms of methodology and often reduces them to questions of method.Footnote ⁴⁵ On the other hand, these distinctions also obscure differences within the respective approaches, as they collapse the two axes of fundamental philosophical differences into a single differentiation. As a result, the positivism versus post-positivism distinction as commonly understood primarily refers to the diagonal axis between neopositivism and reflexivity, thereby leaving aside the other two quadrants. However, while critical realism, reflexivity, and analyticism could technically all be subsumed under the label of post-positivism, they have, in fact, very little in common. Distinctions such as positivism versus post-positivism or positivism versus interpretivism are thus too broad to identify and discuss the relevant underlying methodological differences between contributions to OSS and the implications emerging therefrom. As a result, I draw on Jackson’s taxonomy to refocus attention on the philosophical and methodological issues ‘at the heart of the distinction properly understood.’Footnote ⁴⁶

Analyzing methodological implications for the study of ontological (in)security also requires reflecting on the underlying perspective on knowledge production in this paper. First, Jackson’s fundamental wagers represent ‘questions to which particular scholars give answers that depend, in the final analysis, on a measure of faith.’Footnote ⁴⁷ As these methodologies cannot be adjudicated from an external point of view, none of them is better or worse than another.Footnote ⁴⁸ Therefore, in line with Jackson, I subscribe to a commitment to pluralism and epistemic tolerance as research should be evaluated for its ‘internal validity: whether, given our assumptions, our conclusions follow rigorously from the evidence and logical argumentation that we provide.’Footnote ⁴⁹ Therefore, I refuse to judge between different methodologies and do not make a case against any of them.

Second, while every empirical issue can in principle be studied based on every methodology, scientific methodologies influence the questions that can reasonably be asked, the kind of knowledge that can possibly be produced, and the methods required, thus setting ‘the stage for the kinds of empirical and theoretical puzzles and challenges that a scholar takes to be meaningful and important.’Footnote ⁵⁰ Thus, while I do not judge between different methodologies, I outline the implications of OSS based on different methodologies in terms of the knowledge claims that can be made, their metatheoretical implications, and their critical potential.

Third, while methodologies are the result of faith-based ‘wagers,’ it is my belief that these wagers can be deployed in a functional and even opportunistic way. Scientific methodologies are ‘sweaters, not skins.’Footnote ⁵¹ It is thus possible to purposefully choose between these wagers and place different bets at different points in time, and ‘there is no reason why a single published work cannot contain multiple independent arguments … drawn from different methodologies.’Footnote ⁵² In effect, accepting that as researchers we all resemble blind people trying to figure out the shape of the elephant ‘ontological (in)security,’Footnote ⁵³ I regard accepting that there is no single best methodological approach to doing so as the most promising pathway for knowledge production.Footnote ⁵⁴

In the following, I outline the basic parameters of research on ontological (in)security grounded in the different methodologies and provide examples thereof. This requires two caveats. On the one hand, I refer to contributions that I regard as typical examples and do not provide a conclusive or even exhaustive list of contributions to OSS associated with the respective methodology, which would go well beyond the scope of this paper. Instead, the purpose is to illustrate what grounding OSS in a specific methodology looks like. On the other hand, given the lack of explicit considerations about methodological questions, the allocation of contributions to the different methodologies is necessarily based on my exegetic reading of them. Thus, while some scholars might disagree with their categorization, I argue that the contributions display at least characteristics of the different methodologies.

Combining mind-world dualism and phenomenalism, neopositivism seeks to generate generalizable knowledge through the systematic testing and falsification of hypotheses based on the identification of empirical regularities or covariations and their subsumption under covering laws.Footnote ⁵⁵ As neopositivist research reduces causality to covariation, explanations consist of ‘bringing prior observations of empirical regularities together with specific information about a particular situation and demonstrating that the outcome was to be expected.’Footnote ⁵⁶ As valid claims ‘correspond to the (mind-independent) world,’ hypothesis testing is the ‘key neopositivist procedure’ to evaluate whether the conceptions in our minds in fact correspond with the world ‘out there.’Footnote ⁵⁷

Within OSS, contributions grounded in neopositivism seek to establish empirical regularities concerning factors correlating with the occurrence of ontological (in)security, both as causes and effects. One example is Ofek Riemer,Footnote ⁵⁸ who draws on a survey experiment in the context of Israeli foreign policy shifts to test hypotheses concerning the influence of states’ foreign policies on citizens’ sense of ontological (in)security. A similar research design is employed by Nimrod Rosler and Galia Press-Barnathan’sFootnote ⁵⁹ assessment of whether international boycotts of cultural mega events cause ontological insecurity among citizens of the host state. Remaining at the individual level, Tianyang Liu et al.Footnote ⁶⁰ analyze how measures of ontological (in)security in public surveys in different states correlate with threat perceptions of China. Similarly, Sefa Secen investigates covariations between the living conditions of Syrian refugees in Germany and Turkey and different strategies of ontological security seeking, as well as correlations between indicators of physical and ontological insecurity.Footnote ⁶¹

Critical realism, combining the wagers of mind-world dualism and transfactualism, presumes an independently existing world, while also claiming the existence of ‘real but unobservable’ entities, as the ‘mind-independent world is deeper than our perceptual experiences.’Footnote ⁶² This implies a stratified understanding of reality as only partly accessibleFootnote ⁶³: ‘Underneath what we can perceive and detect lies a realm of real-but-undetectable “structures, powers, and tendencies” that give rise to the ordinary empirical sphere.’Footnote ⁶⁴ In effect, our knowledge can go beyond experience and ‘grasp the deeper generative causal properties that give rise to those experiences.’Footnote ⁶⁵ Critical realism thus differs from neopositivism’s focus on observable covariation and instead theorizes the existence of real but undetectable causal mechanisms, which have important effects in the perceivable part of the world.Footnote ⁶⁶ Explanations thus leave behind deductive hypothesis testing and instead focus on abduction as the generation of plausible conjectures concerning underlying principles and factors from the observed phenomena.Footnote ⁶⁷

Within OSS, several contributions display characteristics of a critical realist methodology. On the one hand, researchers abducted the unobservable causal mechanism of ontological security seeking as the best possible explanation behind empirically observable phenomena, such as the existence of materially non-beneficial regional integration efforts in the post-Soviet spaceFootnote ⁶⁸ or the counterproductive pursuit of spheres of influence by self-perceived great powers.Footnote ⁶⁹ Similarly, explaining the proliferation of private security companies in Europe despite increased safety, Elke Krahmann proposes ontological security provision through the identification, profiling, and management of risks as the underlying and unobservable causal mechanisms.Footnote ⁷⁰ On the other hand, research focused on the unobservable processes underlying spatial and temporal dynamics of ontological (in)security. Drawing on the empirical case of securitization of migration in Serbia, Filip Ejdus and Tijana Rečević abduct that emergent processes generated through feedback loops and non-linear horizontal interactions at the micro level represent the conjecture providing the best explanation for the transformation of individual- into collective-level ontological (in)security.Footnote ⁷¹ Further, analyzing ontological insecurity in the Irish border region in the context of Brexit, Ben Rosher argues that ‘memories of the historically securitized and militarized border,’ projected ‘onto and vicariously through the next generation,’ represent the best possible explanation for the observable outcome of ‘intergenerational ontological insecurity.’Footnote ⁷² Similarly, Kathrin Bachleitner abducts temporal security seeking based on collective memories as the best way to ‘specify the precise details of how ontological security may work out as a process in states.’Footnote ⁷³

Reflexivity denotes the combination of the wagers of mind-world monism and transfactualism.Footnote ⁷⁴ As mind-world monism grounds knowledge production in the ‘concrete practical involvements’ of researchers, practices of knowledge production represent the ‘only feasible objects of transfactual knowledge.’Footnote ⁷⁵ As a result, reflexivist research entails reflection and self-awareness toward ‘research practices and the broader context in which they are embedded’ through the application of the ‘tools of knowledge-production’ to the situations and practices of knowledge-production themselves.Footnote ⁷⁶ This entails an understanding of knowledge as ‘both inseparable from and not in any simple sense reducible to the social position and organizational practices of the scientific researcher.’Footnote ⁷⁷ However, the social relations in which the researcher is embedded are always and thoroughly imbued with power relations, which scientific knowledge, always being ‘for someone and for some purpose,’Footnote ⁷⁸ either reproduces or destabilizes.Footnote ⁷⁹ This is not an accidental byproduct but inextricably ‘wrapped up with the very production of knowledge in the first place.’Footnote ⁸⁰ Reflexivity thus inclines researchers toward a ‘systematic effort to analyze their own role as knowledge-producers and to locate themselves with reference to their broader social contexts’ to produce knowledge about these social arrangements.Footnote ⁸¹

Reflexivist contributions, drawing – among others – on postcolonial,Footnote ⁸² gender,Footnote ⁸³ or queerFootnote ⁸⁴ perspectives, are prominent within OSS. On the one hand, these contributions employ ontological (in)security as a critical tool to ‘excavate… the presuppositions that implicitly inform the designation of certain kinds of worldly claims as knowledge.’Footnote ⁸⁵ For example, they analyze how power structures and systems of oppression, such as neoliberal globalization,Footnote ⁸⁶ the Western-dominated international order,Footnote ⁸⁷ or power-infused ‘gendered spaces,’Footnote ⁸⁸ produce ontological insecurity in the subjects they marginalize. On the other hand, contributions unveiled how ‘previously thought to be unbiased’ conceptions of ontological (in)security ‘have tacit biases,’Footnote ⁸⁹ thus highlighting the marginalizing effects of dominant actors’ ontological security seekingFootnote ⁹⁰ as it ‘disciplines or marginalizes modes of subjectivity which resist the closure of ontological security-seeking strategies.’Footnote ⁹¹ This is evident, for example, in the inherent exclusionary effects of states’ autobiographical narratives, which are gendered masculine and constituted upon the stigmatization of alternative narratives,Footnote ⁹² as well as in dominant narratives of global governance and transnationalism that maintain ontological security through the othering and securitization of subaltern actors.Footnote ⁹³

Closely related to pragmatist perspectives in IR,Footnote ⁹⁴ analyticism places its wagers on mind-world monism and phenomenalism. The mind-world monist rejection of Cartesian dualism locates knowledge neither in the mind nor the world but in the practices which mediate our access to the world.Footnote ⁹⁵ The phenomenalist wager entails that knowledge cannot transcend experience but is always grounded in the concrete practices of knowledge production of communities of researchers.Footnote ⁹⁶ This perspective is reflected in pragmatist perspectives embracing non-foundational epistemological and ontological positions.Footnote ⁹⁷ While, from this perspective, scientific knowledge is always practical knowledge,Footnote ⁹⁸ the intersubjective character of these experiences distinguishes the analyticist understanding of knowledge from notions of arbitrary subjectivity or crude idealism.Footnote ⁹⁹ As a result, knowledge is a ‘social product,’Footnote ¹⁰⁰ and a discipline such as IR represents a scholarly community of practice characterized by a ‘web of justified beliefs’Footnote ¹⁰¹ based on specific rules determining the (always temporal) acceptability of arguments.Footnote ¹⁰²

If knowledge as a social product is located not in the world but in experiences organized through social practices, it should not be evaluated by its correspondence with the world but by the contribution that it makes, as experiences can be organized in more or less useful ways.Footnote ¹⁰³ In this regard, scientific knowledge is generated by ‘systematically organizing experiences so as to generate useful insights.’Footnote ¹⁰⁴ Analyticism thus relies on the process of ideal typification as a form of ‘disciplined ordering of the facts of experience.’Footnote ¹⁰⁵ As ideal types are neither a representation of reality nor solely reducible to the researcher’s perspective,Footnote ¹⁰⁶ this involves abstracting away from the details of individual experiences to instead create analytical tools through the instrumental oversimplification of value-based commitments into deliberate ‘caricatures’ or ‘conceptual filters.’Footnote ¹⁰⁷ Since the value of an ideal type depends on its usefulness as ‘an appropriate means to the analytical end’ to make sense of concrete situations, ‘it is quite literally nonsensical to speak of an ideal-type as being “valid” or “invalid,”’ and ideal types can be discarded not ‘for being false, but for being useless.’Footnote ¹⁰⁸

An analyticist methodology is evident in a number of contributions to OSS. This is very explicit in Dingott Alkopher’s analysis of reactions to migration in the EU, which ‘builds a typology of three analytical ideal-type socio-psychological lenses’ to understand host societies’ dealing with migration-related uncertainty.Footnote ¹⁰⁹ More implicit analyticist perspectives are already evident in foundational contributions to OSS in IR. For example, while Mitzen argued to regard states ‘“as if” they are ontological security-seekers,’Footnote ¹¹⁰ Steele conducted ‘an ontological security interpretation of British neutrality’ during the United States Civil War.Footnote ¹¹¹ Analyticist characteristics can also be found in more recent contributions to OSS, asking whether the concept of ontological (in)security ‘helps us make sense of current global political trends.’Footnote ¹¹² In this regard, Cingöz et al. ‘employ… the concept of ontological security … with the objective of elucidating [the] character and the underlying motivations’ concerning Iran’s ‘axis of resistance.’Footnote ¹¹³ Focusing on the effects of current developments in the cyber domain, Lupovici suggests ‘that the scholarship of ontological security … provides intriguing analytical tools’ and ‘a useful lens through which to consider the challenges [that] interactions in the cyber domain may create for states.’Footnote ¹¹⁴ Mitzen, again, argues that an ‘ontological security lens helps … in ways that potentially shed light on prospects for continued EU integration.’Footnote ¹¹⁵ Similarly, focusing on nineteenth-century Anglo-Afghan relations, Bayly states that from ‘an analytical perspective, … ontological security can bring to the surface relations between the “buffer state” and the neighbouring power.’Footnote ¹¹⁶ In effect, these contributions proceed from the idea that ‘the assumption that states are motivated by ontological security seeking … illuminates blind spots in existing works.’Footnote ¹¹⁷ However, contributions with more conceptual orientations also displayed characteristics of analyticist methodology. This is, for example, evident in Bolton’s analysis of the role of collective emotions and affect for ontological (in)security, arguing that ‘conceptualizing OS in relation to moral orders interrelated with the sacred provides a new optic through which to account for the dynamism of foreign policy.’Footnote ¹¹⁸ In a similar way, Verleye distinguishes several ideal-typical state shame narratives to create ‘a suitable and broad analytical tool for IR,’ which is able to produce ‘a thoughtful ordering of empirical reality.’Footnote ¹¹⁹

Valid knowledge claims

Scientific methodologies differ concerning their standards for the production and evaluation of valid knowledge claims. While standards for the evaluation of knowledge claims are generally important, this is especially the case for OSS, which reflects Berenskoetter’s call for deep theorizingFootnote ¹²⁰ insofar as it understands anxiety and the search for ontological security as a transhistorical constitutive part of the human condition.Footnote ¹²¹ As Anthony Giddens argues, ‘to be ontologically secure is to possess … answers to fundamental existential questions which all human life in some way addresses,’Footnote ¹²² concerning issues such as existence, finitude, relations, and autobiography.Footnote ¹²³ Similarly, drawing on existentialist philosophy, Felix Berenskoetter argued that as ‘a fundamental feature of this [human] condition is uncertainty[,] … anxiety enters our world once we become self-conscious and reflexive subjects,’ leading human beings to embrace a variety of ‘anxiety-controlling mechanisms.’Footnote ¹²⁴ Also, for Lacanian contributions, ‘anxiety is an ever-present generator of agency.’Footnote ¹²⁵ Given that ‘it seems fair to say that people have always lived with anxieties,’Footnote ¹²⁶ the search for ontological security has been proposed as the constructivist equivalent of utility maximization in rational choice theory.Footnote ¹²⁷

However, if anxiety and ontological security seeking represent constitutive elements of the human condition, every human action or interaction could in one way or another be seen as reflective of them. OSS thus encounters a problem similar to rational choice theory: Regarding actors as rational utility maximizers, whichever choice an actor made must have been rational from this point of view. However, to put it with Wendt, this always only appears in the rearview mirror.Footnote ¹²⁸ In the same way, every action, including its opposite, could retrospectively be explained through the search for ontological security. Thus, while voting for populists represents an ontological security-seeking strategy, the opposite is equally true, as for some people, it is exactly voting against populists that stabilizes their ontological security. OSS thus faces the ‘problem of codetermination:’Footnote ¹²⁹ if ‘everyone is anxious all the time, how do we account for differences in actor behavior?’Footnote ¹³⁰

Drawing on existentialist accounts, several contributions proposed to address this problem by distinguishing between ‘normal feelings of anxiety,’ which are compatible with ontological security, and ontological insecurity as extraordinary neurotic anxiety.Footnote ¹³¹ I agree that the latter can neither be regarded as empirically prevalent nor as constitutive of the human condition in general.Footnote ¹³² I also agree with the critique against the often very loose ascription of ontological insecurity in empirical research.Footnote ¹³³ However, as neurotic anxiety ‘is not what scholars of ontological security have in mind when they speak of ontological insecurity,’Footnote ¹³⁴ and as ‘[n]ormal anxiety … is a necessary part of the human condition,’Footnote ¹³⁵ separating between normal and extraordinary anxiety does not solve the problem of codetermination. Instead, this requires focusing on the evaluation of knowledge claims based on their methodological underpinnings.

Neopositivist methodology moves beyond the general assumption of ontological security seeking as a fundamental human condition to instead test propositions concerning factors which, under specified conditions, contribute to the ontological (in)security of actors with specific characteristics, as well as the effects emerging therefrom. The goal is thus the subsumption of increasingly fine-grained knowledge claims under generalizing laws. This would, for example, entail identifying covariations between factors that might contribute to anxiety, indicators of individual-level ontological insecurity, and support for populist parties. While it is at least questionable whether a concept such as ontological (in)security displays sufficient regularities in terms of its causes and effects to be able to produce knowledge claims generalizable under deductively generated laws, this is an empirical question.Footnote ¹³⁶ However, the thornier problem relates to the issues of precise conceptualization and measurement of ontological (in)security for qualitative or quantitative comparative analyses aimed at identifying systematic covariations. On the one hand, this requires decisions, for example, about whether ontological (in)security is conceptualized as a continuous or binary variable and the underlying thresholds.Footnote ¹³⁷ On the other hand, to make knowledge claims within the neopositivist framework, the indicators used to operationalize and measure ontological (in)security and its effects must be both valid and reliable. Thus, testing the hypothesis that anxious people are more inclined to vote for populists requires indicators that accurately measure individual-level ontological (in)security and not some other concept, and which do so regardless of the researcher conducting the study. Making knowledge claims about ontological (in)security that are valid within a neopositivist methodology is no trivial task.Footnote ¹³⁸

For critical realism, producing valid knowledge claims about ontological (in)security requires establishing that ontological security seeking is not just some constitutive part of the human condition. Instead, the strive for ontological security must be explicitly established as a causal effect resulting from ontological insecurity and as the relevant – and ontologically real – causal force in specific contexts which ‘correspond to their innermost essence.’Footnote ¹³⁹ Knowledge claims about ontological (in)security and its effects thus have to be evaluated in terms of ‘demonstrating that it is impossible to ignore,’ thereby connecting ‘undetectable causal powers with manifest actions and behaviors.’Footnote ¹⁴⁰ This involves the formulation of arguments connecting a priori assumptions grounded in a common-sensical starting point through a series of ‘indispensability claims,’ which then allows applying these ‘causal powers’ to a specific empirical case.Footnote ¹⁴¹ For example, establishing anxiety and ontological security seeking as the real causal mechanism behind populist voting in a specific situation requires determining its role in a set of conditions which, while individually insufficient, in combination are sufficient (but not necessary) to produce this outcome.Footnote ¹⁴² Thus, a critical realist perspective demands the difficult task of ensuring that the theoretical accounts of ontological (in)security ‘accurately represent the causal powers actually possessed by [it] in the real world.’Footnote ¹⁴³

From a reflexivist perspective, knowledge claims are not warranted through their external correspondence but through researchers’ systematic reflection about their role as knowledge producers in a broader social context and the knowledge claims present within it.Footnote ¹⁴⁴ Reflexivist knowledge production thereby entails the twofold challenge of maintaining that its claims neither correspond to the world ‘out there’ nor are mere ‘partisan political interventions.’Footnote ¹⁴⁵ It thus draws on pieces of knowledge and tacit assumptions present within specific social arrangements and relates them to their ‘function in reproducing a particular kind of social order.’Footnote ¹⁴⁶ On the one hand, reflexivist researchers draw on knowledge within the ‘social arrangement’ of OSS to outline the biases inherent therein and the functions they fulfill,Footnote ¹⁴⁷ arguing that ‘OSS has yet to confront its own complicity in broader racialized orders of knowledge.’Footnote ¹⁴⁸ This entails, for example, the move from showing that states construct their sense of Self through biographical narrativesFootnote ¹⁴⁹ toward outlining how these narratives uphold dominant social orders by marginalizing alternative perspectives.Footnote ¹⁵⁰ On the other hand, reflexivist research draws on pieces of everyday knowledge within particular social groups to uncover their function in upholding the existing social order. For example, starting from the observation of ‘leave’ voters’ frustration both before and after the Brexit referendum, and systematizing this into an argument about the perpetually frustrated desire for wholeness helps to uncover how tacit assumptions within a particular social group contribute to the upholding of social order.Footnote ¹⁵¹ Thus, within a reflexivist methodology, the question is not how to evaluate knowledge claims based on the assumption of ontological (in)security as a constitutive human condition, but to reflect on the effects of this assumption as well as to use it in order to expose the effects of pieces of everyday knowledge.

An explicitly analyticist methodology provides a possibility to clarify the epistemological status of knowledge claims about ontological (in)security and to evaluate them. Deploying anxiety and ontological security seeking as analytical ideal types, they provide a ‘conceptual baseline’ against which to understand a concrete empirical case.Footnote ¹⁵² In this regard, ontological (in)security can be regarded as causal if, for example, in a counterfactual consideration, the occurrence of support for populism appears implausible in its absence.Footnote ¹⁵³ However, this also implies that for ideal types to generate valuable knowledge, researchers must ‘acknowledge their limitations.’Footnote ¹⁵⁴ Exactly due to the one-sided over-accentuation and simplification entailed in the process of ideal-typification, ideal types necessarily and intentionally leave out ambiguities, opposing factors, or simply alternative perspectives, and no empirical situation is ever reducible to just one ideal type.Footnote ¹⁵⁵ Thus, approaching empirical events through an ideal-type lens will always produce incomplete accounts.Footnote ¹⁵⁶ In effect, while ideal types provide categories which help to generate and sort empirical observations, ‘it is those moments where the messy complexity of actual concrete experience resists such categorization that an analytical narrative takes off.’Footnote ¹⁵⁷ Approaching an empirical situation through the ideal-type lens of ontological (in)security will most likely reveal aspects of the situation that it cannot make sense of, thus highlighting in which respects an ontological (in)security lens is not useful in the sense of ‘revealing intriguing and useful things about the objects to which it is applied.’Footnote ¹⁵⁸ It could, for example, be the case that in specific situations, the analytical lens of ontological security seeking provides less insight into the populist voting behavior of people who do not appear to be ontologically insecure. Or, put differently, in some instances, the counterfactual consideration that people would still support populists if they were not anxious is also plausible. Enabling researchers to evaluate knowledge claims about ontological (in)security by recognizing their explanatory limits, analyticism provides researchers with a clearer understanding of its explanatory contributions. This encourages researchers to identify alternative and potentially conflicting ideal types to generate further insights into the empirical phenomenon.Footnote ¹⁵⁹ Thus, ideal-typical analytical lenses such as socialization, habit, economic interest, or political culture may have greater analytical purchase in explaining support for populists in a specific empirical context.

Metatheoretical challenges

In this section, I draw on two metatheoretical challenges – at the level of ontology and epistemology – which research on ontological (in)security faces. While these metatheoretical challenges also occur in other strands of research, such as (international) political psychologyFootnote ¹⁶⁰ or research on emotions in IR,Footnote ¹⁶¹ they have at least partly been debated within OSS. On the one hand, methodological considerations provide a new perspective on these debates. On the other hand, the challenges serve as useful focal points to discuss the metatheoretical implications of grounding OSS in different methodologies. While the metatheoretical challenges are inherent to research on the concept of ontological (in)security, at least in the conventional conception based on the psychoanalytical works of Ronald Laing and its sociological adaptation by Giddens, depending on the methodology, both the challenges and the way of addressing them differ significantly.

Research on ontological security faces the ontologicalFootnote ¹⁶² challenge of dealing with the actor’s Self, which a focus on the ‘security of the Self’ risks presupposing and reifying as the starting point of its analysis. This argument moves beyond the much-debated issue of transposing the psychological concept of ontological security to collectives such as states.Footnote ¹⁶³ Instead, more fundamentally, the Self is inherent in the concept of ontological (in)security. As ‘the self … occupies a critically important terrain of analysis for OSS,’Footnote ¹⁶⁴ this is the case regardless of the level of analysis. The ontological challenge is thus not that OSS treats states as personsFootnote ¹⁶⁵ – which would not be a problem specific to OSSFootnote ¹⁶⁶ – but that it risks to essentialize the Self of whichever actor it analyzes, regardless of the level of analysis, be it individuals,Footnote ¹⁶⁷ societal groups,Footnote ¹⁶⁸ states,Footnote ¹⁶⁹ state–society complexes,Footnote ¹⁷⁰ international organizations,Footnote ¹⁷¹ or the international community.Footnote ¹⁷²

Within the conventional conception of ontological (in)security, researchers have pursued two main strategies to address this challenge. On the one hand, contributions sought to unpack the actor’s Self by moving beyond the presupposition of any specific substantive content, to instead point toward the openness of its construction.Footnote ¹⁷³ Thereby, however, the Self remains the starting point and center of the analysis. On the other hand, scholars deconstructed the Self as the result of processes of ontological security seeking by other actors, usually at a ‘lower’ level of analysis. For example, understanding states’ identities as the result of multiple conflicting identifications and narrations attached to them by societal actors,Footnote ¹⁷⁴ the analysis shifts toward the ontological security seeking of the different societal actors participating in this discursive process. However, moving beyond the state’s Self as the subject of ontological security to instead take as given the Self of these societal actors leads to a regress. The state’s Self is thus the result of ontological security seeking by societal groups, the societal groups’ Self the result of ontological security seeking by subgroups, and so forth, until reaching the level of the individual. At this point, the Self is either deconstructed from a Lacanian perspective (see below)Footnote ¹⁷⁵ or finally essentialized. This, however, does not remove the ontological challenge as, ultimately ‘“the individual” is just as unobservable as the state is.’Footnote ¹⁷⁶

Additionally, OSS also faces an epistemological challenge consisting of three aspects. First, ontological (in)security is not just an attribute of an actor but a very fundamental experience.Footnote ¹⁷⁷ While identities and perceptions of security can be discursively constructed, an actor either has a stable sense of self, or they do not have it; either an actor experiences anxiety or it does not. Ontological (in)security, as the (absence of) a ‘stable sense of self,’Footnote ¹⁷⁸ thus refers directly to an actor’s ‘real’ condition – not real in physical or material terms but in an ontological sense. OSS thus faces the challenge that it is inclined toward assuming that there is something ‘real’ behind the discourse.

Second, this is especially the case when deploying the concept of ontological (in)security in arguments about their (constitutive) causes and effects. In order to make arguments about the effects of social processes on an actor’s sense of ontological (in)security, or, conversely, the effects of ontological security seeking on the actor’s observable actions,Footnote ¹⁷⁹ OSS thus entails a disposition toward assuming that actors are ontologically (in)secure or are driven by ontological security seeking. Claiming that people’s support for populist parties represents an attempt of ontological security seeking in response to anxieties induced through processes of globalization, this argument would not make sense if people were not really anxious due to globalization.

Third, however, this requires an epistemological leap of faith. This is most evident when focusing on ontological (in)security as a subjective conditionFootnote ¹⁸⁰: ‘Since we cannot know what an actor actually experiences at any given moment, we have to infer their feelings and experiences from their behavior.’Footnote ¹⁸¹ However, there always remains ‘a large and significant epistemic gap between the observation of behavior and the conclusion that this behavior is the … motivated action … of an individual person.’Footnote ¹⁸² It is, at an epistemological level, quite impossible to know whether a person was really ontologically insecure and whether that was the reason behind their populist voting decision. However, the challenge also exists for strands of OSS that regard ontological (in)security as an intersubjective phenomenon in the form of collective emotions,Footnote ¹⁸³ ‘public moods,’Footnote ¹⁸⁴ or affective environments,Footnote ¹⁸⁵ as it would require contrasting actors’ intersubjective reality to the researcher’s ‘objective’ reality. In other words, the epistemological basis for the statement that ‘we are in an age of anxiety’ is not at all clear.Footnote ¹⁸⁶ As a result, OSS faces the challenge that the epistemological impossibility of differentiating between (inter)subjective constructions of ontological (in)security and both individual and collective actors’ real condition implies that the former is understood as the latter.Footnote ¹⁸⁷

OSS grounded in different methodologies perceives of and deals with these metatheoretical challenges in different ways. Neopositivism, assuming that ‘we can never directly perceive a causal effect, and are limited to inferring it from our observations,’Footnote ¹⁸⁸ enables OSS to ignore the metatheoretical challenges, as there is no reliable way to evaluate causation as such.Footnote ¹⁸⁹ In effect, regarding causal relationships as ‘added onto the experience by the mind,’Footnote ¹⁹⁰ neopositivism focuses solely on empirical covariations between indicators of anxiety and indicators of factors theorized as causes and effects thereof. It would thus, for example, be sufficient to test systematic covariations between hypothesized causes of anxiety, such as socioeconomic decline, psychological indicators of anxiety, and people’s populist voting choices. Methodologically, there is thus neither the possibility nor the requirement to go beyond this surface-level covariation to instead make assumptions about ontological realness or real causal effects. As a result, neopositivist OSS avoids the metatheoretical challenges as it can remain agnostic about the existence of actors’ Selves and whether ontological (in)security constitutes a real condition.

Critical realism enables scholars to tackle the metatheoretical challenges inherent in the concept of ontological (in)security head-on. Generating explanations based on the unobservable mechanism of ontological security seeking to account for observable phenomena, for this conjecture to exercise empirical effects, ‘it must be taken to be real, to actually exist, and thus to be something other than an instrumental theoretical convenience.’Footnote ¹⁹¹ Thus, if ontological (in)security provides ‘the best explanation for various puzzling things that we can see,’Footnote ¹⁹² it must be regarded as real insofar as it produces real effects.Footnote ¹⁹³ In this regard, if anxiety and ontological security seeking can be conjectured as the best explanation for the populist voting behavior of individuals or societal groups, understanding as ontologically real both these actors’ Selves and their anxiety and ontological security seeking is not just possible but in fact necessary. OSS grounded in critical realism thus explicitly addresses the metatheoretical challenges.

Reflexivist OSS can ignore the metatheoretical challenges as its mind-world monist position renders any claims about actors’ Selves and their condition of ontological (in)security as ‘really’ existing in the world ‘out there’ nonsensical. More specifically, reflexivist contributions to OSS address the ontological challenge by drawing on non-essentialist notions of the Self and criticizing the biases inherent in its essentialization. For example, outlining the racialized colonial ordering functions and power dynamics inherent in the notion of the Self, Thomas Gehring proposed an ‘autopoetic’ conception of the Self and humanness as a praxis.Footnote ¹⁹⁴ Reflexivist research on ontological (in)security has also drawn on the critical potential of Lacanian non-essentialist notions of the Self. These understand subjects as constituted ‘around a void’ and engaged in perpetually frustrated attempts to restitute their ‘primordial loss,’ thus rejecting the very ‘idea of a single, coherent, and unified subject.’Footnote ¹⁹⁵ Lacanian accounts start from the assumption that subjects do not have pre-formed identities but are ‘impelled to engage in acts of identification, identifying with various socially inscribed signifiers’ that help them to generate a social identity in the form of a place in the social order and recognition through others.Footnote ¹⁹⁶ However, as these signifiers are ‘embedded in preexisting social understandings, discourses, and practices of what being those things means,’ they always remain foreign to the subject.Footnote ¹⁹⁷ As a result, being thrown into a world of alien structures of meaning, the ‘subject can never fully possess or “own” the signifiers it identifies with,’Footnote ¹⁹⁸ thus rendering them perpetually insufficient to ground the subject’s Self-identity.Footnote ¹⁹⁹ Subjects, therefore, subscribe to ‘fantasmatic narratives’ which display their desires for fulfilment and transform them into an empirical lack. This lack is both directed toward specific objects of desire and caused by them through their central position in the symbolic framework.Footnote ²⁰⁰ While these objects promise closure and fulfillment, ‘no empirical “object” can ultimately resolve the ontological lack and make us complete again’ rendering subjects in a continuous process of becoming.Footnote ²⁰¹

On the one hand, reflexivist research on ontological (in)security deployed Lacanian perspectives concerning subjects’ strive for closure and completeness of the Self to unveil exclusivist and oppressive practices, including the appeal of populist discourses,Footnote ²⁰² the racialized othering of migrants,Footnote ²⁰³ political homophobia,Footnote ²⁰⁴ or the persisting power dynamics emerging from shared desires between states and specific societal groups.Footnote ²⁰⁵ On the other hand, reflexivist OSS used Lacanian conceptions of the Self to criticize the marginalizing effect of power structures depriving subjects of complete and unified Selves. This is the case, for example, for post-colonial states, which in an international order characterized by post-colonial hierarchies can never achieve full recognition of the Self on their own terms.Footnote ²⁰⁶ In effect, reflexivist contributions to OSS address the ontological challenge by highlighting the biases in essentialist notions of the Self and the power dynamics in subjects’ strive for completeness, thus excluding conceptions of an ‘essentialized or pre-social self.’Footnote ²⁰⁷

This also holds implications concerning the epistemological challenge. As Lacanian arguments make ‘an ontological claim about the “real” nature of the individual self’Footnote ²⁰⁸ as being constituted by lack and driven by desires embedded in fantasies of wholeness, this would again require an epistemological leap of faith. However, reflexivist contributions to OSS are not concerned with the correspondence of knowledge to anything external and instead reflect upon the social-ordering effects of practices of knowledge production within OSS and specific social groups.Footnote ²⁰⁹ For example, focusing on anxiety in the EU, Thomas Diez and Franziskus von Lucke emphasize that their interest ‘is not about the degree to which the EU as such, or individuals in the EU, feel or are ontologically secure or insecure’ but about ‘how actors … articulate … and thus discursively construct the EU’s being and becoming.’Footnote ²¹⁰ Additionally, reflexivist contributions to OSS outline, for example, how populist discourses always produce the anxiety that they then promise to alleviateFootnote ²¹¹ or how populist politics are motivated and sustained by desire as ‘our attachment to socially available “objects” pushes us to engage in practices through which social orders are reproduced.’Footnote ²¹² This focus on the effects of knowledge production enables reflexivist research on ontological (in)security to leave the epistemological challenge aside.

An analyticist methodology enables researchers to circumvent the metatheoretical challenges. On the one hand, analyticism evades the ontological challenge of the actors’ self. From a phenomenological perspective, OSS can bracket the question of whether a specific actor has or seeks to construct a self – and even if it represents an actor in the first place – to instead regard the entire empirical situation under an ‘as if’ assumption.Footnote ²¹³ For example, regarding a societal group as an actor seeking to stabilize its Self by seeking to inscribe a certain narrative onto the state can generate useful insights into the group’s inclination to support populist politicians. Thereby, however, the ontological status of the actor’s self is irrelevant, and seeking to make statements about it is neither necessary nor useful. Instead, reifying or essentializing the actor’s Self as anything more than an analytical concept is not necessary for it to serve its explanatory function.Footnote ²¹⁴ On the other hand, analyticism’s mind-world monist perspective allows moving around the epistemological challenge of OSS as scholars remain deliberately agnostic about whether ontological (in)security represents an actor’s real condition with real effects. The assumption that globalization causes anxieties which lead people to vote for populists promising to alleviate them provides valuable insights into the current populist ‘wave.’ However, from an analyticist perspective, the epistemological can of worms of the realness of these connections may remain closed since ‘as if’ assumptions suffice to produce these knowledge claims.Footnote ²¹⁵ As a result, faced with metatheoretical challenges inherent in the concept of ontological (in)security, ‘[i]nstead of wading into a dense philosophical thicket and trying to fight their way through it,’ analyticism enables researchers to ‘simply go around … a set of non-issues.’Footnote ²¹⁶ This, however, requires that analyticist research must be self-conscious of and explicit about its methodological positioning.

Critical potential

The critical potential of research on ontological (in)security depends on the scientific methodologies on which it is grounded. For contributions to OSS based on neopositivism and critical realism, the critical potential is limited due to their mind-world dualist grounding of knowledge claims in the real world, either in the form of a mirror image based on covariation or the uncovering of its deep and unobservable properties. Analyzing, for example, the correlation between socioeconomic decline and populist voting behavior, or the underlying mechanisms through which anxieties are caused and (potentially) alleviated in this process, both methodologies take the external world as a given. As a result, their critical potential remains limited to evaluating the results of the research. However, the normative argument that increasing socioeconomic inequality should be addressed to tackle support for populists is neither part of the analysis as such nor does it in any way necessarily follow from it.Footnote ²¹⁷

While the critical potential of both analyticism and reflexivity is much greater due to their mind-world monist position, it differs significantly.Footnote ²¹⁸ For analyticists, ideal-typified value commitments cannot be evaluated through empirical analysis, and disputes about them cannot be scientifically solved. In contrast, for reflexivists, the fact that the methods of knowledge production are not turned against these value commitments to make ‘themselves subject to further scientific critique’ is insufficient.Footnote ²¹⁹ For example, even though from a reflexivist perspective ideal types could be fruitfully incorporated into critical discourse analysis, this would also require analyzing the researcher’s discursive embeddedness from which ideal types emerge.Footnote ²²⁰ As ‘knowing the world and changing the world are inseparable,’ the whole purpose of this endeavor of eliminating biases in knowledge production is ‘that a clearer view of things can be achieved,’ which may then open up possibilities for the non-arbitrary transformation of social conditions.Footnote ²²¹ Uncovering, for example, that support for populist politics is connected to desires for closure and wholeness may provide the basis for inducing transformative change.Footnote ²²²

Analyticist OSS thus necessarily falls short of the reflexivist critical potential. However, this does not require forsaking all critical ambition. In IR, the relationship between pragmatist-analyticist perspectives and critical approaches has been somewhat diffuse: Pragmatism has been characterized as a strand of post-positivism,Footnote ²²³ as an alternative moving beyond both positivism and post-positivism,Footnote ²²⁴ as closely connected to critical theory,Footnote ²²⁵ and as the more relevant point of critique against theory-centered mainstream IR than post-positivism.Footnote ²²⁶ While this conundrum is somewhat representative of the ‘less-than-useful’ categorizations outlined above,Footnote ²²⁷ there are at least two ways in which critical potential can be integrated into analyticist OSS.

First, the value commitments underlying an ideal type do not have to be normatively desirable. Instead, ‘it is quite possible to acknowledge the scientific value of an ideal type, the value-laden basis of which one quite vehemently rejects’Footnote ²²⁸ and thereby to highlight and criticize the fact that a particular empirical situation appears to function according to the logic of this (negative) normative value. For example, showing that ideal-typical notions of people striving for homogeneity, closure, and coherence elucidate popular anti-immigratory sentiments does not mean to endorse them. Instead, as the outcome appears to be implausible in their absence, analyticist research contributes to critique by showing the very prevalence of (normatively undesirable) value commitments.

Second, analyticism entails an inclination toward reflection. Critics might argue that analyticist research inevitably reifies both the objects of study and the researcher’s (biased) value commitments deployed to them. In terms of the former, I argue that analyticism implies an agnostic position, enabling researchers to explicitly bracket questions concerning ontological realness. If made explicit and reflected upon concerning potential ‘lapses’ of ontological reification, this position differs from implicit (neopositivist) or explicit (critical realist) reification processes. Concerning value commitments, reflection is an inherent part of the process of ideal-typification, requiring the researcher to be explicit about the ethical claims transformed into ideal types. Outlining the biases, tacit assumptions, and silences inherent in the deployed ideal types demands the researcher ‘to be reflexive about her own categories used and the theories from which she thinks and acts,’Footnote ²²⁹ as well as their embeddedness in the social and discursive context.Footnote ²³⁰ This entails, for example, reflecting on the implications of analyzing populism primarily through the ideal type lens that people are generally anxious and driven by anxiety minimization, instead of the ideal typified value commitment of radical agency based on the embrace of anxiety.Footnote ²³¹ In fact, some forms of (radical) support for populist politics, such as the January 6 attacks on the United States’ Capitol, might be better understood as attempts to claim radical agency through the embrace of uncertainty. These reflections about the ideal type value commitments deployed to empirical situations thus have the potential to elucidate biases in our analyses.

Moving beyond the potential for critique, analyticist research on ontological (in)security differs from reflexivist contributions in that it constitutes the methodological basis for a pragmatist-oriented normative turn in OSS, as proposed by Jason Ralph.Footnote ²³² While there have been extensive debates about the conservative status quo bias inherent in understandings of ontological security as generated through closed identities and fixed routines,Footnote ²³³ according to Ralph, this is precisely the reason why the call to ‘embrace anxiety’ is not very appealing for many people.Footnote ²³⁴ Instead, drawing on pragmatist ethics, Ralph proposes a focus on improving people’s lived experiences in constantly changing environments. This perspective regards normative value not as substantive but as processual, located in the practice of learning and growth,Footnote ²³⁵ which may contribute to overcoming violent and exclusivist orders without abolishing all forms of order as such.Footnote ²³⁶ This pragmatist ethical perspective is constituted on an analyticist methodology in two ways.

On the one hand, focusing on the ability of agents to assert control over their experience of being entails that these questions can only be answered in specific contexts. Whether an actor’s situation-specific experience of being would improve through the continuation or adjustment of routines and identities cannot be answered in the abstract.Footnote ²³⁷ This resonates with analyticism’s phenomenological orientation, which generates knowledge about specific situations and the role that value commitments, such as inclinations for stability and openness for change, play in them, while recognizing that these insights do not have transfactual value. Identifying the role that ideal typical (non-)exclusionary value commitments play in a specific situation and the potential ways to improve people’s situation-specific experience of being without recurring, for example, to populist politics, analyticist OSS can thus contribute to learning and growth.

On the other hand, Ralph argues that the question concerning the possibility of nurturing healthy growth depends on a certain ‘faith’ in the functional value of thoughts:

Thus, thoughts do not necessarily have to be true to be of practical value, e.g., by helping actors to overcome anxiety in a specific context.Footnote ²³⁹ This is reflective of the analyticist mind-world monist position, which recognizes the functional value of ideal-type ethical claims in producing insights about specific situations while rejecting both to securely ground these insights ‘within the world’ and to regard them as ‘true’ beyond their functional value. Grounding functional knowledge claims in value commitments which do not have to correspond to the world ‘out there,’ analyticist OSS provides the basis for learning processes whereby people’s situation-specific experience of being can be improved by focusing on the functional value of ethical commitments, for example, in overcoming exclusionary populist politics through processes of altercasting.Footnote ²⁴⁰