Reintroducing friendship to international relations: relational ontologies from China to the West

Chinese government representatives and scholars have attempted to ameliorate fears about China’s rise by portraying China as a new and friendlier kind of great power. It is claimed that this represents a new way of relating which transcends problematic Western understandings of Self–Other relations and their tendency to slip into domination and enmity. This article takes such claims as a point of departure, and analyses them with focus on the explicit discussions of friendship in international relations theory. Paying attention to current Chinese thinking which emphasizes guanxi relationships, friendship can contribute to the development of genuinely relational international relations thinking and move beyond a focus on ossiﬁed forms of friendship and enmity

centred on the anxious self. The vantage point of friendship suggests a way out of the dangers of theorizing Self in contrast to Other, and reopens the possibility to conceptualize Self with Other.

Introduction
Chinese policymakers, scholars, and pundits have attempted to ameliorate fears about China's rise by portraying China as a new and friendlier kind of great power. It is claimed that this represents a new way of relating which transcends problematic Western understandings of Self-Other relations, and their tendency to slip into domination and enmity. Claims along such lines can be seen in President Xi Jinping's official discourse, which portrays the Chinese nation as culturally predisposed to friendly, peaceful and harmonious behaviour abroad, and which lists friendship as one of 12 key terms for his socialist 'core value system' at home (People's Daily Online, 2014;Xi, 2014). These claims have been illustrated in various international nation-branding events, often through the Confucian adage that 'it is glorious to receive friends from afar' (Callahan, 2010, p. 2). They have also been an important part of emerging debates over a possible Chinese school of international relations (IR) (Noesselt, 2015). Famously, Zhao Tingyang claims that Chinese traditions offer a 'Chinese ontology, the ontology of relations, instead of the western ontology of things' which enables the peaceful transformation of enemies into friends (Zhao, 2006:33, 34; for a discussion, see Nordin, 2016a;2016b), and researchers discuss 'China's self-perceived role of a friend versus the (often Western) exploiter' (Shih and Yin, 2013, p. 81). Although few foreign policy documents or academic texts focus solely or even primarily on friendship, this terminology is persistent and claims are repeatedly made through the language of friendship.
Taking such use of the term 'friendship' as a point of departure, this article discusses the connection between friendship and IR (cf. King and Smith, 2007;Devere and Smith, 2010;Koschut and Oelsner, 2014). It develops scholarship which understands friendship as a conceptual tool, pointing to a concern with relations between Self and Other. Rather than merely denoting a personal and private relationship, friendship denotes a way of thinking about the co-constitution of Self with Other, and theorizes the dynamics of such co-foundation. Such a concern is as pressing for states as it is for individuals, as much a matter for the political as for the personal. A contribution to theory rather than policy, this article brings together and juxtaposes a number of ways of thinking through the language of 'friendship' and analyses the implications of different theorizations for thinking about world relations.
The article progresses in three main parts. Part 1 summarizes the historical role of friendship in political thought, drawing attention to similarities between the Chinese and European ancient traditions. In these ancient traditions, friendship played a role in cultivating the self and promoting virtues in politics at all levels. Later thought under the state system transformed the meaning and importance of friendship and politics. On the one hand, friendship was increasingly associated with the personal and emotional, and confined to the private realm. On the other hand, 'friendship' was able to appear in the public political realm if it was transformed into overarching and general forms of community such as citizenship and nationality. Importantly, this public friendship was bound to the idea of enmity, and its logic of Us versus Them.
Part 2 builds on these introductory observations. It develops the argument that the discipline of IR has been shaped by a form of thinking which privileges 'things' over 'relations' (Jackson and Nexon, 1999;Nordin and Smith, 2017). This distinction contrasts an analysis which starts by positing or assuming that there are given 'things' (such as states, nations, and power) that pre-exist their relations; and an analysis which starts from the position that 'things' are always the effects of relational processes. To the extent that we can speak of 'things' in the latter case, these are not substances, but formations, contexts, or differences that are always changing and incomplete (Jackson and Nexon, 1999, pp. 304, 314-15). We show how even some theoretical approaches to IR that aver to focus on relations fall back on an ontology of things. This is connected to the marginalization of friendship as a key relationship through which Self and Other are co-constituted. In the process, the central question of friendship -the question of what it means to become with others, and what it means to share and shape a world with others -has been lost to much analysis in IR.
Part 3 suggests that recovering friendship enables a recast IR based instead on 'relational ontologies'. It starts by focusing on friendship through the concept of guanxi relationships that has recently been advanced by Qin Yaqing as a Chinese relational ontology. It then develops Qin's account by discussing Felix Berensk€ otter's argument that friendship between states is important because friends help calm the anxiety that is an effect of ontological insecurity. Drawing on Qin's work, alongside similar concerns of L.H.M. Ling, we argue that contemporary developments of traditional Chinese thought are particularly significant for IR debates because they indicate a co-constitutive Self-Other relationship which does not emphasize anxiety and fear of difference or of misrecognition.
Finally, we conclude that the real divide in understandings of friendship and IR is not between China and the West (nor, for that matter, between ancient and modern). Instead we contrast the role of friendship in the ontology of things on the one hand, and relational ontologies on the other. The first ontology tends to reproduce an essentialist Self-Other dichotomy and ossifies friendship as a role or attribute; the second tends to allow for transformation, and so is open to the coconstitutive dynamics suggested by friendship. Paying attention to the current trend in Chinese IR theory which emphasizes relationships and guanxi friendship can contribute to the development of such IR thinking, and move beyond a focus on ossified forms of friendship and enmity centred on the anxious self. In so doing this article joins in efforts already under way to recentre the relations of IR. Relational thinking can be seen in growing efforts to decolonize the discipline, in favour of a pluriverse of relational ontologies that draw inspiration from non-Western traditions or 'epistemologies of the South' (Santos, 2014;Ling et al., 2017).
In summary, this article shows that an explicit focus on friendship in IR adds a distinct and valuable vantage point on relationality by refocusing attention on the co-constitution and dynamism of self in relations with others. It therefore makes a contribution to the ontology of IR, and offers a resource for understanding the complex phenomena that form the focus of its discussions and concerns.
1 Transforming friendship: from friend-friend to friend-enemy Standard accounts of the vocabulary of politics and IR treat the inclusion of 'friendship' in their lexicon as somewhat of a novelty. Instead, they centre the state, power, sovereignty, citizens, nations, and people.
If friendship appears at all it tends to be tied to enmity, which has also shaped much thought in IR. Whilst many accept that states and nations can have enemies, there is much wider scepticism about the possibilities for friendship (e.g. Keller, 2009;cf. Koschut and Oelsner, 2014, 6-8).
So, why does scepticism persist? In answering this question considerable light is brought to what talk of friendship in IR theory might mean. Many of those who are sceptical about friendship in both politics and IR typically rely on an unexamined assumption about what friendship is and what it must be: the 'contemporary-affective' view of friendship (Smith, 2011b, 12ff). This account assumes friendship to be a private and voluntary relationship between individuals characterized by emotional attachment. A body of scholarship has emerged which challenges this assumption. This challenge has two dimensions. The first dimension draws on Wittgenstein to show that rather than there being one core definition or instance of friendship, friendship comes in varieties, and no one use or practice can be privileged over the others (Smith, 2011a(Smith, , 2014Digeser, 2016). Friendship has been understood as a public and a private relation; applied to states and individuals; obligatory and voluntary; hierarchical and equal; spiritual and secular. That this might sound surprising to the contemporary Western ear illustrates how parochial the use of friendship has become in the Western analytical lexicon, and it underlines the need for critical re-evaluation. The second dimension of the response is to focus on what the critic of friendship claims to know: what 'friendship' is. Here, historical, philosophical, and theoretical work has demonstrated that the 'contemporary-affective' or so-called 'ordinary' use of friendship is far more complex, confusing, and even contradictory than it first seems. Right from the start of the European tradition philosophers have struggled to say what friendship is. The 'ordinary' account might be oblivious to these problems but they are problems nonetheless. If, then, the ordinary language account of friendship is at least not 'the only game in town' (Digeser, 2009), then this opens the possibility for a wider investigation and theorization of friendship in politics and IR. In what follows we offer a sketch of some of the salient features of friendship in ancient and modern Western and Chinese traditions. We aim to demonstrate the transformations that friendship has undergone, and to draw points of contrast and comparison between traditions. This sets the scene for the discussion and re-theorization of friendship in contemporary IR scholarship which follows.
Friendship was a central category for the European ancients, and there is increasing recognition that this was not just an ethical, but also a political category (Price, 1989). This is also the case for friendship in Chinese tradition whose principal source is the thought of Confucius and his heirs (Analects, 1940, p. 1.8, 9.25, 19.3;Kutcher, 2000Kutcher, :1615. Incongruous though this might seem to many contemporary IR scholars, it was not simply that the ancients fused what scholars might currently be inclined to treat separately: ethics and politics, the personal and the public. Instead, ancient theorizations constructed a concept of friendship which was just as central to theorizing political life as 'justice' or 'equality' in Greek tradition, and 'unity' or 'harmony' in Chinese tradition. Thus, although there are differences between seminal accounts of friendship in Chinese and European thought, both traditions understood friendship as an important connector of politics, ethics, and human flourishing. Ancients in both Greek and Chinese traditions were concerned with one having the right kind of friends. Although they were often candid about the emotional, material and social advantages expected from friendship, these were not necessarily the principal reason for friendship. Notably, whilst affection might emerge, praiseworthy friends were to be chosen not so much for their particularity, but because they provided moral nourishment (Hall and Ames, 1998, pp. 254-69). On this logic, Plato suggested that friendship could have a higher purpose connected to politics and the state. In Phaedrus, Plato stages a dialogue that forms the basis of the emerging friendship (philia) between Socrates and Phaedrus. Socrates seduces Phaedrus away from the influence of rhetorician Lysias, first by constructing an impressive speech on love, and then encouraging Phaedrus to become his fellow enquirer into the Good through dialogue. The dialogue ends with Socrates imploring the gods to give him inner beauty, temperance and wisdom, to which Phaedrus responds: 'Please include me in your prayer, for friends hold everything in common' (Plato, 1956, p. 75). It is Socrates' aspiration and ability to be virtuous that makes him a suitable friend for Phaedrus, and it is Phaedrus' sharing of this desire that enables the friendship. For Plato, friendship becomes a necessary condition for realizing the good. It builds on a desire for self-improvement that the friends share, and that the friendship helps them augment and sustain this (Hall and Ames, 1998, p. 256). This common search for the good also linked Platonic friendship to the production of the good state in his Republic (Plato, 1987;Price, 1989;Smith, 2011a). In these accounts, the key usefulness of the friend was to transform the self into something new and better, with more virtuous desires, which would in turn lead to harmony and stability in the state.
In a similar way, Confucian friendship could and should sustain moral growth in support of family-state hierarchies. Confucian thought took five key relationships as its foundation, one of which was that between friends (you 友). The other four relationships denoted mutual obligation between father and son, ruler and minister, husband and wife, and older and younger brother. If everybody fulfilled their role in these relationships, families would be stable and harmonious and productive of the good subjects that would ensure stability and harmony for the state. Friendship should be supportive of the other four relationships, where the good friend might both offer respite from their demands and act as a virtuous example to emulate in order to become a better son, a better official, and a better subject (Kutcher, 2000). The importance of the friend as a virtuous example to emulate is so strong that Confucius repeatedly urges: 'Do not have as a friend anyone who is not as good as you are' (Analects, 1940, p. 1.8, 9.25). Put differently, although one should be benevolent to those less virtuous than oneself, it is only one's betters who support one's self-improvement that can be considered true friends. The key virtue of concern here is the central Confucian value ren (仁), which is usually translated as 'benevolence' or 'love'. The Xunzi recalls a conversation between Confucius and a number of his disciples on what it means for a person to embrace ren, which concludes that the truly enlightened person 'is one who loves himself' (Hsun Tzu [Xunzi], 1966, p. 105.29.29, cited in Hall andAmes, 1998, p. 258). This self-love (zi'ai 自爱) is reflexive, where the self incorporates the entire field of concerns of Self and Other. Ren does not refer to an isolated agent, but as Hall and Ames (1998, p. 258-259) explain, it describes a 'mutually beneficial relationship between self and other . . . a complementarity grounded in the specific conditions of one's cultivated relationship with another person'. Hall and Ames (1998, p. 259) emphasize that the 'self' here is not an 'egoself', but 'an incipient, inchoate self that is radically situational, and hence reflexive'. Moreover, personal and communal realization is made possible by the cultivation and extension of this reflexive self with others. This relationality of the good ren is repeatedly underscored in the Analects, and is immediately visible from the written character ren 仁, which is made up of the radical for 'person' 人 and the number 'two' 二 (see Analects, 1940, p. 6.30, 3.3, 15.39).
These ancient pictures of friendship contrast with views that have become dominant after the advent of the nation-state. Of those thinkers in the European tradition who actively engage with friendship as a feature of politics, Carl Schmitt is infamous. Schmitt identifies the friend and enemy distinction as the defining feature of the political. Yet, in contrast to the rich theorization of friendship in the ancient literature, Schmitt is surprisingly silent on what 'friendship' is. When Schmitt does elaborate on the meaning of friendship it is cashed-out in public terms and in relation to the homogenous identity of a people. He claims that such identity is one of the foundational principles of the state. When a people is conscious of its own identity as a nation 'it has the capacity to distinguish friend and enemy' (Schmitt, 2008, p. 247). Moreover, whereas the ancient literature is able to theorize political forms of friendship without reference to enmity, Schmitt appears unable (or unwilling) to do so. The definitions of enemy and friend are tied together and linked to the possibility of killing and being killed (Schmitt, 1996, pp. 26-27, 37, 45-46). Although Schmitt maintains that we cannot determine in advance who is friend or enemy, these categories remain permanent features of his political thought, seemingly incapable of transformation. In many ways, Schmitt's thought is paradigmatic of the fate of friendship in the modern state system; a fate which understands friendship as the less significant other of enmity, and ossifies both categories.
It is perhaps not surprising that this view of friendship is paralleled in Mao Zedong's thought insofar as it embraces the state, and a different applied article could profitably read Mao's China through a Schmittean lens (indeed, Schmitt sometimes portrayed the dialectic between the proletariat and their oppressors in terms of the friend and enemy distinction). Fairly soon after China's integration into the modern interstate system, society became dominated by Mao's interpretation of Marxism-Leninism, which challenged the ancient understanding of friendship and to a large extent broke away from Confucian thinking (Kam, 1980). The fundamental importance of friendship can nonetheless be seen to have remained, albeit in a new guise. In Mao's China this took a domestic and an external form. Domestically, friendship cast the highest friend as the selfless communist comrade (tongzhi 同志). The external form developed in the international arena was foreign-friendship (youyi 友谊). These Maoist terms captured the terrain of friendship for the communist cause, and realigned it from emphasizing harmony to a more confrontational polemicization. It no longer focused on mutually constituting elements in a harmoniously transforming relationship (as in Confucianism), but based its dialectic on dichotomized units that clashed in painful revolutions to push history forwards. Under Mao, such dichotomized entities included the 'friend' and the 'enemy'. As Mao posed the question: Who are our enemies? Who are our friends? This is a question of the first importance for the revolution . . . To ensure that we will definitely achieve success in our revolution and will not lead the masses astray, we must pay attention to uniting with our real friends in order to attack our real enemies (Mao, 1961, p. 13).
In summary, this sketch of friendship in ancient and modern state-centric thought illustrates a displacement of friendship and what it originally represented: profound relationality and co-constitution of Self and Other. The effects of this displacement are reflected in contemporary thought about politics and IR. In ancient thought friendship had a structuring role for individuals and political systems. Connected to virtue, it was a means by which political life could be stabilized and made harmonious. In modern politics, where the modern state has become dominant, the purview of friendship has been bifurcated. Along one branch, personal and individual friendships based around the emotions are now allowed to flourish -but strictly in the private sphere. Along the other, friendship is abstracted and put to work by the state in relations of group-belonging such as comradeship and nationality. Friendship is formulated as an Us in opposition to a Them; a Self opposed to Others. This suggests a profound shift in thinking about politics, which turns attention away from the possibility of relational production, reconciliation and even combination of distinctive and contrasting components. Instead, this shift emphasizes the assertion and preservation (or annihilation) of distinct and Reintroducing friendship to international relations 377 antagonistically opposed things. This latter view characterizes the ontology of much IR scholarship.
2 Marginalized friendship: IR and the 'ontology of things' If it is the case that the state system has marginalized friendship as a concern with co-constitutive Self-Other relations from the political, how is this more broadly reflected in IR? This part revisits the ontological assumptions of 'mainstream IR' arguing that much contemporary IR rests not on an ontological focus on relations, but on an 'ontology of things' which has marginalized friendship. Surprisingly, not only is this true of Realism and Liberalism, but also of Wendtian Constructivism where discussions of friendship explicitly appear. The fuller meaning of relational ontologies and the contribution of friendship to IR is explored in the next part.
Common stories about the development of IR depict a Eurocentric discipline shaped by two related concerns. The first of these is an attempt to understand states as elements of a systematic whole (Waltz, 1979). The second is a focus on the establishment and maintenance of peace, or at least the avoidance of unnecessary conflict between states (Jervis, 1999, p. 42). These two objectives have produced a range of theories of which varieties of Realism, Liberalism, and Constructivism are often said to be 'mainstream'. Of these, Realism is habitually said to be dominant. One reason for this is that Realism emerged near the conception of the discipline of IR, as it is commonly rehearsed, and so other theories are in some way a response to Realism. More importantly, Realism set the tone for both the ontological assumptions of IR and its lexicon. This vocabulary is more than a choice of words, it identifies the 'things' of IR. Ontologically speaking, Realism set the rules of the game by setting the first lexicon of the discipline as it came to imagine itself. Key within this lexicon are 'sovereignty', 'power', 'state', and 'nation'. Realists tend to view the international state system in terms of 'anarchy' constituted by sovereign states, seeking survival and locked into a 'security dilemma' as a result (Morgenthau, 1972;Mearsheimer, 1990, p. 5-56). From its inception, then, IR was infused with an ontology of things. The state was taken as the object of IR, and assumed to be a self-contained, independent, and unchanging unit. This view is exemplified by Waltz when he writes that states 'are unitary actors who, at a minimum, seek their own preservation and, at a maximum, drive for universal domination' (Waltz, 1979, p. 118).
Realism has adapted, transformed, and has been remarkably influential. As a result, even theories that offer an alternative approach or methodology for understanding the international system have tended to do so by positing themselves as critiques of, or alternatives to, Realism. Moreover, they have tended to accept the lexicon of Realism and its ontology of things. For example, most Liberals accept the basic Realist assumptions concerning the state and the basic rules of the game, in terms of the wider ontology (Milner, 1993, p. 9;Keohane, 1993, p. 272;Jervis, 1999, p. 43-44). Of course, Liberals differ from Realists by pointing to the role of norms and actors other than the state. However, by doing so they merely stress an additional consideration to that of the state; they do not seriously challenge its importance. As Robert Jervis (1999, p. 45) has observed, what differs is the level of focus. Liberals and Realists have conducted a lively debate, but they have been able to do so precisely because they share a lexicon and an ontology. Having identified their objects of study, they are only then concerned to theorize their interaction. In the words of Erik Ringmar (1996, p. 441) 'the state is given exogenously to the analysis . . . and hence endowed with something akin to a transcendental ontological status'. States are treated as pre-constituted calculating machines much in the same way that Hobbes theorizes human beings (Ringmar, 1996, p. 447). Factored out of IR is precisely the idea that the units in question can be transformed -moreover, constituted -through their interaction.
If Realism and Liberalism share a core set of assumptions (and ontology of things), Constructivism appears to offer something new. Indeed, at first blush Constructivism appears to move away from the assumptions that underpin debate between Realists and Liberals, and to offer an alternative ontology. Here Alexander Wendt's 1999 book Social Theory of International Politics is paradigmatic. In this book, Constructivism is not a theory of IR, but a way of approaching ontology that is applied to a subject matter or field of enquiry (Wendt, 1999, p. 7). Wendt then uses the tools of Constructivism to intervene in the debate between Realists and Liberals. His purpose is not to challenge their use of 'the state', but to confront how this is understood ontologically (Wendt, 1999, p 1). Wendt accepts that there are relations within the state, but this is not his concern (Wendt, 1999, p. 246). Instead he attempts to show how states are to be considered persons. For Wendt, state behaviour does not depend on hard structural facts relating to 'anarchy' or fixed intentions (as Realists and Liberals might be inclined to think), but on how roles and identity are formed and maintained by interaction with others (Wendt, 1999, p. 257).
Wendt focuses on three roles that he says states cast for themselves and each other: enmity, rivalry, and friendship. Wendt's point is that (contra Realists) anarchy has no predictive power for state behaviour. State behaviour is determined by how states identify themselves and others. This identity formation is relational. As Wendt writes: What this means is that in initially forming shared ideas about Self and Other through a learning process, and then in subsequently reinforcing those ideas casually through repeated interaction, Ego and Alter are at each stage jointly defining who each of them is (Wendt, 1999, p. 335).
Yet, despite appearing to offer an alternative to the 'ontology of things', Wendt's theory of international politics falls back on it. Ironically, the feature of Wendt's social constructivism which connects it squarely to an ontology of things is precisely the one which suggests his thought might exemplify a relational ontology: his discussion of the social construction of 'the identities and interests of purposive actors' (Wendt, 1999, p. 1), and in particular his use of the terms 'Self' and 'Other'. Whilst this language might suggest that the ideas of Self and Other are doing significant work in Wendt's theory (he himself capitalizes them), in fact they elide complex theorization rather than signal it. Wendt's terms are little more than placeholders for what he must assume, but not theorize. This is best illustrated by comparing Wendt's thought to other traditions that also employ these terms, not least feminism, post-colonialism and post-structuralism. In these traditions, ideas of Self and Other are drawn from a wider philosophical debate focused on the alterity of Self and Other. These traditions also point to the fluid and even contradictory meaning of the Self. This deeper meaning is not Wendt's concern. Indeed, whilst Wendt makes the rather strong claim that the state is a kind of person, he does not spend much time thinking about the tensions inherent in the notion of personhood itself. Whilst real persons can have a range of relationships and roles, Wendt limits these to three in his cultures of anarchy: enmity, rivalry, and friendship. Wendt has a rather limited and static view of the relationship of friendship. Indeed, as Wendt claims, it is not a relationship but a role.
It is worth thinking here about what roles mean. Wendt uses the example of the role of the president of the USA as an illustration (Wendt, 1999, pp. 258-259). This is instructive as the powers of the president are famously defined and limited by the constitution of the USA. Irrespective of the person who fills this role, qua president that person can only act in certain prescribed ways. In Wendt's world, friends must rehearse their friendship in a narrowly scripted way. This is peculiar when we consider the rich variety of friendships individual persons conduct. Wendt's limitation of the possible relationships that a state as person can have, and in particular his limitations of the transformative nature of friendship, does not indicate a relational ontology but an ontology of things. The state's relations are not transformative of Self and Other. Although the roles that states adopt can change their interests, the roles themselves are fixed.
The consequences of this identity fixing becomes especially evident when Wendt conflates Self and Other with Ego and Alter. These terms carry different connotations. Whilst self and other (and especially Self and Other) can indicate radical difference or alterity, Ego and Alter are, in fact, a linked pair. In Latin, 'Ego' denotes I, we, myself, and us, whereas 'Alter' does not mean Other in the strong sense that links it to difference, strangeness, and alterity, but the other of two, the second, the other one. Thus the idiom alter ego: the other, which is linked to the first I and has identity with it.
Thus, although Wendt talks about a constitutive relation between Self and Other, we might question how serious he is about this. The other encountered in this relationship is really a form of self. Wendt's three cultures of anarchy do not depend on difference between actors, but on what they share. The only real encounter of difference in his book (between Self and Other, rather than alter and ego) is that of Cort es and Moctezuma (Wendt, 1999, p. 158). This encounter is telling as it exemplifies exactly what Wendt cannot theorize with his view of 'relations'. The encounter between Cort es and Moctezuma is a real encounter between Self and Other. The two are alien to each other; they do not fit into one other's script. By using this encounter to illustrate the claim that culture needs to be shared, Wendt betrays the fact that such shared culture depends on a relation of self-self, not Self-Other. Furthermore, what the Self and Other are in Wendt's thought remains 'fixed' (at least at the level of his analysis). What is transformed is not the self or the state (the ontologically pre-existing things) but the identity, intentions, and behaviour of those units. Identity is central and possible precisely because in others the self sees not alterity and difference, but an Alter Ego.
From this discussion of Realists, Liberals and Constructivists it is clear that even accounts that have tried to theorize the relationships of IR, including friendship, have tended to fall back on an ontology of things. In doing so, they have failed to account for states as embroiled in complex processes of becoming with others. In this way, much IR literature fails to adequately engage with the central questions raised by friendship, questions concerned with the co-constitution of Self and Other. Nevertheless, friendship persists in IR as an intellectual space or question in need of theorization.

A return to friendship: 'relational ontologies', guanxi, and Daoist dialectics
The first of the previous two parts outlined the bifurcation of friendship in much modern political thought. This bifurcation saw friendship displaced from the possibility of theorizing Self with Other, to an understanding of friendship as either private (and thus not truly political), or a community bond which opposes 'Us' (the Self) to 'Them' (the Other). Such 'Us versus Them' relations tend towards mutual antagonism. In the second part this bifurcation was related to the development of IR based on an 'ontology of things'. This ontology theorizes the existence of discrete entities as prior to any relations that they might have.
We are now in a position to consider the contribution of Chinese scholarship to these debates about ontology and Self-Other relations in IR. In what follows we show how Daoist dialectics can complement the 'ontology of things' that has stressed the conflictual incompatibility of Self and Other. The focus on relational ontologies in recent Chinese thought provides a platform to reintroduce friendship to the IR discipline. Such a reintroduction does not only refocus on relations, but on the very possibilities of thinking Self with Other. We develop this line of thought by discussing friendship and relationality in three key theorists: Qin Yaqing, Felix Berensk€ otter, and L. H. M. Ling. Qin argues that 'guanxi' relations (关系) should be the hard core of a Chinese IR theory, and uses Daoist 'Chinese dialectic' to overcome what he sees as the conflictual understanding of dialectic in the West. Such a dialectic denies dichotomy and suggests mutual structuring. Berensk€ otter is also concerned to overcome the dichotomy of Self-Other, and suggests friendship can tame the ontological anxiety of the state. Such a relation accepts, but reconciles, difference. Finally, Ling develops relationality through Daoist dialectics. Rather than separate Self and Other, and attend to the anxiety induced by this separation, Ling points to the co-dependency and intermingling of Self with Other.

Qin: guanxi (关系) and dialectics
A proponent of a 'Chinese school' of IR theory, Qin Yaqing, has been key to refocusing relationality in IR. Qin argues that the basis for Western IR theory is 'individuality' whilst the Chinese model is focused on 'relationality', upon which he proposes a 'relational theory of world politics' (Qin, 2009, p. 5;Qin, 2016;2017;2018). Furthermore, Qin (2009, p. 5) suggests, 'mainstream International Relations theories that have arisen in the past thirty years . . . have all missed an important dimension, i.e., the study of processes in the international system and of relational complexity in international society'. He draws on sociologist Fei Xiaotong, who famously argues that 'Western society' is based on independent individuality, like bundles of rice straw tied together by social contract and institutions. 'Chinese society' is instead like the continuous circles that ripple outwards from a pebble dropped on the surface of water. The ripples spread social relations and each circle is connected in one way or another (Fei, 2007;see Qin, 2009, p. 6).
Qin emphasizes friendship 'guanxi' (关系), in his development of Chinese relational ontology. In contrast to sociological accounts which focus on guanxi's role is support networks and welfare provision, or its Reintroducing friendship to international relations 383 role in diplomacy in creating friendly 'feeling' (ganqing 感情) between people, which can help them develop a guanxi relationship (Brady, 2003, p. 15), Qin (2005a; argues that guanxi should form the 'hard core' of a Chinese IR theory. He focuses on guanxi as an ontological assumption of IR which differs from 'Western' interpretations, as embodied in theories like structural Realism, Neo-Liberal institutionalism and structural constructivism (Qin, 2005b;2009, p. 5). Qin argues that taking relations as the focal point of IR steers away from understandings of relations between states that start with state units or individuals and conceive of their relations as secondary. In Qin's view, a reliance on guanxi means Chinese people have a distinct and geoculturally determined way of thinking about relations between people, which is different from Western thinking. Whereas in European social science 'rationality became the dominant word' in Chinese thought the counterpart is 'relationality' (Qin, 2009, p. 5).
In his account of guanxi relationality, Qin takes processes and agents to be symbiotic and 'inter-constitutive' in an intermingled practice of socialization. There can be no one-way causality between the two, because neither precedes the other and neither is external to the other (Qin, 2009, p. 9;2016, p. 39). Qin illustrates these relations through the yin-yang symbol that is common to explaining Daoist thought (see Figure 1). This symbol consists of a black and a white half that together form a 'harmonious and holistic' circle (Qin, 2009, p. 9;2016, p. 39). The circle does not exist without the halves; the halves cannot form a shape without the circle. In this way, '[y]in, yang, and the circle are in and of one simultaneously' (Qin, 2009, p. 9). The relationship between agents and process must therefore be interpreted in terms of circular constitution, rather than linear causality. Qin further explains this in terms of what he calls a 'Chinese dialectic' of change and inclusiveness, which he contrasts with a 'Hegelian dialectic' (Qin, 2009, p. 10;2016, p. 39). In his view, the 'Western way of thinking' focuses on the independent entity and tends to assume discreteness (Qin, 2009, p. 10;2016, p. 39). On its dichotomizing understanding, A can never be non-A, because the two have essentially different properties. In contrast, in Qin's Chinese dialectic, A can be non-A or includes non-A; it is inclusive and puts emphasis on change. On this understanding, there can be no social actors that pre-exist social relations and process. The process of relationships moreover transform both the behaviour and the essential properties of actors involved. A can transform non-A or be transformed into non-A. In contrast to Qin's reading of the Hegelian dialectic where thesis clashes with anti-thesis, in Chinese dialectics thesis and antithesis complement one another to make a harmonious whole. In this way, yin-yang relationality, 'denies the dichotomously structured concept of "thesis vs. anti-thesis" or "us vs. them"' (Qin, 2016, p. 40).
In recent writing Qin highlights the importance of friendship as a neglected kind of relationship in IR, with reference to its theorization by Felix Berensk€ otter (Qin, 2016, p. 37). Although Qin does not discuss Berensk€ otter other than to underline the importance of friendship, dwelling on it here can bring out the distinct contribution of relational ontologies that draw on Chinese thought to the broader discussion of this topic.

Berensk€ otter: friendship and anxiety
Berensk€ otter, like Qin, sees limitations in an ontology that assumes 'the individual (state) as an autonomy-seeking entity' common to mainstream Western IR (Berensk€ otter, 2007, p. 653). Berensk€ otter draws on Heidegger to advocate an 'evolutionary ontology of the state as something which is neither static nor ever complete but a work in progress, something always in the process of becoming' (Berensk€ otter, 2007, p. 655). Berensk€ otter draws on feminism, post-structuralism and other European thought to propose that states are not primarily concerned about other states that threaten their survival (as Realists, for example, might have it), but rather about uncertainty as such (Berensk€ otter, 2007, p. 655). Anxiety about uncertainty and incompleteness provides 'the foundational sentiment defining the human condition' Reintroducing friendship to international relations 385 (Berensk€ otter, 2007, p. 655). It is because of such fundamental anxiety that people and states are said to 'look for what Anthony Giddens calls "anxiety-controlling mechanisms" employed to gain "ontological security", or a stable sense of Self' (Berensk€ otter, 2007, p. 656). States, in his view, seek friendship to control anxiety (Berensk€ otter, 2007, p. 656). Berensk€ otter bases his understanding of friendship on that of Aristotle, and highlights a number of features that we suggest Aristotle's account has in common with that of Confucius': that true friends share a common goal of virtue, that such virtue is obtainable primarily through activity with virtuous friends, that this process is what can lead to harmony (Berensk€ otter, 2007, p. 664-68).
However, Berensk€ otter's friendship is not simply rooted in a sense of group-membership or identification with humanity in general. Part of friendship's ability to control anxiety stems from its capacity to 'sustain the individual's sense of self by treating [it] particularistically' (Berensk€ otter, 2007, p. 664). A similar concern with particularity or alterity has led philosophers such as Derrida to worry that Aristotle's understanding of friendship as an extension of self-love collapses the differentiation between Self and Other, and therefore negates the very possibility of a friendship relation in the first place (Berensk€ otter, 2007, p. 667;Derrida, 1997, p. 11). Here, the worry is that understanding the friend as 'another Self', as Aristotle does, makes it a narcissistic extension of that Self, and therefore treats it as derivative. In this sense, the hierarchy of the 'total' construct obliterates or ignores difference. If we follow Derrida and understand politics to be made possible by a plurality of being, or better perhaps of becoming, this merging of selves would make friendship apolitical. Therefore, Berensk€ otter notes, 'in order to conceptualise friendship as a relationaship in which politics occurs, friendship relations must allow for heterogeneity and be conceived "through a philosophy of difference so as to be rendered politically relevant"' (Berensk€ otter, 2007, p.668;Chiba, 1995, p. 522).
The tendency to construct difference negatively that Derrida identifies in Aristotle's friendship is sometimes claimed to be a wider feature of 'Western' language or thought. Some have also claimed that it is Derrida's deconstructive reading of Western thought that creates the hierarchical and totalizing understanding of difference (Massey, 2005, p. 49-54). Berensk€ otter's reading of friendship, in Aristotle and in IR more generally, also centres on concerns with the individual's sense of particularity and self and the anxiety that stems from it. He is not alone in such a focus, which regularly occurs in literatures interested in ontological security, including those that focus on China (Gustavsson, 2016;Pan, 2014, pp. 455-456). In contrast to those who understand this type of ontological anxiety to be a feature of Western thought, Shih Chih-yu argues that one of the things that distinguishes the Chinese relational turn in IR, is that it concerns anxiety rather than passion. He sees in both 'China' and 'the West' a 'general feeling of anxiety ingrained in relationality', which can be calmed when positive feeling in relationships provide individualized and mutually assuring recognition (Shih, 2016).

Ling: the intimacy of 'self and other', and the possibility of multiple worlds
So far, we have seen that relationality and 'friendship' have played a key role in both Qin's and Berensk€ otter's thinking about the nature of IR. In particular, both link relationality to questions of the co-construction and dependence of Self and Other. Berensk€ otter seeks to show how friendship can allay a state's anxiety as it seeks recognition of its self from others. Qin develops his notion of relationality through Daoist yinyang dialectics, which he renationalizes as 'Chinese' dialectics. For him the issue is not so much reconciling Self and Other, but realizing that Self and Other are inter-constitutive. Thus, although Berensk€ otter and Qin agree about the co-constitution of Self and Other, they place different importance on the origin and permanence of antagonism between Self and Other. For Berensk€ otter anxiety appears to be a permanent ontological fact of Self-Other relations. It can be tamed, but not eradicated. In contrast, Qin sees no such tension as ontologically foundational. However, this also indicates another contrast. Berensk€ otter's Heideggerian Self and Other have the potential to be radically different, whereas Qin's Wendtian foundations attenuates difference.
A similar understanding of yin-yang relationality is also at the root of L. H. M. Ling's 'worldism'. Ling critiques what she calls 'Westphalia World', the common understanding or hegemonic vision of IR that includes the 'mainstream approaches' that we have shown to be characterized by an ontology of things. Ling shows how this view and its ontology 'perpetrates a profound violence' by denying its reliance on those it excludes, as well as their knowledges and ways of Reintroducing friendship to international relations 387 knowing, what she calls 'Multiple Worlds'. Ling also reacts against Wendt's claim that first encounters like those between Cort es and Moctezuma led to an accretion of culture at the systemic level, leaving the enemy, the rival and the friend as the only roles available to others, locking out any other considerations of relations among worlds (Ling, 2014, p. 30). As a result of such an ontology of things, Ling describes how a '"postcolonial anxiety" festers in Multiple Worlds that, in turn, aggravates a "colonial anxiety" in Westphalia World' (Ling, 2014, p. 3). This leads to a nihilistic logic where the lives of others need to be forfeited in order to save one's own.
Ling offers an alternative to that violent and anxious worldview in the form of a Daoist dialectic similar to Qin's. Ling writes that in such a Daoist dialectic the 'complementarities (yin) prevail despite the contradictions (yang) between and within the polarities. Nothing remains static or the same' (Ling, 2014, p. 15). This worldview strives to recentre contributions to world politics that have been marginalized from it, and to conceptualize these as having ontological parity with Westphalia world. It is thus a response to the negative spiral of violence and anxiety in the relation between Westphalia World and Multiple Worlds: A dao of world politics propels us from this dilemma. In recognizing the ontological parity of things, a post-Westphalian IR experiences the constant potential of creative transformations due to the mutual interactions that transpire, especially between opposites. Multiplicity and difference manifest, enacted by local agents and their transformations of knowledge (Ling, 2014, p. 3).
In this way, and in contrast to Wendt's account of relations which falls back on an ontology of things, this worldview emphasizes a recognition of the complexity of the self, which includes traces or elements of the other in the self. Intimacy, rather than autonomy, marks its condition (Ling, 2014, p. 12). However, the point is not to replace Westphalia World with Multiple Worlds, just as the point is not for the ontology of things to be superseded by relational ontologies. Rather, the Daoist dialectic urges us to move closer towards balance and engagement. In Ling's terms, '[f]ortified with Daoist dialectics, worldism re-visibilizes Multiple Worlds in relation to one another as well as to Westphalia World' (Ling, 2014, p. 18). Equally, we might say that it makes multiple relational ontologies visible again, both in relation to one another and in relation to the ontology of things. The area of intersection between different ontologies forms a dialogical space. However, unlike the Socratic dialogue on friendship, Daoist dialectics do not presuppose that there is a stable and discoverable truth independent of human perspectives (Ling, 2014, p. 66).
Ling's insistence on this contrapuntuality between West and Rest, Self and Other, 'to jointly produce the complicities that endure despite and sometimes because of the mutual conflicts that tear them apart' adds an important emphasis to Qin's account (Ling, 2014, p. 45). Qin is clearly aware that the relational ontology he advocates is not uniquely Chinese, that it has both ancient and contemporary parallels in Europe and elsewhere. Our previous discussions of friendship have highlighted further commonalities between Confucian and ancient Greek understandings of friendship in their focus on friendship as a relationship of learning for the purpose of developing virtue. Ling's efforts to articulate her 'Multiple Worlds' without falling back on dichotomizations of 'the West' and 'China' (or 'the Rest') helps to further underscore that the 'Chinese view' that Qin describes need not be exoticized as a geo-culturally specific example. On the contrary, it might even be contemporary IR theories which have assumed an ontology of things and marginalized friendship relations that should be considered to be a highly specific exception to the more general global and historical trend.
Furthermore, the move away from a focus on individuality in Qin's yin-yang processual constructivism and Ling's Daoist dialectic decentres the prior focus on anxiety. We are not suggesting that Berensk€ otter, Shih and others are wrong in observing existing anxiety. We are, however, suggesting that these emotions are as constructed as the relations that are said to provoke and sooth them. In Qin's guanxi relations, affect and emotion have an important role, but do so in terms of 'collective emotion', rather than in terms of the anxiety that resides within the self (Qin, 2009, p. 12). On a similar note, Ling's Daoist dialectic of multiple worlds is offered as a 'social ontology', 'a vision, an understanding, a state of being to treat and put into remission this "postcolonial anxiety"' (Ling, 2014, pp. 31-32). We may all be in a process of becoming, but there would be no reason to be anxious about this if we never attached ourselves to an ontology of things or of being in the first place. To Berensk€ otter, to reach harmony means to 'tame anxiety', and so friends matter because they can help us provide some sense of ontological security (Berensk€ otter, 2007, p. 666). Shih sees a similar role for friendships or 'non-competitive relationships' drawing on Chinese tradition (Shih, 2016). On at least one reading of the yinyang dialectic, harmony is not the opposite of anxiety. Granted, harmony depends on our ability to manage relationships in a way that mediates disagreement, but this process as described by Qin and Ling is very different from that of taming the anxious self.

Conclusion: rediscovering friendship in IR
This article started with claims that China will be a new and friendlier kind of great power, because it relies on a Chinese relational ontology instead of a Western ontology of things. Our focus has been on friendship as a component of relational ontologies in the theories of IR. We have suggested that it is a mistake to essentialize or exoticize relational ontologies as being specifically Chinese; the predominance of an ontology of things in IR may be the exception rather than the rule in global and historical perspective. Chinese thought is a useful reminder to scholars of resources and ways of thinking that contemporary views of politics and IR either ignore or neglect. Chinese thought on friendly relations can make a distinct contribution to disciplinary efforts to develop relational ontologies. Relational ontologies are an essential counterpart to the 'ontology of things' which is so foundational in much contemporary IR. Chinese relational ontologies suggest that understanding the co-constitution of Self and Other is both necessary and useful if we are to have a fuller understanding of international politics. Moreover, Chinese thought shows that the relations between Self and Other need not be conflictual or colonial. On the contrary, they can be the basis for a dynamic of interdependent growth and change.
Thus, the relational ontologies considered in this article provide an alternative starting point for understanding China's friendliness in IR, which differs from common accounts in Chinese official, academic, and diplomatic discourse. Through them, we come to see how China's friendliness and 'peaceful rise' will not depend on the autonomous actions of some imagined independent 'China', or on some essential characteristics of the Chinese people, nation or state. Nor does it require conformity, integration or socialization into an imagined 'international society' (cf. Ling, 2014, p. 91). Most importantly, scholars do not have to assume that IR is built-up of state-units that are made anxious by the incompleteness and change that is indicated by the presence of others. Nor do they have to assume that the only possible significance of friendship is to sooth the anxious self. Instead, friendship can be understood as that which creates and maintains our continuous becoming with others, and the ontological parity of multiple worlds.
That we have found contemporary discussions based on Chinese epistemic legacies to add to the debate in constructive ways does not indicate a necessary link to subject positions designated as 'Chinese'. Specifically, much policy and discourse of the Chinese state is wedded to 'Westphalia world', as expressed in its fixation on territorial sovereignty, the claim that others should not voice opinions about China's 'internal affairs', the demand that those who are considered 'insiders' be patriotically loyal to the Party-state, and strong attachment to the 'ontology of things' more generally. We want to be clear, therefore, that the Daoist political imaginary suggested here offers a new vocabulary for those who want to think differently about relationality and anxiety in IR, regardless of whether they are speaking from China, about China, or neither. This is not 'how Chinese people think, feel and behave', it is how people could think, feel, and behave. It offers IR a different starting point to what many of us are used to, with the potential to help us know our worlds in a different way and to produce knowledge about those worlds in a different way. It offers one possibility, which does not exclude or denigrate other possibilities for thinking world politics.
Some critics will object that while it might be a better world if state identities were accepted as insecure and both elites and the public were to see each other as parts of a whole, this seems unlikely. Indeed, since the Western world does largely accept the ontology of things, it remains unclear how the view espoused here could come to fruition without a fundamental change in world view. Such critics speak from what Robert Cox famously referred to as problem-solving theory, which 'takes the world as it finds it, with the prevailing social and power relationships and the institutions into which they are organised, as the given framework for action' (Cox, 1981, pp. 128-129). To these critics we want to suggest that now commonplace ideas like democracy, human rights, the abolition of slavery, or gender equality were all criticized as unlikely to be implemented at some point in history. Nonetheless, they continued to be developed as alternative vocabularies and ideas to those that dominated intellectual discourse at the time of their emergence. The dominance of one set of ideas in a particular society or system is not a good reason to refrain from exploring alternative ways of thinking and doing world politics. Quite the opposite, as critical academics we should explore what may be made possible by mobilizing alternative vocabularies, ideas, and traditions of thought, even if we do not make hubristic claims that our writing alone will transform the world.
It is as a possible alternative starting point for thinking that the concern of friendship is useful. As we have argued, friendship should not be understood as simply denoting a concern with the personal and private. In fact, its historical and cultural usage is far broader and more complex than that. Friendship is useful to IR insofar as it helps us to refocus on relations, and to conceive of those relations as a constitutive dynamism of Self with Other. Bringing friendship back to parity can help us grapple with something that is lost when we focus on enmity, conflict, war and disjuncture: what it means to 'become' in relation with Others. This is different from discussions of agreements and alliances, or understandings of international community that still see peoples and states as discrete entities based on an 'ontology of things'. Chinese thought thus reintroduces an ontology of relations to the West and to the discipline of IR, and acts as a reminder of what has been forgotten. Such a meeting does not reinforce the supposed differences between China and the West, but acts as a reminder that they are part of creating each other. Whilst an ontology of things tends to cast friendship as a conflictual 'Us and Them', the relational ontology of contemporary Chinese debates on friendship can contribute to such debates by viewing other possibilities for Self and Other. Friendship achieves this through its focus on the relationship of the friends, and the way in which this relationship is formed by, and forms, both Self and Other. This ontology reopens the possibilities for friendship as a way of conceptualizing Self with Other, rather than Self in contrast to Other.