Abstract
Research on ontological security in International Relations (IR) has grown significantly in recent years. However, this scholarship is marked by conceptual ambiguity concerning the meaning of and relationship between the key concepts of ontological insecurity and anxiety. In addition, ontological security scholarship has been criticized for applying a concept that was originally developed for understanding individuals to states, and for being excessively concerned with continuity while largely ignoring change or seeing it as a negative force to be avoided. Despite such issues, however, reflection on the theoretical origins of ontological security remains limited. Based on such reflection, the present article argues that these issues can be circumvented if we return to one of the theoretical precursors of ontological security studies, the existentialist literature on anxiety. R.D. Laing, who coined the term ontological security, was strongly influenced by the existentialist anxiety theorists. Anthony Giddens, however, who drew on Laing and whose understanding of ontological security permeates IR scholarship, explicitly rejected the distinction between normal and neurotic anxiety, which was central to the work of existentialists like Rollo May. This article reintroduces this distinction. Doing so is useful, the article argues, both for providing conceptual clarity and for moving beyond the criticisms of ontological security mentioned above. More generally, the article suggests that ontological security studies has much to gain from drawing on the insights of the existentialist literature on anxiety to a greater extent than has hitherto been the case.
Introduction
Coined by psychologist R.D. Laing (1990 [1960]), the concept of ontological security entered International Relations (IR) more than two decades ago, primarily through the work of Anthony Giddens (1991) (Huysmans, 1998; McSweeney, 1999). Since then, ontological security scholarship in IR has grown exponentially into a diverse literature that has produced novel insights into a number of important issues in IR. However, this scholarship is also marked by conceptual ambiguity concerning the meaning of and relation between key concepts such as anxiety and ontological insecurity. In addition, ontological security scholarship has been subjected to a number of criticisms, two of which this article addresses. First, while the concept of ontological security was originally developed for understanding individuals, sophisticated arguments have been presented for applying it to states (Mitzen, 2006; Steele, 2008). Such arguments notwithstanding, some still criticize its application at the state level for treating states as people with psyches and emotions (Lebow, 2016)—a move seen as taking states as monoliths and reifying them (Croft, 2012; Krolikowski, 2008; cf. Delehanty and Steele, 2009; Steele, 2017). Second, ontological security scholarship in IR has been criticized for being excessively concerned with continuity and focusing mostly on how states or other agents seek to protect their identity, while largely ignoring change, or seeing it as a negative force to be avoided (e.g. Browning and Joenniemi, 2017).
Despite the existence of these issues, reflection on the theoretical origins of ontological security remains limited. Engaging in such reflection, this article argues that the issues described above can be circumvented and the research program advanced if we return to the existentialist psychological literature on anxiety, a key theoretical precursor of ontological security. Some existing IR research has already begun to integrate theoretical insights from work on anxiety—primarily that of Kierkegaard and Tillich—into ontological security scholarship (Browning, 2018b; Ejdus, 2020; Gustafsson, 2016; Rumelili, 2015a). However, such scholarship has only selectively incorporated insights from parts of this rich literature and has not reflected on or addressed the fundamental conceptual issues and criticisms mentioned above. The importance of the anxiety literature for the concept of ontological security and its usefulness for developing ontological security studies (OSS), we argue, has not been fully recognized. While most IR scholarship has primarily drawn on Giddens, this article demonstrates that the theoretical pedigree of OSS stretches back further than that, even beyond Laing. When Laing developed the concept of ontological security, he was influenced by the existentialist psychologists, and the distinction between normal and neurotic anxiety found in the work of existentialist anxiety theorists seems to have informed his understanding of ontological security. However, in his adaptation of ontological security, Giddens explicitly rejected this distinction. This article seeks to recover some of the insights from the existentialists, in particular the important distinction between neurotic and normal anxiety, and to integrate it into ontological security research.
Drawing on scholars like Rollo May (1977), this article argues that the distinction between normal and neurotic anxiety is useful for providing greater conceptual clarity. While much scholarship conflates ontological insecurity and anxiety in general, the article suggests that ontological insecurity and neurotic anxiety are highly pathological and debilitating conditions that can be seen as more or less synonymous. Normal anxiety, on the other hand, is not necessarily debilitating and experienced by all of us to some degree. This contradicts Giddens’ assertion that all anxiety is necessarily neurotic. Further, normal anxiety, as understood by May and others, is not just an individual-level concept and May himself, along with thinkers like Paul Tillich, also applied it to societies. This suggests that applying it to collective actors like states is not a great leap and therefore less problematic than critics assert. In addition, by incorporating the distinction between neurotic and normal anxiety, it becomes possible for ontological security scholarship to move beyond its continuity bias, as normal anxiety is linked to creativity and change.
The next section reviews the IR literature on ontological security and some of the criticisms it has been subjected to. The section that then follows returns to Laing’s original theorization of the concept as well as Giddens’ adaptation and links it to conceptual issues in OSS in IR. In the penultimate section, we discuss how the anxiety literature can be used to provide greater conceptual clarity, help strengthen the case for applying a theory developed for understanding individuals to states, and address the “continuity bias.” The conclusion summarizes the article’s findings and discusses potential avenues for future research.
Ontological security in International Relations
In IR, the concept of ontological security typically refers to “security not of the body but of the self, the subjective sense of who one is, which enables and motivates action and choice” (Mitzen, 2006: 344). It thus denotes “security as being” rather than “security as survival” (Giddens, 1991) and derives its explanatory power from its opposition to the kind of physical security IR has traditionally been concerned with. In doing so, the study of ontological security seeks to question the primacy of physical security in IR—a critical edge that has arguably contributed to the concept’s wide appeal.
This common starting point notwithstanding, IR research on ontological security is characterized by increasing diversity with constructivist (Berenskoetter, 2014; Berenskoetter and Giegerich, 2010; Flockhart, 2016), post-structuralist (Browning, 2019; Eberle, 2019; Kinnvall, 2004, 2018), and post-colonial (Agius, 2017; Shani, 2017; Untalan, 2020; Vieira, 2018) approaches being developed. Recently, this internal heterogeneity seems to have prompted a terminological shift away from Ontological Security Theory (OST), a label that might be taken to imply a single standardized theory, toward OSS (Donnelly and Steele, 2019; Steele, 2019; Steele and Homolar, 2019) as a way to better capture the plurality of approaches. Irrespective of the precise denomination, however, ontological security has proved fruitful for addressing a wide variety of theoretical and empirical concerns. It has allowed scholars interested in status (Pacher, 2019; Zarakol, 2010, 2011), revisionism (Behravesh, 2018), ideology (Marlow, 2002), and nationalism (Kinnvall, 2004; Skey, 2010) to enter into a conversation with scholars working on identity practices (DeRaismes Combes, 2017), material environments (Ejdus, 2017, 2020), collective memory (Gustafsson, 2014; Mälksoo, 2015; Subotić, 2019), transitional justice and reconciliation (Gustafsson, 2020; Mälksoo, 2019; Rumelili, 2018), diasporas (Abramson, 2019; Kinnvall and Nesbit-Larkin, 2009), regionalism (Russo and Stoddard, 2018), foreign policy (Darwich, 2016; Lupovici, 2012; Mitzen and Larson, 2017; Oppermann and Hansel, 2019), power transitions (Chacko, 2014; Young, 2017), popular protests (Solomon, 2018), populism (Browning, 2019; Kinnvall, 2018; Steele and Homolar, 2019), or security communities (Berenskoetter and Giegerich, 2010; Greve, 2018). Thus, ontological security scholarship has certainly succeeded in inaugurating a new research agenda and in generating new interpretations of a great variety of issues in international politics.
Despite its many merits, however, and in spite of the by now fairly large body of research, ontological security scholarship in IR is also marked by a considerable amount of conceptual ambiguity, especially with regards to the key concepts of anxiety and ontological insecurity. These conceptual issues, we suggest, can be traced to Giddens’ treatment of the concepts (1991: 35–55). Those who define anxiety typically do so by clearly distinguishing it from fear. While fear is seen as linked to a specific object or threat, anxiety is understood as involving a more diffuse or general feeling of insecurity (e.g. Chernobrov, 2019; Ejdus, 2020; Rumelili, 2015a). Definitions of ontological insecurity typically contrast it with ontological security. While ontological insecurity is, according to one very clear definition, a “deep incapacitating state of not knowing which dangers to confront and which to ignore, i.e. how to get by in the world,” ontological security, is by contrast, “the condition that obtains when an individual has confident expectations, even if probabilistic, about the means-ends relationships that govern her social life” (Mitzen, 2006: 345). Ontological insecurity and anxiety, however, are not defined in relation to each other in such a fashion. Rather, some seemingly prefer the term anxiety to ontological insecurity (e.g. Chernobrov, 2019), but more often than not, the terms are conflated in practice with both tending to be treated as opposites of and threats to ontological security (e.g. Giddens, 1991: 35–55). Those who do provide clear definitions of both anxiety and ontological insecurity separately sometimes seem to suggest that ontological insecurity is a more severe condition than anxiety, and that the latter might turn into or cause the former (e.g. Ejdus, 2020; Karp, 2018; Klose, 2019; Mitzen, 2006). Moreover, when reading empirical applications, one sometimes gets the impression that ontological insecurity, used synonymously with anxiety, is observed more or less everywhere. To the extent that anxiety is understood as leading to ontological insecurity under certain conditions, distinguishing between the two is an analytical step that is often skipped over. Because they are not discussed in relation to each other, the differences and relations between anxiety and ontological insecurity remain unclear. This raises several important questions: Are anxiety and ontological insecurity synonymous? Or does anxiety lead to ontological insecurity? Is ontological insecurity a particular kind of anxiety?
Conceptual issues aside, the literature on ontological security has also been subjected to a number of criticisms, two of which we seek to address here. One prominent criticism relates to whether a psychological concept like ontological security is applicable to collective actors like states in the first place, considering that it was originally coined purely for the purpose of understanding individuals. Not all ontological security scholarship in IR is state-centric and there exists a growing body of work that is concerned with everyday practices and ontological security at the level of the individual (Browning, 2018b; Innes, 2017a; Mac Ginty, 2019; Steele, 2019) or the community (Cash, 2017; Kinnvall and Nesbitt-Larkin, 2009). However, the majority of works still takes the state as their unit of analysis, which is perhaps why the debate about ontological security’s applicability to the state is so enduring. Mitzen (2006: 351–353) offers three justifications for “scaling up” from the individual to the state level. Her first observation is that the attribution of individual needs to states actually underlies all of IR, even those accounts that presume that states crave physical security and act out of fear of others. Second, the application of the ontological security concept to states can be justified by the fact that states, in turn, satisfy the ontological security needs of their members. Individuals become attached to the state, which subsequently acts to protect its members’ group identity. Third, even though states are not people, thinking of them as such is theoretically useful because it can explain macro-level patterns. Steele (2008) takes issue with Mitzen’s black-boxing of processes within states. Consequently, he offers an alternative defense centered on state representatives. In his view, because state leaders have the moral burden to represent the state, they “are the state” (Steele, 2008: 18), or, to put it differently, “ontological security obtains for states because state agents seek to satisfy the self-identity needs of the states which they lead” (Steele, 2008: 19). Likewise, Roe (2008: 785–787) holds that ontological security-seeking can be ascribed to states if we accept that they, like individuals, promote and are constituted by values.
These justifications notwithstanding, several scholars have challenged the application of ontological security to state actors. Lebow (2016) is especially critical of how ontological security research treats states as persons with psyches and emotions, stating that it “makes no sense to speak of their psychological needs, especially anxiety reduction” (2016: 36). Krolikowski (2008) argues that the assumption of state personhood is too often simply assumed to be theoretically useful, rather than empirically demonstrated to be so, and that problems arise when concepts are taken from other disciplines “without due consideration for their theoretical origins” (2008: 110). She finds that the concept of ontological security is better applied to individuals, because only an individual-level application can shed light on how the state as an institution impacts the ontological security of its members. For Krolikowski, a state-as-actor analysis obscures this aspect and impoverishes Giddens, who was after all concerned with how modernity, including states, affects the human condition. Croft (2012) holds that ontological security should be utilized to investigate the collective identities of individuals within and across state boundaries. This is because any application of the concept to states ends up reifying the state as an institution by separating it from other social institutions, and thus overlooks the fact that insecurities are intersubjectively constructed among individuals. Similarly, others have suggested that rather than treating the state as a monolith, scholars should explore various sites that influence state ontological security (Steele, 2017). Some have also warned of the “mistaken analytical move by OST [Ontological Security Theory] scholars to homogenize the national self” and point toward a need for listening to the marginalized voices within a national discourse (Delehanty and Steele, 2009: 526). In this way, ontological security research has been criticized for sometimes uncritically applying an individual-level concept to states.
A second important strand of criticism focuses on how ontological security has been adapted and put to use in IR, and identifies a bias in favor of continuity and the maintenance of the status quo. For example, Browning and Joenniemi (2017) argue that because the literature tends to conflate the self with identity, the meaning of ontological security has been reduced to one of mere identity preservation. This is indeed reflected in the vast majority of empirical applications, which treat ontological security as “confidence in one’s identity” (Della Sala, 2017: 547) or “an actor’s ability to defend its identity” (Lupovici, 2012: 812) rather than as “security of the self.” Even though evidence for ontological security-sustaining behavior, as a presumed constant, should be visible across the full spectrum of behavior, including cooperative relations (Greve, 2018), ontological security is in practice usually applied to situations of identity-related conflict. Examples are geostrategic competition, as when Russia is said to reaffirm its identity through conflict with the EU (Hansen, 2016) or vice versa (Browning, 2018a), the aftermath of civil unrest, as when ontological insecurity is taken to persist in Northern Ireland (Cash, 2017; Kay, 2012) and Cyprus (Innes, 2017b), or inner-organizational tensions, as when the EU and its members are described as in identity competition with each other (Della Sala, 2018).
As a consequence, ontological security has often been equated with the absence of contestation, a move that leads scholars to detect ontological insecurity almost everywhere given that contestation is elementary to politics. This has prompted some scholars to conclude that ontological security is in fact not achievable (e.g. Cash and Kinnvall, 2017; Kinnvall and Mitzen, 2018). Furthermore, the equation of self and identity has led to an association of ontological security with securitization practices and of desecuritization processes with insecurity (Browning and Joenniemi, 2017). This bifurcation has arguably made it difficult to view change as a positive force since all change automatically comes to signify instability and anxiety. While exceptions exist that have explicitly sought to separate the self from identity (Chernobrov, 2016; Flockhart, 2016), the widespread conflation of the two in the literature has led to an excessive preoccupation with identity maintenance and continuity, which comes at the expense of being able to account for change.
It is, of course, possible to adopt a pragmatic stance and argue that it does not matter whether the concept of ontological security is “actually” applicable to states or whether ontological security scholarship is biased toward continuity and protecting the status quo, because its introduction into the discipline has proven to be analytically useful and productive. However, due to the significant liberties the literature has taken with the original formulation of the concept, OSS in IR remains vulnerable to criticism concerning the interpretation and application of the term. This issue is compounded by a relative lack of reflection on the origins of the concept and the fundamental differences between the ways it is used in IR and how it was originally conceived. A notable exception to this is Croft and Vaughan-Williams (2017) who point out that the notion of ontological security has undergone a substantial rereading in IR to better “fit” into the discipline and speak to a state-centric literature. The adaptive changes have in fact been so far-reaching, according to them, that the concept is changed almost beyond recognition. But while they note that the term ontological security was specifically coined with individual human beings in mind, and is thus “a function of the problem that he [Laing] sought to address – understanding schizoid and schizophrenic persons – [and] not a philosophical expression of the nature of the ontology itself” (Croft and Vaughan-Williams, 2017: 15), they do not elaborate on this point or draw out what it means for IR to designate states as ontologically insecure. For this we turn to the next section.
Laing, Giddens, and the origins of ontological security
As outlined in the previous section, the criticism that the concept of ontological security is not applicable to states is not new. What has so far not been spelled out, however, is that the concept of ontological insecurity was not developed to characterize individuals in general, but particular kinds of individuals suffering from a pathological condition. R.D. Laing, who coined the term, was a psychiatrist who in his 1960 study The divided self seeks to convey an understanding of his schizoid and schizophrenic patients—individuals who cannot keep up with the expectations put on them by society. Laing defines ontological security as experiencing oneself as “real, alive, whole, and, in a temporal sense, a continuous person” (1990: 39). Such a sense of ontological security, “ontological” here being used as the adverbial form of “being,” allows one to face the challenges of life head-on and to encounter others and the world from an essentially firm point of view. This feeling of being securely oneself is thereby contingent on recognition and congruence between subjectively felt and externally ascribed identity (Laing, 1990: 35–37), which in turn validates each person’s self and makes relating to others potentially gratifying. It is because ontological security thus enables one to feel like a separate and autonomous being that one can genuinely relate to others. In this way, a sense of ontological security forms the basis for living a “normal” life.
However, the individuals Laing is concerned with are not ontologically secure. They experience themselves as split in a double sense: from the world and from themselves (Laing, 1990: 17). They are schizoid. They do not feel themselves to be real, alive, and whole, and because they do not have a firm sense of who they are, no one can ever recognize them for who they take themselves to be. Their being-in-the-world cannot be validated, which means they do not come to view themselves as independent and autonomous persons and therefore cannot relate to others as ontologically secure individuals can (Laing, 1990: 42–47). For them, every aspect of life, every personal contact poses a potential threat to their very existence, generating debilitating forms of anxiety like that of being completely engulfed by others, imploding under the pressure of a reality “out there,” or being turned into a mere object without any subjectivity. Importantly, ontological insecurity in this sense refers to an utterly subjective feeling. It does not matter whether a person’s existence is “really” under threat; what matters is that this is how that individual experiences the world and that the corresponding feelings of dread are real.
According to Laing, for most people establishing a basic sense of ontological security does not require much effort. But for ontologically insecure people all their energy goes into maintaining a tenuous sense of self, to the point that they struggle with tasks that go beyond the mere defense of their being. When dealing with this constant existential anxiety becomes too much, some individuals will split themselves further and invent fake persons with which to face the world. This is when schizophrenia becomes a coping mechanism to sustain a minimal sense of self. Because schizoid persons do not have a firm sense of self, they are continuously terrified of simply ceasing to be. They experience a severe kind of anxiety that in effect threatens their existence. This is what it means to describe someone as ontologically insecure. It should be obvious that this is not the experience of most people in the world. Importantly, while the nature of Laing’s work means that he is largely concerned with individuals suffering from a pathological condition, he recognizes that ontologically secure people also feel anxious when they are misrecognized or when their needs for significance and genuine relationships are not met (Laing, 1990: 80). However, their anxieties are different because they “do not arise with anything like the same force or persistence” as they do in fundamentally ontologically insecure persons (Laing, 1990: 65). While not being explicit about it, Laing implies that ontologically secure and ontologically insecure persons experience different kinds of anxiety; the former experience it in a “normal” and temporally limited way, while for the latter anxiety is permanent and existentially threatening.
In other words, the term ontological insecurity was coined to help describe a pathological condition that is quite rare and that can lead to extreme coping mechanisms. Even though Laing’s work has resonated widely, it was never meant to be transposed to a different context and Laing himself never saw any applicability beyond the kinds of patients he was trying to understand (Prince, 2005). However, Giddens (1991) borrows the term from Laing, although he cites him but sparsely, for his inquiry into the impact of modernity on the human condition. Giving the concept a sociological spin, Giddens defines ontological security as possessing “on the level of the unconscious and practical consciousness, ‘answers’ to fundamental existential questions which all human life in some way addresses” (1991: 47). In how we “live” those answers, Giddens emphasizes the importance of reflexively monitored routines as well as the ability to keep one’s biographical narrative going (1991: 37–41, 52–54). Unlike Laing, he is much more concerned with the individual’s being in society than with the individual’s psychological make-up. Nonetheless, he follows Laing in describing “feelings of ontological security [as] characteristic of large segments of human activity in all cultures” (Giddens, 1991: 36) and as something we normally take for granted, while ontological insecurity, conversely, denotes a condition that makes it hard to cope. Consequently, for ontologically insecure individuals, anxiety can have a debilitating effect that manifests itself in overly rigid routines or in phobic and compulsive behavior, like anorexia or overeating (1991: 45, 105–107). Hence, Giddens, too, understands being ontologically insecure as something not ordinarily experienced by most people and as closely related to neurotic conditions, even though some of the examples he mentions seem perhaps not as severe as the ones discussed by Laing. Importantly, however, he explicitly rejects the distinction between normal and neurotic anxiety, which, as related above, is implicit in Laing’s understanding. According to Giddens, “[a]ll anxiety is both normal and neurotic” (Giddens, 1991: 45), so that pathological behavior is not a function of different kinds of anxiety but of different degrees.
What does this tell us about the usage of the concept of ontological security in IR? For one, it tells us that, technically, to label states as ontologically insecure is to go beyond the conventional “as if” treatment of states as persons and to ascribe to them a human-level of psychological depth. Not only does it imply that states are capable of experiencing all-out terror and anxiety that threatens their existence, but also that they have the potential to develop a pathological condition to the point where they can no longer cope with everyday life. It is clear that this is not an “accurate” description of most states in the international system and that it also does not correspond to the experiences of most of the states empirically investigated in the ontological security literature. Those states seem to know quite well who they are and how to cope with the challenges of life, i.e. with the challenges of international politics. On occasion, they are even depicted as doing so reflexively and creatively. They react by securitizing others or their own identity (Darwich, 2016), by adjusting their biographical narratives (Subotić, 2016), or by pursuing policies to increase their standing in the international community (Zarakol, 2011). These states, described as ontologically insecure, do not find it hard to participate in international society most of the time—something that would be difficult if they were all-consumed by securing their existence.
Second, a closer look at Laing and Giddens tells us that something happened in the transposition of the concept of ontological security from psychology to sociology to IR. This is also evidenced by the fact that nearly every empirical study seems able to identity feelings of ontological insecurity. Is it really likely that every group of people and every state under investigation can reasonably be described as experiencing debilitating anxiety, and as suffering from a pathological condition? Clearly this is not the case, and it is not what scholars of ontological security have in mind when they speak of ontological insecurity. Rather, they seem to have in mind the kind of more limited anxiety that Laing suggests ontologically secure people might experience. It thus seems that ontological security gets at something important, but that by over-relying on Giddens the language it employs does not quite capture what most IR scholars actually do.
One might object that, as long as a concept is useful for providing novel interpretations of international politics, and as long as there is internal consistency in how a term is used, it is not necessary to remain true to the original conceptualization when a term is incorporated into a new discipline. The point we wish to make is not that adapting the concept is wrong. Adapting it has, we would argue, clearly been fruitful. However, the concept has been adapted without discussion of the fact that ontological insecurity was originally used to understand pathological individuals. This is in stark contrast to the debates about scaling up from the level of the individual to that of the state. If it is seen as problematic that the concept was originally developed to be applied to individuals rather than to states, then it is surely also potentially problematic that these individuals were pathological. It is the lack of reflection on this issue that makes ontological security scholarship vulnerable to criticism.
More importantly, we believe that the conceptual vagueness in OSS can, in part, be traced to Laing and Giddens, and particularly to Giddens’ rejection of the distinction between normal and neurotic anxiety. As a consequence of this move, the insights of works that highlight this key distinction were lost. The nature of Laing’s original inquiry and Giddens’ subsequent adaptation give the impression that actors are either ontologically secure, in which case they are well-functioning and appear to experience anxiety only to a limited extent, or ontologically insecure, in which case they are highly pathological. As outlined above, within IR scholarship this apparent dichotomy is reflected in that both anxiety and ontological insecurity are often treated as opposites of a condition of ontological security, which makes it unclear exactly how, if at all, these concepts are distinct or related.
In what follows below, we suggest that this ambiguity can be overcome if we return to the existentialist anxiety literature by which Laing was greatly influenced. This literature offers a clear distinction between normal and neurotic, or pathological, anxiety. The distinction between normal and neurotic anxiety is useful, since it clearly acknowledges that we all experience anxiety and that anxiety is a normal part of the human condition, while still allowing for the possibility that it can also be pathological (with more severe consequences). Drawing on this distinction, we suggest, can bring greater conceptual clarity to OSS and aid in addressing the issues concerning the application of the concept to states as well as its difficulties in accounting for change.
Returning to the distinction between normal and neurotic anxiety
This section discusses how a return to the existentialist anxiety literature can contribute to OSS. To be clear, this article is not the first to incorporate insights from this literature. Some existing IR research has already drawn on thinkers such as Kierkegaard and Tillich in fruitful ways (e.g. Browning, 2018b; Ejdus, 2020; Gustafsson, 2016; Rumelili, 2015a). For example, Rumelili’s edited volume, which draws on Kierkegaard, shows that identity-related anxiety does not necessarily lead to conflict, but that it can also be a catalyst for more peaceful intergroup relations (2015a). The idea that insecurity or anxiety is not something that should always be avoided is also echoed by Browning (2018b). Employing Tillich’s notion of “the courage to be,” he emphasizes the importance of facing anxiety as well as the possibility of obtaining a feeling of ontological security in spite of such anxiety. These existing studies have successfully incorporated selected insights from parts of the rich existentialist literature on anxiety. Still, they have not fully recognized the importance of this literature for the concept of ontological security and its usefulness for developing OSS. We seek to go further than these studies by demonstrating its relevance for understanding and addressing fundamental conceptual issues at the heart of ontological security scholarship.
Before we can outline how the anxiety literature can be used to address the conceptual issues and criticisms of ontological security scholarship discussed above, it is necessary to demonstrate that the two literatures can indeed be integrated. That the two literatures are closely linked should not come as a surprise given that Laing drew explicitly on philosophers in the phenomenological existentialist tradition, including Heidegger, Merleau-Ponty, and Sartre, and was heavily influenced by Kierkegaard (Beveridge, 2011). However, the anxiety literature’s relevance for IR becomes particularly clear in how it treats the term “security.” IR research on ontological security relies on a sharp distinction between security as traditionally understood in IR, i.e. as physical security or survival, and ontological security (cf. Mitzen, 2006; Rumelili, 2015b; Steele, 2008), and existentialist psychologists have a similar distinction in mind when they discuss “security.” While the descriptor “ontological” is not used, distinctions between physical needs and the need for (ontological) security are made. For example, existentialist psychologist Harry Stack Sullivan differentiates between “satisfactions” and “security.” Bodily activities such as eating, drinking, and sleeping are satisfactions, while security is related “more closely to man’s [sic] cultural equipment than to his bodily organization” (Sullivan, 1946: 6). “Cultural,” in this context, refers to “all that which is man-made” (Sullivan, 1946: 6), meaning that what matters for an individual’s security is related to ideas about societal conventions that are imbued into the individual through its significant others (Sullivan, 1946: 6).
Several existentialist psychologists identify “security” as something that is not physical but sought and developed in interpersonal relations, and that is thus fundamentally relational (Horney, 1999; Sullivan, 1946). Security is threatened by anxiety when a value central to a particular subject’s sense of self is somehow at risk (Horney, 1999; May, 1977; Sullivan, 1946). For Karen Horney, anxiety is closely linked to hostility. When an individual’s sense of security is threatened and they are made to feel anxious or helpless, they often become hostile toward those they take to be responsible for their anxiety and helplessness (Horney, 1999). For Sullivan, the apprehension of disapproval from significant others can lead to anxiety. Disapproval threatens the subject’s interpersonal world. In order to protect oneself from feelings of insecurity and powerlessness, the subject avoids actions that lead to disapproval and develops a set of “actions, thoughts, foresights and so on” (Sullivan, 1946: 6).
These examples suggest that “security” is understood in the existentialist psychological anxiety literature in a way similar to how ontological security tends to be used in IR. This is not to say that the concepts are completely identical, only that they are similar enough to justify greater integration of the two literatures. We can thus speak of anxiety, as understood by the existentialist psychologists, as akin to a threat to ontological security as conventionally understood in IR scholarship. Having said this, we are now in a position to return to the conceptual issues in ontological security scholarship in IR and the two prominent criticisms outlined above.
Addressing conceptual ambiguity
As mentioned above, the individuals that Laing originally sought to understand were highly pathological (1990). When we say that a state is ontologically insecure, we are thus not only seeking to understand state behavior by using a theory originally developed to understand individuals, but we are using a theory that was developed to understand individuals who suffer from a severe pathological condition which is quite rare. It follows that if we describe states as ontologically insecure, we might be criticized for suggesting that they are capable of being psychologically ill. Perhaps states could be seen as suffering from such a pathological condition, but it would probably be quite uncommon. This might be taken to imply that the applicability of the ontological security concept in IR is actually quite limited but that is not our intention. Rather, we aim to extend the analytical reach of the concept by incorporating the distinction between normal and neurotic anxiety.
If instead of taking as the key distinction that between ontological security and insecurity, we turn to the distinction between normal and neurotic anxiety, an avenue is opened up for specifying how ontological insecurity and anxiety might relate to each other. This makes it possible to achieve greater conceptual clarity. While Giddens rejects the distinction between pathological and normal anxiety on the grounds that all anxiety is both normal and neurotic (1991: 45), we argue that it is central to making sense of both the concepts of anxiety and ontological security. This is because if anxiety is seen as normal, it provides firmer theoretical ground for applying it both to individuals and states than a concept originally meant only to be applied to pathological individuals. The distinction between neurotic and normal anxiety can be found in the work of several anxiety theorists, including Horney (1999) and Tillich (1952), but Rollo May discusses it in most detail and is therefore the main focus of this section.
According to May, anxiety is “the apprehension cued off by a threat to some value that the individual holds essential to his [sic] existence as a personality” (1977: 189, emphasis in original). This threat to the individual’s “security as a personality” (May, 1977: 190, emphasis added) is a threat to the “meaning” of one’s “existence as a person” (May, 1977: 193). The occasions of anxiety thus vary between different individuals depending on what values they hold dear. Unlike fear, which is related to an external object that can be dealt with as such, anxiety is internal. Because anxiety strikes at the foundation of the personality, it is not possible to “stand outside” of it and treat it as an object to run away from. In other words, it is a “threat to the essential, rather than to the peripheral security of the person” (May, 1977: 191). We cannot therefore view the threat “separately from ourselves, for the very perception with which we look will also be invaded by anxiety” (May, 1977: 191). May’s description of anxiety and its relation to security applies to both neurotic and normal anxiety. However, a key difference between the two is that normal anxiety can be constructively confronted and consciously dealt with if the triggering situation is altered, while neurotic anxiety involves repression or other types of mechanisms of intrapsychic conflict, leading to the creation of defense mechanisms in order to manage anxiety (May, 1977: 193). This process of establishing defense mechanisms is reminiscent of Laing’s account of why people become schizophrenic, and in this way neurotic anxiety is similar to ontological insecurity. Normal anxiety, by contrast, is a necessary part of the human condition and therefore experienced to different degrees by all individuals from time to time.
As an example of normal anxiety, May mentions a socialist who lived in Germany when the Nazis assumed power. As his colleagues were taken away to concentration camps or met other unknown fates, he was aware that he was also in danger, but he did not know for sure if he himself would also be arrested, and if so when that might happen. He also did not know what would happen to him if he were to be arrested. This was not merely a physical threat—it was not just related to the possibility that he might be killed or physically abused. Instead, what was threatened was his freedom to work for his beliefs. This freedom was a value that he associated with the “meaning of his existence as a person.” For May, this individual clearly experienced anxiety, however not of the neurotic but of the normal kind. His reaction was not excessive or compulsive (May, 1977: 193). Normal anxiety can be managed constructively: Every individual experiences greater or lesser threats to his [sic] existence and to values he identifies with his existence in the course of his normal development as a human being. But the human being normally confronts these experiences constructively, uses them as ‘learning experiences’ (in the broad and profound meaning of that term), and moves on in his development. (May, 1977: 194)
In everyday life, normal anxiety, as May sees it, is dealt with through numerous measures: “In all sorts of subtle ways, the manner in which people talk, joke, argue with each other demonstrates their need to establish their security by proving that they are in control of the situation, avoiding what would otherwise be anxiety-creating situations” (May, 1977: 346). Avoiding anxiety-generating situations is indicative of normal anxiety and only becomes neurotic when it is done excessively or compulsively (May, 1977: 346). This normal anxiety is arguably quite similar to what much of the IR literature on ontological security is referring to when they describe particular states as ontologically insecure. Those states are not usually described as neurotic or as trapped within a psychological prison, but as actors capable of coming up with responses to anxiety. States that are dissatisfied with their perceived ranking in the international hierarchy pursue status-enhancing policies (Zarakol, 2011); states that consider their memory narratives to be threatened take action to protect those narratives (Gustafsson, 2014; Mälksoo, 2015); states whose distinctiveness is threatened by similar others seek to distinguish themselves (Darwich, 2016). This discussion suggests that the experience of anxiety does not principally preclude feeling ontologically secure. However, if it is not constructively faced head-on, normal anxiety might later reappear as neurotic anxiety (May, 1977: 356).
In other words, by describing actors, regardless of whether they are individuals or states, as anxious in a normal sense rather than ontologically insecure, we are in a much better position to contextualize and understand their ways of coping with the normal uncertainties of everyday life. This theoretical foundation has the additional advantage of reserving the descriptor ontological insecurity for actors that exhibit neurotic forms of behavior, and thus of remaining true to the original meaning of the term, even though in the case of states such an application would still need to justify why the concept is applicable to collective actors in the first place. But here, too, the anxiety literature provides firmer ground.
Circumventing the problem of applying an individual-level concept to states
As outlined above, applying an individual-level concept like ontological security to states is often regarded as problematic because the utility of such a move is more often assumed rather than demonstrated, and because it simplifies reality significantly as states are not individuals in the true sense of the word. Again, the anxiety literature’s distinction between normal and neurotic anxiety provides a way to sidestep this criticism. Ontological insecurity as understood by Laing is arguably rare in the case of states and large collectives, because it is a purely subjective and uncommon disturbance that tends to have its roots in very specific individual experiences. The same goes for neurotic anxiety as described by scholars like May.
Normal anxiety, by contrast, is not an exclusively individual-level concept. Some scholars working on anxiety, like May and Tillich, are quite explicit in their discussions of normal anxiety as being something not solely experienced by individuals. Instead, they regard it as a cultural and societal phenomenon, to the extent that collectives, societies or particular periods in history can come to be characterized by grave anxiety (May, 1977; Tillich, 1952). For example, Tillich describes the situation in Germany in the 1930s, in which the Nazis came to power, as permeated by anxiety: “First of all a feeling of fear or, more exactly, of indefinite anxiety was prevailing. Not only the economic and political, but also the cultural and religious, security seemed to be lost” (Tillich, 1947: 245, emphasis in original). Similarly, May noted that fascism is “born and gains its power in times of widespread anxiety” (1977: 11).
For our purposes this means that, if societies or historical periods can be characterized by anxiety to a greater or lesser extent, applying it to collectives like states is arguably not a big leap. Again, normal anxiety is a necessary part of the human condition and it arises when a value which is essential to one’s sense of self is threatened. It follows that if a value cherished by a collective is threatened, those who belong to the collective can feel collectively anxious. Indeed, May provides several examples of such collectively experienced anxiety (May, 1977: 8–12; ch. 6). This makes it less contentious to describe a state as experiencing normal anxiety in situations where a value, which its citizens and representatives regard as central to what it means to be a member of that state, is threatened. By this we do not mean to suggest that ontological insecurity, or neurotic anxiety, is never applicable in IR. We could still imagine that in extreme circumstances a state might experience neurotic anxiety, or that it matters at the individual level in a situation where the president of one of the world’s most powerful states is regularly described as a pathological narcissist. What we are saying is that the points made by most ontological security scholars are better captured if understood through the notion of normal anxiety as outlined in the existentialist literature, and that describing states as anxious in this sense circumvents the charge of applying an individual-level concept to states. This is because anxiety as discussed in the anxiety literature, as opposed to ontological insecurity, is not a solely individual-level descriptor to begin with.
Accounting for change and the creative potential of anxiety
Finally, the criticism of ontological security research having difficulties accounting for change can also be remedied by the distinction between normal and neurotic anxiety. Simply put, the former enables change while the latter prevents it, or at least makes it unlikely. The ontologically insecure individual’s rigidity and tendency to cling to the familiar at any cost, as described by Giddens (1991), makes change unlikely if not impossible. Based on the anxiety literature, we propose two links between anxiety and change. The first appears in the understanding of scholars like Horney and Sullivan of anxiety as an essentially interpersonal phenomenon. It is in interpersonal relations that anxiety occurs, and it is only in an interpersonal context that it can be managed and that security can be established. For Sullivan, for example, interpersonal acts are processes of becoming that form patterns. Because they are reciprocal, these patterns are transformative and amenable to change: The interaction brings about something different, new. Because these acts are in [the] process of change, instead of being rigidly fixed and predetermined or haphazardly and willy-nilly occurring, they can to some extent be redirected. Goals can, therefore, be redefined, modified, changed. Within the human situation, there is a possibility of creativeness, of discovering and inventing new goals, new purposes (Mullahy, 1946: 122–123):
A second way in which anxiety can lead to change, rather than inflexible behavior, draws on the human capacity for creativity that May introduces. Again, the distinction between normal and neurotic anxiety is central. May finds that a high degree of neurotic anxiety occurs together with a contradiction, or a gap, between expectations and reality. The neurotic individual constructs an idealized understanding of the world, but this understanding is contradicted by what that individual actually experiences. It might be said that there is insufficient empirical evidence to support the neurotic’s understanding of the world (May, 1977: 332–336). This kind of gap, between expectations and reality, May argues, characterizes not only neurotic, but also normal anxiety. However, unlike in neurotic anxiety, in normal anxiety it can be bridged because one’s situation is more realistically appraised and faced (May, 1977: 355–356). By bridging the cleavage between expectations and reality in this way, a creative act is carried out and something new is brought into being (1975: 39; 1977: 368). May provides several illuminating examples that are worth repeating in their entirety (1977: 368–369): The artist conceives a landscape in his [sic] imagination that has significant form. This is always partially in the way he sees the natural scene and partially produced by his imagination. His painting is the result of his capacity to wed his own expectations — in this case, his artistic conception — with the reality of the scene before him. The man-created picture then has beauty which is richer and more gripping than the inanimate nature from which it was painted. Similarly, every scientific endeavour consists of the scientist’s bringing to bear his own expectations — in this case, his hypotheses — upon reality, and when this process is successful he uncovers some reality which was not known in that way before. In ethics, the person brings to bear his expectations — in this case, ideals of more desirable relations — upon the reality of his immediate relations with other people, and by this means some transformation of his interpersonal relations occurs.
From this discussion, it seems clear that the existentialist anxiety literature provides theoretical accounts of change that are useful for ontological security scholarship. Such a theoretical foundation is better equipped to account for how actors of all kinds, whether states, individuals, or communities, adapt their routines or biographical narrative to a new situation, or how they respond otherwise creatively to perceived threats to the self. In this way, ontological security scholarship in IR could address its continuity bias.
Conclusion
This article has argued for a return to the existentialist literature on anxiety, which can be understood as having been concerned with ontological security before R.D. Laing even coined the term. Such a move could address a number of issues within OSS. Within research on ontological security in IR, there is a lack of clarity concerning the key concepts of ontological insecurity and anxiety. The two are sometimes conflated or treated as synonymous, and it is not always clear how they are related. IR scholarship on ontological security has also been criticized for applying a concept originally developed for analyzing individuals to states, and for prioritizing continuity at the expense of change. In addition, we have highlighted that the individuals Laing had in mind when he wrote about ontological insecurity were highly pathological, which might be seen as making the application of ontological security to states even more vulnerable to criticism. We have launched a friendly intervention that has sought not only to criticize, but also to move ontological security scholarship forward.
In discussing anxiety, existentialist scholars like Rollo May distinguish between neurotic and normal anxiety while still treating the two as similar in important ways. This distinction, which Giddens explicitly rejected, we have argued, is more useful than the common distinction between ontological security and insecurity, as it provides greater conceptual clarity and can help us deal with issues in ontological security scholarship in IR. We have suggested that ontological insecurity and neurotic anxiety can be seen as synonymous, while normal anxiety is different and something that we all experience to some degree. Contrary to Giddens’ argument, not all anxiety is necessarily neurotic. Further, drawing on the existentialist anxiety literature can help us circumvent the charge of falsely applying an individual-level concept to states, because normal anxiety, as understood by scholars like May and Tillich, was never a purely individual-level concept in the first place. It also counters the bias in favor of continuity by providing an avenue for meaningfully accounting for change, including identity change.
In other words, while IR has mainly drawn on Giddens, we have suggested that the theoretical pedigree of OSS stretches back even further than Laing. By drawing on the existentialist anxiety literature, we have emphasized that normal anxiety, unlike ontological insecurity and neurotic anxiety, is a normal and unavoidable part of the human condition, and that it therefore makes sense to think that this is also true for all subjects of IR scholarship, states as well as individuals. Unlike what is sometimes suggested, not all anxiety is utterly debilitating. Not all anxiety causes unhealthy and overly rigid behavior. On the contrary, anxiety can be a catalyst for change and creativity. This point is often obscured by the tendency in ontological security scholarship to use the terms “anxiety” and “ontological insecurity” synonymously. Greater conceptual awareness and precision can add clarity to ontological security research. Thus, we are not advocating for IR scholars to abandon the concept of ontological security completely. That would be unwise given how widely the concept has resonated and how it has contributed to IR. Rather, our suggestion is specifically to differentiate between ontological insecurity, understood as neurotic anxiety, on the one hand and normal anxiety on the other, and generally to pay greater attention to the thinkers who inspired Laing when he first developed the ontological security concept. The existentialist literature on anxiety ought to be seen as a rich theoretical resource that OSS would benefit much from drawing on to a greater extent than has hitherto been the case.
In this spirit, while this article has exemplified how the existentialist anxiety literature can be used to inform IR scholarship on ontological security, it also lays the foundation for further integrating insights from this literature into OSS. While IR scholars like Rumelili (2015a) and Browning (2018b) have already begun to fruitfully incorporate theoretical ideas from work on anxiety, we suggest that a more in-depth engagement with the broader anxiety literature and related research could enrich ontological security scholarship even more. For example, the emphasis in the work of scholars like Horney and Sullivan on anxiety as an essentially interpersonal phenomenon, where for Horney hostility is central and for Sullivan apprehension of disapproval, could perhaps be used to provide a different understanding of relationality within the context of ontological security in IR. While much ontological security research in IR has stressed the importance of recognition in relational processes, hostility and apprehension of disapproval appear to provide a slightly different approach to relationality and its links to anxiety (Horney, 1999; Sullivan, 1946). Moreover, while numerous contributions to OSS have highlighted shame and shaming (e.g. Gustafsson, 2015; Steele, 2007, 2008; Youde, 2014; Zarakol, 2010, 2011), the focus on hostility and apprehension of disapproval in the work of Horney and Sullivan could help broaden the range of emotions focused on in OSS in useful ways.
Footnotes
Disclosure
The authors report no conflict of interest. The authors alone are responsible for the content and writing of the paper.
Funding
The author(s) received financial support for the research, authorship, and/or publication of this article: The authors research received financial support from the Johan and Jakob Söderberg Foundation.
