Multiplicity and the problem of ‘society’

Abstract

This article seeks to address a major shortcoming of the idea of Societal Multiplicity in international theory: its lack of a worked-out definition of what counts as ‘a society’. The article supplies this definition in four steps. First, it reviews some difficulties in defining ‘society’, arguing that these arise in part from the historical separation of social and international theory. Second, it interweaves resources from both traditions to argue: (a) that societal multiplicity reflects the specifically political separation of human populations; (b) that this separation arises in part due to centripetal properties of the political as a feature of social life; and (c) that an individual ‘society’ is best defined as a unit of the resultant multiplicity: a social formation that exists simultaneously as a geopolitical sub-division of the social world. Third, the article defends this definition against a series of objections – that it is too general, that it reifies the state, that it is methodologically nationalist and that it suppresses historical variety. And finally it inserts the definition into the idea of Societal Multiplicity in order to test how far, and with what consequences in International Relations and beyond, this closes the definitional gap left by the original formulation of that idea.

What is [society]? If no-one asks me, I know: if I wish to explain it to one that asketh, I know not . . . [After Augustine, 1907: 262]

The idea of Societal Multiplicity seeks to bring into view the significance of ‘the international’ right across the field of human experience (Rosenberg, 2016). ¹ International Relations, it claims, amount to more than the specialised domain of specifically inter-state politics, trans-border economics and inter-governmental organisations. Rather, they are a framing condition for all branches of social life – from politics and economics through science and technology, religious and other belief systems, artistic and literary production, language, fashion and even cooking. All these domains of practice exist in the societal plural, with their multiple instances both differing from and yet also chronically interacting with each other, giving the international a much broader and deeper existence than is conventionally assumed. This also suggests that the academic subject of International Relations (IR) is potentially much more than a passive ‘inter-discipline’ (Jackson, 2017, emphasis added) where other, more established disciplines meet, a mere sub-discipline destined to remain trapped within ‘the prison of political science’ or, worst of all, a failing ’sub-field’ that, due to its intellectual underperformance, scholars are tempted to ‘leave behind’ altogether (Reiter, 2015, emphasis added). Quite the opposite: the fact that all social life occurs within this context of societal multiplicity means that there is an ‘international of everything’ (Rosenberg and Tallis, 2022), just as there is a Politics, an Economics, a Sociology and a History of everything. And while each of these other disciplines elaborates a distinctive method of analysis, IR too can derive an analytical framework of its own from the five generic implications that multiplicity has for the nature of social reality: co-existence, difference, interaction, combination and dialectical change (Rosenberg, 2016: 135–141). IR’s potential – and perhaps one day its cohering identity – is to be the discipline which reveals and explores these ‘consequences of multiplicity’ in their many different manifestations across the subject matter of the social sciences and humanities. In short, behind the inter-state multiplicity that is generally taken to comprise the core subject matter of IR lies the deeper and broader multiplicity of interacting societies.

Whatever its aspirations, however, the Multiplicity research programme has also, from the start, been dogged by a fundamental question which remains largely unanswered: ‘what counts as a society?’ (Wæver, 2023: 1). The question was first posed by George Lawson in 2015 in a referee’s report on the earliest published statement of the Multiplicity thesis. It was later raised again with reference to historic or prehistoric collective actors which exercised significant agency with respect to other groups without themselves being recognisably societal entities (Powel, 2018, 2020); Meanwhile, critics have seen in the undefined idea of ‘societal’ multiplicity a veiled attempt to drag the discipline of IR, so recently liberated from the tyranny of political realism, back into the confines of a state-centric and Eurocentric ‘straitjacket’ (Blaney and Tickner, 2017; Drieschová, 2019). Other contributors to the debate have switched away from the specific significance of the multiplicity of societies to focus instead on multiple cultures (Wiener, 2022), multiple types of actors (Spalińska, 2023) and even multiple biological species (Kurki, 2020). Seven years into the programme, there was still arguably only one organised attempt to provide a general definition of ‘society’ – Viacheslav Morozov’s (2021, 2024) ingenious discussion of ‘hegemonic formations’; and that arguably produced a multi-scalar model in which the distinctive properties of the international scale are no longer visible (Wæver, 2023). Meanwhile, others (notably Rosenberg and Tallis, 2022) continued to argue that the historical variety of societal formations was simply too wide to support a single definition.

This was the context in which Ole Wæver forcefully re-posed the question of ‘society’ on three main grounds. First, the concept of Societal Multiplicity willy-nilly presupposed some notion of the entities whose quantitative plurality composes this condition. And it must therefore be possible in principle to draw out this presupposition and give it a clear statement. Second, in the absence of such a clarification, the potential value of the research programme would continue to dissipate as participants took it as a licence to explore any and every heterogeneity in the human and natural worlds. ‘Strict multiplicity’, by resisting the necessary next step in its own self-clarification, would end by being submerged in a flood of ‘mushy multiplicity’ which was no longer focused upon the international per se (Wæver, 2024). Third, defining society was an unavoidable challenge because it remained to be seen whether such an exercise could avoid tipping over into one of two familiar outcomes: either a proto-realist assimilation of society to the state or a proto-liberal collapse into an indeterminate pluralism. Either of these would spell failure for the research programme; but without a definition of some kind, the outlines of its supposed third alternative could not be fully seen. ²

The purpose of this paper is to supply the missing definition of society – a definition which shows how societal multiplicity forms the underlying substance of International Relations, while avoiding the pitfalls of both realist statism and liberal pluralism. The argument is set out in four steps. First, we shall briefly review some difficulties involved in defining society; and we shall identify key requirements that our own definition will need to fulfil. Second, we shall try to find our way to a definition that meets these requirements. Three clues will mark our route: Adam Watson’s classification of different international systems in terms of the varying degrees of political separateness among their members; Aristotle’s (and others’) identification of the unique properties of the political as a type of human association; and Adam Ferguson’s argument about the active role of societal multiplicity in the constitution of the political per se. Together, these clues will bring us to what might at first appear to be a disappointingly obvious (even banal) definition: societies, we will say, are social formations that also exist as geopolitical sub-divisions of the human world. ³ What exactly these sub-divisions are – the form they take and even the degree of their separateness – will vary across time and space. They are historical: we have to look and see. But they will always be found in some form because human existence has always been multiple. This multiplicity, with its many-splendored variety, is the ‘international’ dimension of social reality. And ‘societies’, whatever their form, are the constituent units thereof. This is a definition that will need immediately (in step three) to be defended against a series of objections. By then, however, the route we have taken will have revealed that there is both more and less to ‘geopolitical sub-division’ than meets the eye. And in that further calibration may lie a key part of our solution to the challenge of definition. To confirm whether this is so, step four inserts our definition directly into the theory of Societal Multiplicity. Does it fit? And is it really the definition that has been implicit in the theory all along? To find out, we shall revisit all five of the ‘consequences of multiplicity’ which, according to the original theory, compose the international as a dimension of the social world. We shall examine whether these consequences proceed logically from the nature of society as we have now defined it. If so, we will have arrived at the missing definition which in fact underpins the theory of Societal Multiplicity. And the gap that has threatened this theory from the outset will finally have been closed.

Society – the challenge of definition

Numerous writers have wrestled with the question of defining society, but they have often come away with underwhelming results. Consider, for example, Michael Mann’s initial attempt in The Sources of Social Power. Mann (1986) was so determined to overcome the anachronistic and misleading view of a society as a ‘bounded totality’ that he actually declared: ‘It may seem an odd position for a sociologist to adopt; but if I could, I would abolish the concept of “society” altogether’ (pp. 1–2). It was indeed odd, and he recognised that it could not be done. So he tried instead to produce a definition which avoids the ‘bounded totality’ problem:

a society is a network of social interaction at the boundaries of which is a certain level of interaction cleavage between it and its environment. (p. 13)

Mann himself was pleased with his formula, immediately adding that ‘[f]ew historians, sociologists, or anthropologists would contest this definition . . .’. But there was something peculiar about it: although he was trying to avoid boundedness, the only thing in his definition that distinguished a society from other networks of social interaction was – of all things – that it had a boundary where there was some kind of separation, which he called an ‘interaction cleavage’. Thus, he apparently redefined society in terms of the very characteristic he had sought to exclude. Something similar happened with Anthony Giddens (1984) in The Constitution of Society:

‘Societies’ . . . are social systems which ‘stand out’ in bas-relief from a background of a range of other systemic relationships in which they are embedded. They stand out because definite structural principles serve to produce a specifiable overall ‘clustering of institutions’ across time and space. (p. 164)

There it is again: trying to avoid the ‘bounded totality’ idea, but then re-inserting boundedness under a different name – this time not ‘interaction cleavage’ but ‘standing out’, ‘bas-relief’ and ‘clustering’. As we shall see later on, Giddens and Mann did of course go beyond these definitions. But the key point for now is that neither of the definitions in itself tells us why (or even how) societies are spatially finite entities, or which particular social relations it is that produce or undergo the ‘interaction cleavage’ at the boundary. Exactly what is it, then, that makes a society a society? And how, correspondingly, does this relate to our understanding of the international as societal multiplicity?

The answers are not obvious: ‘no basic agreement exists in sociology about perhaps its most basic concept – the very idea of society itself’ (Hamilton, 1986: 7). Wallerstein cites the Oxford English Dictionary as listing 12 ‘principal meanings’ of the word (Wallerstein, 2001: 64). For our purpose, however, three meanings are most relevant. ‘Society’ can refer to the fact of sociation, meaning that humans are social animals and they exist only in society, in interrelation with other humans. As Montesquieu (1989) put it: they are ‘[m]ade for living in society’ (p. 5). This usage of the word, says Wallerstein, occurs only in the singular. Second, society is also used to refer to particular kinds or types of society – such as capitalist society, tribal society, feudal society and so on. Here a multiplicity is recognised, but it exists only in the typological space of comparison. And third, society can also refer to a particular collective entity, an aggregate of individuals who make up a social unit: ancient Athens for example, or mediaeval Japan, or present-day Germany. In this case, as Wallerstein notes, it can take the indefinite article – ‘a society’ – implying that a given instance is one of a number of potentially co-existing entities.

Each of these three meanings, (which we shall refer to below as meanings #1, #2 and #3), yields a fundamental principle of social theory. Meaning #1 gives us the ‘social construction of reality’, the claim that human behaviour is not explained by individual factors alone but always involves relationally generated logics, identities and capacities. Meaning #2 gives us the principle of historical specificity. This is the idea that these logics, identities and capacities are not homogeneous, universal tendencies; they differ according to the particular social structures that obtain in particular times and places, differences which can be explored through comparative historical analysis. And meaning #3 entails the distribution of human existence across a variety of interacting social entities; and that gives us – or should give us – ‘societal multiplicity’: the international condition as a framing context of any given social reality or analysis thereof.

Finally, all three meanings – society as sociation, structure and entity – refer to perennial features of human existence. No one will argue that there was a time when humans were not social creatures. Equally, no one denies either that the structure of societies has varied across time and space, or that social existence has always been distributed interactively across a plurality of finite instances. ⁴

Nevertheless, it is widely recognised that the 19th- and early 20th-century classics of social theory concentrated heavily on the first and second of these meanings, while largely neglecting the third. Charles Tilly (1984) listed the resulting (though tacit) assumption that society was ‘a thing apart’, a self-enclosed entity subject to internal causes alone, as the first of eight ‘Pernicious Postulates’ inherited from ‘ideas built up by 19th-century intellectuals’ (p. 11; 2). Three decades earlier, Ralf Dahrendorf (1958) had identified the same assumption as one of two which condemned the leading social theory of his day to an absurdly ‘utopian’ worldview: ‘it is obvious that such [self-contained] societies do not exist’ (p. 118). Gianfranco Poggi suggested that what would later be termed ‘methodological nationalism’ (Martins, 1974) had resulted in ‘a sort of “learned incapacity” for tackling problems relating to the external dimensions of social phenomena’ (Poggi, 1965: 284). And although Theda Skocpol called explicitly for ‘an inter-societal theory’ in 1973, Zygmunt Bauman (1992) could still lament, nearly 20 years later, that ‘sociology so far is poorly equipped to treat the social space beyond the confines of the nation-state as anything else but the analytically compressed “environment”’ (p. 59).

As this suggests, the classical lacuna has endured into contemporary social theory. Thus, while Giddens was emphatic that societies exist only in complexes of other societies, this crucial insight did not (in his writings) give rise to any particular consequences for the nature of social reality. It did not even make it into the premises of structuration theory in either A Contemporary Critique of Historical Materialism (Giddens, 1981: 26–29) or The Constitution of Society (Giddens, 1984: 281ff). Wallerstein’s bold revisioning of society as world-system emphasises structural inequality and subordination deriving from historically specific forms of social system, rather than anything deriving from the fact of societal multiplicity per se. Meanwhile, Margaret Archer’s magisterial reworking (1980) of social theory in terms of agency, structure and culture simply did not address the inter-societal dimension of social reproduction. Michel Foucault’s account of disciplinary power avoided it too; for, as Nancy Fraser (2009) pointed out, it ‘assumed that disciplinary ordering was nationally bounded’ (p. 121) Globalisation theory rested upon a deep premise that the international was at best a gradually fading reality (Rosenberg, 2000). The work of Niklas Luhmann, in the words of his foremost IR interpreter, ‘completely avoids dealing with . . . international politics (or, for that matter, world politics or international relations’ (Albert, 2019). And finally, in perhaps the most remarkable near miss of all, Gilles Deleuze and Félix Guattari’s A Thousand Plateaus (2013) advanced ‘multiplicity’ itself in the boldest way possible, propounding it (in the form of ‘the rhizome’) as nothing less than a thorough-going alternative principle of thought and being at every level of existence; and yet even this did not bring into focus the international question – that is, the specific implications of the multiplicity of societies. ⁵

But if social theory has struggled to incorporate our third meaning of society into the other two, international theorists have likewise found it a challenge to integrate all three meanings. Kenneth Waltz chose instead to theorise international politics sui generis, famously declaring that ‘students of international politics will do well to concentrate on separate theories of internal and external politics until someone figures out a way to unite them’ (Waltz, 1986: 340). Meanwhile, liberal, Marxist, feminist and postcolonial IR theorists have largely denied that meaning #3 adds fundamental determinations of its own. These writers have generally viewed the international as shaped by the contents of meaning #2 – that is, by democracy, capitalism, patriarchy, racism, Western culture and so on. And outright rejections of ‘the international’ continue to abound in the field. ‘The international’, writes Nicholas Onuf (2023), for example, ‘is a sentimental allusion to an undisciplined subject matter’ (p. 2).

The result of this impasse is that at the deepest level – the level of ontological premises – social and international theory have been severed from each other. This disconnection violates the first principle mentioned above: namely that the human world – all of it, including the geopolitical realm hived off by realism and spurned by realism’s critics – is socially produced.

Thus, we can see that any definition of society produced in response to Wæver’s critique must not only overcome the ‘boundedness’ dilemma encountered by Mann and Giddens. It must also reunite in itself all three meanings of the term discussed above. This is the challenge of definition which haunts the concept of ‘society’ whenever it is viewed in the context of the international. It implies that the stakes of our definitional exercise are much larger than the travails of the theory of Societal Multiplicity alone. And it suggests several criteria that our definition must meet.

First, our definition must be consistent with meaning #1, the principle of sociation, so that societal multiplicity becomes visible as continuous with the nature of social reality in general, rather than being treated (as in neo-realism) as a theoretically separate object domain. Second, however, (and in order to avoid tipping over into a ‘reductionist’ liberal perspective), it must also enable us to see the international as a specific dimension of this social reality, with constitutive and causal properties of its own. Third, it must nonetheless incorporate (and survive the implications of) meaning #2 – namely that societies only ever exist in particular historical forms, forms that vary widely across time and space. And finally, our definition must enable us to derive from within it the last of the meanings we have described: meaning #3. It must visibly entail the quantitative multiplicity of societal entities, rather than treating this as a purely empirical happenstance that would still allow us to define society implicitly in the ontological singular. Only in this way can it be adequate to the transhistorical claim made by Societal Multiplicity, namely that the human world has always comprised a plurality of social entities, and that this fact defines a subject matter for IR with deep and theorisable implications for all the social sciences and humanities.

Are there also criteria we do not need to meet? Apparently so. Our definition of society, for example, does not need to be original. The purpose of this caveat is not to set an artificially low bar for the argument. It is to keep our attention on the main goal: namely, to establish that the theory of Societal Multiplicity, with its dramatic implications for IR’s place among the human disciplines, can be grounded in a coherent definition of its constituent elements. This main goal implies a further limit to our exercise too. Excavating the particular definition of society buried within the concept of Societal Multiplicity is not quite the same as providing a general-purpose conceptualisation of society per se. After all, the findings we unearth will inevitably be oriented towards society’s interactive dimension. However, insofar as that dimension is a necessary (and too often neglected) part of any general conceptualisation, our more limited labour may still contribute to that wider ambition too. How much it can contribute will depend on just how important the international turns out to be. ⁶

Defining society

In The Evolution of International Society, Adam Watson provides our first clue, taking us an initial step beyond the point reached by Mann and Giddens. All international systems, he says, may be located on a spectrum which measures the degree of separateness among the entities involved (Watson, 1992: 13ff). ⁷ At one end of the spectrum is the hypothetical condition which he calls ‘absolute empire’. In this condition – hypothetical because it lies beyond the vanishing point of multiple entities per se – the units involved have lost all authority to a central power, with regard to both their internal laws and their external relations. The erstwhile units are thus units no longer, and the relations among them are not inter-societal but rather intra-societal. At the other end of the spectrum lies the opposite condition: ‘absolute independence’. Here the units are completely separate and entirely free to be and act as they please. This condition too is hypothetical, this time because it is not possible for societal entities to co-exist in space and time without that co-existence affecting their range of choice in both positive and negative ways. As Hugo Grotius put it, ‘[t]here is no State so powerful that it may not some time need the help of others outside itself, either for purposes of trade, or even to ward off the forces of many foreign nations united against it’ (Grotius, 2002: 330).

With both extremes of the spectrum ruled out, all international systems in history must fall somewhere between the two. Watson subdivides his spectrum into a number of broad conditions which shade into each other at their margins, hence forming a continuum that includes every possible increment of integration and separation: from (non-absolute) empire, through dominion, suzerainty, hegemony and on to sovereignty and (non-absolute) independence. He also notes that particular international systems occupy different positions on the spectrum at different times, oscillating between greater and lesser separateness among their members. And we may further add that no international system is internally uniform: at any given point in time, the degree of separateness varies among its members too. Nonetheless, a social system which cannot be placed on this spectrum (in however qualified and provisional way) is not an international or inter-societal system at all. And its parts are therefore not societies in the sense that we are discussing them: they are not the elements of a condition of societal multiplicity.

Suppose, then, that we call the range of conditions between the extremes ‘multiplicity’, and we name as ‘societies’ the parts whose separateness is being measured on the spectrum. In this case, what are we saying that societies actually are? If we examine the increments, (dominion, suzerainty, hegemony, etc.), we can see that what the spectrum indicates is the extent to which de facto authority over a population’s own affairs lies within the population itself – where ‘affairs’ denotes both the internal relations that constitute it in a given form, and the external orientation that comprises its relations to other such populations. Correspondingly, we are also measuring the extent to which this population recognises de facto authorities or powers beyond or outside itself. Thus, the variable which allows us to identify the existence of a ‘society’ is the same that Watson used to identify an international condition, but this time viewed from the perspective of the parts: it is political separateness, and the degree thereof.

But what is political separateness, as opposed to other forms of separateness – say, economic, cultural, linguistic and so on? Here, en route to our general definition of society, we find (not for the first time (Rosenberg, 2000: 77)) that we need first to deploy a general abstraction of ‘the political’, a definition which captures the nature of this dimension without presupposing any of its myriad historical forms. This step Watson has already taken, in very broad terms: ‘the political’ concerns authority or government, and its distribution in relation to given entities and populations. Let us look a little further into this.

In order for any human group to endure over time, it must have some way of constituting and reproducing itself as such. That is to say, it must have some means, however informal, of coming to collective decisions and giving effect to those decisions – whether they relate to the inner structure of the group or to its relations with other groups (Rosenberg, 2000: 78). These means constitute and reproduce the group as both agent and social structure.

But these points apply to all forms of sustained association, including families, clubs, corporations, religions and so on. And it is surely not these that delineate the ‘societies’ we are looking for. They do not make entire populations politically separate from others. ⁸

And there, perhaps, is our second clue: entire populations. In his Politics, Aristotle notes that among the many kinds of association which structure human existence there is one which ‘embraces all the rest’. While all those others have as their object some specialised goal – the pursuit of wealth, knowledge, friendship, beauty and so on – this superordinate association originates, he says, ‘in the bare needs of life, and [continues] in existence for the sake of the good life’. And since the ‘bare needs’ include both the ability to uphold the laws within and ‘to meet dangers coming from without’ (Aristotle, 1942: II.7, 23), this special association also delivers a precondition for all the other forms of association to endure and prosper (p. I.2, 27–30). Without saying it in so many words, Aristotle is implying that ‘society’ in our first two meanings (sociation and structure) must also include a special form of binding association which, as we shall see, leads on to our third meaning: society as entity. This entity includes all the associations among its members, as well as the members themselves. It therefore comprises a unique articulation of social relations embodied in a particular population organised in a particular way, inhabiting a particular territory at a given point in time. In short, a society in this third sense is always a concrete social formation.

In Marxist thought, ‘social formation’ is often counterposed to ‘mode of production’ (see Anderson, 1980: 67ff). The latter provides an abstract mapping of any given type of property relations (e.g. tribal, feudal, capitalist). ‘Social formation’, meanwhile, designates concrete historical instances of a mode (or modes) of production in their complex empirical settings, interwoven with other aspects of social life to produce entities at a given point in their historical existence. Such entities may include societies (e.g. Elizabethan England, Classical Athens, Wilhelmine Germany). But the term may also be used to analyse larger complexes (Ancient Greece, Early Modern Europe, the late 19th-century capitalist world economy) and smaller ones (e.g. Shakespeare’s London, 17th-century New Spain). The key point is that a social formation is a unique concrete combination of different social structures that go to make up an entity in its actual historical existence. It can therefore be used to correct not only for the economic reductionism of orthodox Marxism but also, as here, for the political reductionism of neo-realist theory.

Both corrections are needed, for the ‘bare needs’ of a social formation are not exhausted by purely socio-political requirements. ‘Every child knows’, wrote Marx (1934), ‘that a country which ceased to work, I will not say for a year, but for a few weeks, would die’ (p. 246). This need for a metabolic interaction with non-human nature means that societies are necessarily articulated spatially too. The interaction may be carried on via direct occupation of productive land or via mercantile (or imperial) connection to ‘ghost acres’ elsewhere. But even in extreme cases of the latter, where tiny city-states (or enormous empires) orchestrate trade networks far beyond their borders, societies are tethered to particular spatial configurations which, at any given point in time, form an elemental condition of their reproduction. They are territorial, as well as demographic and social structural, entities.

For Aristotle, the association which ‘embraces all the rest’ was of course the polis, from which our word ‘political’ derives. In the Greek world of his day, this was the formal means of self-legislation and self-defence employed by a given population which was simultaneously involved in all manner of other forms of association for other purposes.

However, the polis was a kind of city-state. And because we are looking for a general abstraction of the political that would apply to non-state societies too, we have to take a further step back and ask: what more general feature of human social life was being accomplished in Greece through the particular historical form of the polis? If we strip away that particular form, we are left with our earlier observation: social existence requires shared norms, understandings and practices; but it also requires the means of upholding these norms and practices for a given finite population – means which range from decentralised social sanctions all the way through to organised physical coercion. ⁹ Any population that separates itself from others must arrogate to itself these means and be able to deploy them against both dissolution and attack. Once again, we see that through this political dimension a population both stabilises its reproduction as a definite social structure and constitutes itself as a collective agent. Because this involves the mobilisation of collective capacities, the political dimension turns societies into concentrations of social power. And we can also see that the political, even if it oversees and upholds deep internal inequalities among its members, is uniquely involved in the survival of the community as such, in its given form, and hence it is organically linked also to the collective identity of the group and its members. ¹⁰

For Emile Durkheim, membership of a society demarcated in this way provides individuals with participation in a system of meanings and capacities which are bequeathed to them at birth and which endure beyond their deaths. Consistently with meaning #1 (sociation), he argues that religion itself is nothing other than the invisible group relations of society hypostasised into (and worshipped as) a transcendent reality. It is ‘the system of symbols by means of which society becomes conscious of itself’ (cited in Larkins, 1994: 245). This is clear even if we move on to meaning #2, and consider the modern secular nation-state. The national flag which is the object of such extreme veneration, is not simply a piece of coloured cloth. It is a sacred totem of the political community (Durkheim, 1995: 221–222). ‘Man’, says Durkheim (1972), ‘is only human because he is socialised’ (p. 232). And because individuals are necessarily socialised into particular historical societies, the ‘social sentiments’ which arise from the individuals’ connection to these wholes ‘manifest themselves primarily in the relationships of the society with other societies, and could be called inter-social’ (Durkheim, 1972: 219). Here Durkheim has even invoked (though without deriving) meaning #3: societal multiplicity.

Max Weber provides a further insight into the properties of the political as a dimension of social formations. For him, the exercise of legitimate violence (again, both internally and externally) invests the identity derived from political community with unique, existential associations with death. Whether as punishment for violating the internal norms of the group, or as ‘self-sacrifice’ in conflict with outsiders, violence, which never disappears as a threat to personal and collective security, is tamed by being made to serve the preservation of the society. This connection is strongest in the case of war. But the point is not simply that war ‘creates a pathos and a sentiment of community’ in some general sense. Rather,

. . . war does something to the warrior which, in its concrete meaning, is unique: it makes him experience a consecrated meaning of death . . . Death on the field of battle differs from death that is only man’s common lot. . . . [because] in war, and in this massiveness only in war, the individual can believe that he knows he is dying ‘for’ something . . . This location of death within a series of meaningful and consecrated events ultimately lies at the base of all endeavours to support the autonomous dignity of the polity resting on force. (Weber, 1970: 335)

As Benedict Anderson (1983) would note again several decades later, it is this transformation of fatality into meaning, in relation to ultimate questions of identity, violence and death, that makes political community – in modern times ‘the nation’ – potentially the site of such powerful affective commitments. ¹¹

One further point needs to be extracted from this second clue – that is, from our brief interrogation of the political as a dimension of social existence. Thus far, we have seen both that the political in some form is a necessary feature of society and that it is bound up with existential features of human subjectivity. Now, however, we must find the link to societal multiplicity. It is not far to seek. As Simon Bromley (2010) has argued in illuminating detail, political authority is always constituted in relation to a given population (thus excluding others), and normally within a given territory (whether this territoriality is exercised through year-round habitation, imperial subordination or nomadic circulation in space and time). Perhaps this demographic and territorial determinacy is due to the role of political structures in overseeing social life in general, or perhaps to the need (as Bromley argues) for physical co-presence in order to deliver their sanctioning role. ‘Political power’, says Mann (1986), presumably for similar reasons, ‘is necessarily centralized and territorial’ (p. 27). Either way, it is unlike, for example, relations of material exchange, which may indirectly link together individuals in completely different and territorially remote social settings. Unlike those other relations, the demarcating quality of the political makes it centripetal and necessarily finite. ‘[I]t is inevitably particular, creating an inside and an outside’ (Bromley, 2010: 244). This points to a condition of political (and hence societal) multiplicity. And the organisational and existential importance of the political imparts a territorial ‘lumpiness’ even to potentially centrifugal dimensions of social existence, such as exchange relations, fostering the socio-spatial ‘clustering’ qualities of society observed by Mann and Giddens.

If we turn now to the Scottish Enlightenment writer Adam Ferguson, the claim that societal multiplicity is somehow integral to the nature of the political per se becomes even more explicit. And this will provide us with our final clue. Ferguson’s 1767 Essay on the History of Civil Society conceives human society as intrinsically plural in the manner required by meaning #3. For Ferguson, the multiplicity of societal entities is not just an ulterior empirical fact about political existence. It is also a necessary theoretical premise of social thought:

If both the earliest and the latest accounts collected from every quarter of the earth, represent mankind as assembled in troops and companies; and the individual always joined by affection to one party, while he is possibly opposed to another . . . these facts [among others] must be admitted as the foundation of all our reasoning relative to man. (Ferguson, 1995: 9)

And he even adds that these same facts of multiplicity and partisanship ‘are to be retained in [our species’] description, as the wing and the paw are in that of the eagle and the lion’. In other words, they are part of the definition of humanity itself.

It is worth looking a little further into Ferguson’s argument, because his ‘troops and companies’ are functioning as a general abstraction of society. We know this to be the case because when speaking of them Ferguson (1995) also says that ‘the multiplicity of forms . . . which different societies offer to our view is almost infinite’ (p. 65). And yet, he argues, there are generic effects of their separation and interactive co-existence which are common to them all, whatever their form. Indeed these effects are even part-constitutive of society and its political structure: ‘[t]he society and concourse of other men’, he says, ‘are not more necessary to form the individual, than the rivalship and competition of nations are to invigorate the principles of political life in a state’ (Ferguson, 1995: 116). Societal multiplicity is thus intrinsic to what society is. But whence arises this ‘rivalship’? It comes from the fact of multiplicity itself: ‘[t]he emulation of nations proceeds from their division’. And this division is the source not only of competing interests, but of inter-subjective processes of identity formation: ‘we love our country, as it is a party in the divisions of mankind . . .’ (Ferguson, 1995: 26).

Correspondingly, there is a special quality to the antagonism between political communities. ‘What is it’ asks Ferguson, ‘that stirs in the breasts of ordinary men when the enemies of their country are named? . . . What interest [have they] . . . in the quarrels of princes?’ (Ferguson, 1995: 27–28). His answer is that ‘they cannot account’ for the passions so aroused. For the latter’s intensity comes not from visible conflicts of interest at all. It derives from the imbrication of individual identity with political community. Through this imbrication, inter-societal opposition promotes internal cohesion and thereby supports and validates the identity of the individual too.

Thus, the multi-societal nature of humans is not just a descriptive empirical universal; it is constitutive of the internal solidarities and dynamics of societies themselves. In fact, ‘[w]ithout the rivalship of nations and the practice of war, civil society itself could scarcely have found an object or a form’ (Ferguson, 1995: 28); and, pushing beyond Aristotle’s tendentially monadic conception of the polis, he adds that ‘Athens was necessary to Sparta, in the exercise of her virtue, as steel is to flint in the production of fire’ (Ferguson, 1995: 61). Correspondingly, if one removed the effects of multi-societal co-existence, societal cohesion itself would be endangered: ‘[c]ould we at once, in the case of any nation, extinguish the emulation which is excited from abroad, we should probably break or weaken the bonds of society at home, and close the busiest scenes of national occupations and virtues’ (Ferguson, 1995: 29). The implication of this statement is as clear as it is remarkable: if there were only one society in the world, it would quickly fall apart into a multiplicity; for without the interactive effects of that more-than-oneness, this society would lose ‘more than half’ (Ferguson, 1995: 116) the sources of its inner cohesion. For society to endure, it must be multiple.

For Ferguson, this positive role of multi-societal co-existence is not automatic: its operation depends on a proportionality of the parts, that is, on the existence of a balance of power. In 18th-century Europe, by contrast, when Ferguson was writing, small republics were ‘choked by the neighbourhood of more powerful states’ (Ferguson, 1995: 61). And the disproportion of power between European and other societies was, he lamented, even more disastrous: ‘the inhabitants of one half the world were let loose on the other, and parties from every quarter, wading in blood, and at the expense of every crime, and of every danger, traversed the earth in search of gold’ (Ferguson, 1995: 201). But the larger points still stand. The rivalry of societies proceeds from the very fact of ‘their division’ (Ferguson, 1995: 61). It generates ‘angry passions, that do not arise from an opposition of interest’ (Ferguson, 1995: 27) alone, but additionally from interactive processes of group identity formation. And this phenomenon is not restricted to state-based groups, but extends to all societies in history. ‘Thus without any settled form of government’, he notes of the stateless Iroquois, ‘they conducted themselves with the concert and the force of nations’ (Ferguson, 1995: 85, emphasis added). And Caribbeans too ‘conduct themselves . . . with a perfect discernment of their national interests’ (Ferguson, 1995: 88).

It is time to formulate our definition. Where then has our argument led us? We have followed three clues. The first was derived from Watson’s typology of historical international systems; and it suggested that the societal multiplicity we are trying to define rests fundamentally on political separateness of varying degrees. The second clue, prompted by Aristotle, led us to identify the political as a necessary dimension of all social existence. This dimension, we argued, did not always find expression in a separate organisation such as the state. But it did nonetheless have properties (of demographic and territorial finitude) which, unlike other dimensions of social existence, implied insides and outsides, and hence a multiplicity of entities. Finally, our third clue was supplied by Adam Ferguson who argued that the multiplicity of these entities was not a contingent happenstance, but a functioning part – even a precondition – of human society itself.

In short, the political in some form is necessary to social life per se; but multiplicity is intrinsic to the political, and it must therefore form part of its definition: the political is always simultaneously (and necessarily) the geopolitical. And at this point we need only to invert this latest step of the argument to reach our general abstraction of society too: if the inner qualities of the political point, among other things, to the necessary existence of other politically-ordered social formations, this tells us that the social world as a whole must sub-divide politically into multiple parts; conversely, those politically differentiated social formations are the elements – the units – of that wider whole.

For the purpose of the theory of Societal Multiplicity, therefore, a society can be defined at the most general level as a social formation which exists simultaneously as a geopolitical sub-division of the human world. Both elements of this definition are indispensable. First, our units must be social formations because polities are not free-standing entities but subtend upon the wider social relations that give rise to them. And yet not all social formations (meaning local complexes of social relations) are simultaneously geopolitical sub-divisions of the human world at any given point in time. London is a social formation. But it is not itself a geopolitical sub-division of the human world. It is nested within another social formation – the United Kingdom – which does simultaneously comprise such a sub-division. The international, we are saying, is not a purely political multiplicity; but without political multiplicity it would not exist at all. For this reason, if Societal Multiplicity is to capture the international in its fullness, it must not only picture the units as social formations; it must also add the second element–‘geopolitical sub-division’ – to the definition of society itself. It is this combination of social structure with geopolitical plurality (meanings #2 and #3) that has eluded both social and international theory in the past, and which the concept of Societal Multiplicity seeks to grasp.

In fact, Societal Multiplicity references simultaneously the fact of such sub-divisions, their differentiated social structures, and also their specific configuration in relation to each other at a given point in time. It identifies both the general object and the particular historical content of what we call ‘the international’.

Having made this claim, we must note that it is subject – and must be subject – to the same variability or flexibility as was invoked by Adam Watson when designing the spectrum on which he placed his historical international systems: across space and time, different social formations exhibit in differing degrees what we have decided is the defining (geopolitical) characteristic of societal multiplicity. If empire, suzerainty, hegemony and independence all fall between the impossible extremes that define the limit cases of the international, then the corresponding degrees of geopolitical agency exercised by the units across that spectrum qualify them in different measures as ‘societies’ composing the societal multiplicity which specifically constitutes the international.

The spectrum is important because it enables us to include apparently ‘hard’ cases such as the Kurds and Palestinians today, Poland during its tripartite partition, the Spanish colony of New Spain in the early 18th century, subjugated indigenous peoples, the present-day member states of the European Union and even that famous bugbear of international theory: European feudalism itself, with its layers of sub-infeudation and overlapping sovereignties. All of these can be placed at their respective points on the spectrum. And together they imply that societies may (and often do) overlap with each other, or contain other societies within themselves, or find themselves divided between two or more other societies.

Thus, when we outline our definition, we can state the key variables which compose it, and the range within which they operate. But anything more specific than this would be a definition immediately defeated by the variety of historical cases. Whether and how far a given social formation, whatever its formal standing, was also functioning as a geopolitical sub-division has to be established empirically through historical analysis. What will not vary, however, is the fact that at any given point in time the social world comprises a multiplicity of such entities of varying degrees, configured in a particular way in relation to each other. And it is from this fact that the five ‘consequences of multiplicity’ flow which make the international such an important dimension of social reality in all its branches.

Perhaps it is understandable that all our historical and sociological instincts still baulk at this general abstraction of society. We assume that it must involve a fallacy of some kind – of anachronism, of essentialism, of ‘politicism’ or Eurocentrism. ¹² But our reluctance might equally point to something else: namely how completely our third meaning of ‘society’, despite having always provided the widest, most encompassing frame of social existence, has been excised from the discourse of modern social theory, a discourse for which it should instead have provided an elemental premise.

Objections, objections!

As we saw at the start of this article, Ole Wæver argued strongly that a concept of Societal Multiplicity requires a general definition of society. Having constructed our definition, however, we now have to address the equally vigorous objections which any such generality is sure to provoke from other quarters. Four of these objections demand our immediate attention, claiming respectively that our definition is empty, realist, methodologically nationalist and disablingly anachronistic. Let us take them in turn.

First, then: is a general abstraction of society so broad as to be useless? Is it like those transhistorical concepts of classical political economy in which, according to Marx (1973), descriptive features of social reality had been ‘hammered out into flat tautologies’ (p. 86) that were incapable of saying anything determinate about real historical events and processes? Certainly, it is true that our definition accommodates geopolitical sub-divisions of almost any size and form, ranging from tiny hunter-gatherer bands through tribal chiefdoms, city-states, territorial empires and all the way to the billion-plus nation-states of India and China today.

And actual historical societies are variable in so many other ways. Different internal structures (capitalist versus state socialist, for example) are surely significant in ways that a single definition must elide. Further, a geopolitical sub-division may be internally diverse or wracked by social and ethnic conflicts that undermine its identity as a unified actor. It may be interrelated with other societies, making it dependent upon them or subordinate to them, once again blurring its existence as separate. It may be in active or frozen conflict with them over the extent of its (and their) territorial jurisdiction so that we cannot even say where its political borders run. And, of course, non-exclusive (e.g. feudal) forms of territoriality have also existed which have allowed multiple political actors to exercise different rights over the same populations at the same time. Finally, geopolitical sub-divisions are continually passing into and out of existence, for no society lasts forever. And hence at any given moment the world of societies also contains both social groups aspiring to the creation of new societies and existing societies that are sinking into dissolution.

These variations make for a bewildering kaleidoscope of social units – a ‘crazy quilt of polities’ (Ferguson and Mansbach, 1996: 60) which might indeed seem to render any general definition implausible. But here we must remember that the purpose of a general abstraction is not to provide a catalogue of all the empirical versions that a given phenomenon may have assumed in history. It is rather to isolate what all these variations share so that the general significance of the phenomenon can be mapped and analysed. And in this case the dazzling qualitative multiplicity of societal entities should not be allowed to obscure the logically prior fact of their quantitative multiplicity.

To put it another way: we know, with rare certainty, that the human world, at least since neolithic times, has always been divided into numerous co-existing entities of wildly differing shapes, sizes, conditions and levels of development. It is this remarkable fact, and not some theoretical caprice on our part, which dictates the need for an equivalently general concept. Precisely this fact is captured by extending the concept of society to embrace meaning #3: society as a geopolitical sub-division of the human world. And as the original argument for Societal Multiplicity sought to show, this sub-dividedness per se, which must be abstracted from qualitative difference in order to come into view at all, is anything but empty. Despite its extreme generality, five major consequences of this quantitative multiplicity shape the social world in quite fundamental ways. They entail: the very existence of that domain of social action which we today call ‘the international’ (and which, as we earlier saw, many social theorists have struggled to incorporate); the proliferation of co-existing difference among the multiple instances of social being which comprise this domain; the imperative to interaction among them through diplomacy, trade, war and more; the overdetermination thereby of ‘internal’ social development by ‘external’ interactions; and the correspondingly non-linear, dialectical patterns of world history both locally and as a whole (Rosenberg, 2016).

Indeed, what is truly implausible is not the breadth of our general abstraction. It is that the fact of dividedness itself should have so rarely been formulated as a theoretical object, and that so many social theories should have therefore proceeded as if it were not the case, thereby enlarging the discursive tension between social theory and history as modes of analysis (Rosenberg, 2006). As Robert Nisbet (1969), Theda Skocpol, Friedrich Tenbruck, Charles Tilly and the many critics of ‘methodological nationalism’ have alleged, it is the traditions of social theory which ignore this fact that must answer to the charge of being unhistorical.

But have we not ourselves, in our attempt to fill this lacuna, fallen back into the reifying, realist perspective that (along with its ‘reductionist’ liberal twin) we were trying to escape? This second suspicion arises because of the centrality we have accorded to political separateness and hence to the political per se in our definition of society. Have we not conceded in this way the fundamental premise of realist international theory? Yes and no. Certainly, we have agreed that the phenomenon of society is demarcated by its political structure. But before taking that step, we had already redefined the political itself not as a free-standing phenomenon, but as an outgrowth of the wider conditions of social existence. The point is crucial: the more we lean on the political in our definition, the more important it becomes to simultaneously deconstruct it into an emergent dimension of the social. Otherwise, we do indeed run the danger of reification. And this applies equally to any other dimension of society such as ‘the economic’, ‘the cultural’ and of course ‘the international’. In this case, we said that the political is the means through which a defined social group constitutes itself as such and manages its own existence (or has it managed by a subgroup that succeeds in capturing and channelling the social power which human interaction everywhere produces). This means of self-management includes certain general features like the connection with violence and the restriction to a given population, But, in line with meaning #2 of ‘society’, (historical specificity), it also varies with the type of society being reproduced – to the extent of sometimes not even entailing a state organisation at all. Even hunter-gatherer bands, if they are to endure, need to secure their reproduction by some socially recognised structure of rules, expectations and sanctions. Thus, a polity is never just a polity. It is always the political dimension of a given social formation, of a given kind, at a given point in its historical existence. And our ‘society’ therefore also includes social, economic and cultural structures. Each of these is affected by its participation in a social world that includes multiple interacting instances of itself. The general formula for Societal Multiplicity (co-existence, difference, interaction, combination and dialectical change) can be used to analyse the international formation of these non-political structures too. And hence a general abstraction of societal multiplicity (the international) can speak far beyond the limited domain captured by realism.

But this apparent solution leads us straight into a third criticism. ‘Social formation’ may well be a useful prompt for reintegrating the political into other dimensions of social reality – satisfying meaning #2 of ‘society’. But as an apparent reference to self-contained units, it resurrects the very fallacy of methodological nationalism that the social sciences of the last few decades have been at such pains to dismantle. The enormous literature on globalisation, flawed though it may have been (Rosenberg, 2000, 2005), was also a sustained campaign to expose this fallacy once and for all, initially in relation to the post-Cold War world, but subsequently for all historical periods (Hopkins, 2002). Are we now to indulge a definition of ‘society’ that blithely disregards this advance?

Fortunately, the solution to this problem is not as remote as it might seem. Recall that our general abstraction of the political included centripetal properties requiring territorial and demographic demarcation. We did not, however, discover the same properties in relation to the other dimensions of social existence in which the political is always grounded. From this, it follows, in principle, that these other dimensions will not be fully co-extensive with the political, even while they operate within its necessarily fragmented world of multiple polities. A hunter-gatherer band may (through exogamous marriage arrangements) extend relations of kinship deep into another band, without itself ceasing to be politically distinct in the sense we have described (Buzan and Little, 2000). A settled tribe like the Baruya of New Guinea could be dependent for their very subsistence upon ‘a vast intertribal and interregional system of exchanges’ (Godelier, 1977: 141) while nonetheless carrying on a political existence of their own within that system. Religious structures may stretch across different societies, as did the Catholic Church in mediaeval Europe. And of course in the modern capitalist era the ‘separation of the economic and political’, (the one globalising and the other territorially delimiting), is perhaps the central institutional feature of social existence on a world scale.

What does this mean for our definition of society? It means that while the political structure of a given society relates to a particular population and territory, that society is simultaneously located within wider economic, cultural and ecological structures that traverse political boundaries. The intersectionalist insight that human individuals always exist at particular points within multiple, cross-cutting structures of identity and power (Crenshaw, 1989) applies to societies too.

It turns out, then, that the attempt by Michael Mann and Anthony Giddens to say that a society ‘‘stand[s] out’ in bas-relief’ and exhibits ‘interaction cleavage’ at its ‘boundaries’, and yet that it is not ‘a bounded totality’, need not be the simple equivocation it appears. Rather, these divergent properties co-exist as different dimensions (political and other) of the same social entities, as indeed Mann (1986) later observes: ‘political power heightens boundaries, whereas the other power sources may transcend them’ (p. 27). It is this duality that gives rise to the ‘gestalt switch’ effect attending the international itself – simultaneously an interconnected structure and yet also a plurality of different entities. By explaining this phenomenon, our definition of society actually releases the idea of Multiplicity from the substantialist ontology in which it might otherwise have been trapped.

And finally, what of the charge that our definition of society is an anachronism which falsely superimposes modern characteristics onto world history at large? Like the accusation of conceptual emptiness discussed earlier, this charge arguably misunderstands the nature of a general abstraction. If our definition of the political as such is sound, and if it is used correctly, it provides no obstacle to a historical sociological analysis of the variety of social formations in world history. For it is not modelled on any particular historical form of polity – the sovereign state, for example. It claims only that every form of social existence has a political dimension of some kind, that the centripetal quality of the political entails limitation and hence multiplicity, and that this limitation is traversed nonetheless by other dimensions of social existence, entailing the chronic interpenetration of structures, influences, dependencies and so on which form the object of so much geopolitical activity.

This claim holds for national, feudal, imperial, city-state, tribal and band societies alike. It takes varied forms in these different cases. But in a general sense it applies to them all. In this way, it is like Marx’s claim that all societies depend upon a collectively organised metabolic interaction between humans and nature (i.e. that ‘production’ exists as a general feature of human societies even though it takes different forms in different cases). This is why it can be expressed as a general abstraction, and why, as such, it belongs among the foundational premises of social theory.

In fact, the argument we have made, depending on our angle of vision, can now provide general abstractions of no less than four theoretical objects that have been circling each other throughout our discussion. It identifies ‘the political’ as the necessarily territorial, authoritative moment of social existence; ‘society’ as a politically demarcated social formation together with its non-political interconnections with other societies; ‘the international’ as the entire dimension of social reality which arises specifically from this entangled co-existence; and ‘the social world’ overall, now reconceived as necessarily including this international dimension for reasons that are, as we have discovered, intrinsic to the very nature of social existence. Via our redefinition of society, undertaken in order to ground the concept of Societal Multiplicity, we have in fact reconnected social, political and international theory.

‘Societal multiplicity’ completed?

Where does all this leave us regarding the criteria we identified near the start of this article? Does our definition – social formations which function simultaneously as geopolitical sub-divisions – incorporate all three core meanings of ‘society’, while also enabling us to visualise both the specificity of the international and its significance for other dimensions of social reality? The answer seems to be as follows. We have argued that social life in general (meaning #1) always includes a political dimension of some kind. We noted also that this will always take a given form, and will have a particular content specific to the kind of social existence involved (meaning #2). At the same time, it became apparent that the centripetal properties of the political entailed two further characteristics: that the political is necessarily expressed in a plurality of entities, and hence in an international dimension to all branches of social existence (meaning #3); and that some non-political dimensions of this existence, which do not share those centripetal properties, are co-substantial with the political and yet not necessarily co-extensive with it, meaning that they may traverse political jurisdictions. We then defined a society as a geopolitical sub-division of humanity arising from these linked conditions – a sub-division which is therefore both circumscribed and interpenetrating with others. And correspondingly, the international is a multiplicity of entities which, far from existing over against societal structures, arises out of them and is intrinsic to social existence per se.

This, admittedly, amounts to a very particular definition of society. It combines the optics of social and political theory to posit ‘the multiple-societal’ as perennial to human life. In fact, however, and as Ferguson insisted, this is the only form in which society has ever actually existed. All other definitions are arguably violent abstractions which tear the phenomenon of society out of its inter-societal integument and thereby disconnect the geopolitical dimension from its ever-present sociological content too. Twin reifications – of methodological nationalism and autonomous geopolitics – are, and always have been, the unavoidable result. As predicted in the Introduction, the fact of geopolitical sub-division has turned out to be both more and less defining of social reality than those reifications imply.

All that remains, therefore, is to test whether this general abstraction fits (and rounds out) the idea of Societal Multiplicity for which it has been designed – and whether, as Wæver requested, it turns out to have been the definition that was tacitly presupposed all along. This we can do by examining how it relates to the five major consequences of multiplicity, as originally identified (Rosenberg, 2016: 135–141). Do these consequences proceed logically from the nature of society as we have now defined it? And does this derivation result in a more convincing account of the international than those available in realist statism and liberal pluralism?

The first consequence – that societal multiplicity entails an ‘international’ level of social reality in which societies co-exist, a ‘second-order’ social formation or ‘society of societies’ (Buzan, 2023) – fits without difficulty. The centripetal properties of the political (especially when added to the spatio-temporal unevenness of social development as a planetary whole) explain in principle the more-than-oneness of social formations and hence the perennial existence in some form of an international dimension to the social world.

Second, the original theory posited that more-than-oneness, conjoined with its necessary distribution across an ecologically varied planet, creates the possibility (and in fact the unavoidability) of ‘difference’. And this too is given within our definition: the general existence of the political, necessarily expressed in multiple societies, evolves different concrete shapes and developmental potentials depending on the forms, scales and varied ecologies of the social formations in which it is grounded. This makes sense of the fact that the human world comprises – and always has comprised – a ‘crazy quilt of polities’ that today contains both continental super-states and tiny ‘island nations’.

Third, the fact of co-existing difference entails both threats and opportunities which compel societies into continuous interaction via diplomacy, trade, war and much else. Not only is our new definition consistent with this claim. It also enables us to visualise the fact that while some of these interactions proceed from the imperatives of geopolitical existence, others arise from the non-centripetal character of other aspects of social life. This distinction, by recognising ‘external’ pressure and ‘transversal’ infiltration alike, preserves us from a definition of societies as ‘bounded totalities’. More positively, it also underpins a sociological definition of the international which is both wider than that of realism while remaining intellectually precise and demanding: the international on this reading is ‘that dimension of social reality which arises specifically from the co-existence within it of more than one society’ (Rosenberg, 2006).

The resulting scenario, in which ‘second-order’ international effects are discovered within all branches of social life, is key to the richer version of IR propounded by Societal Multiplicity, and thereby also to the claimed significance of IR for the human disciplines at large. And in line with this, our definition supports the fourth consequence of multiplicity: that societies, even while they retain their political separateness, always (and even in their political dimension) evolve as combinations of internal and external determinations. As Friedrich Tenbruck pointed out, even when societies have attempted to cut themselves off from others, that very attempt is an external orientation that plays into the inner development of the society (Tenbruck, 1994). The norm, however, is anything but seclusion, even of this contradictory kind. The norm is that the ongoing political, economic, cultural, historical, technological, linguistic, and other development of a society is constantly being compared to and influenced by those of others. Hence, and in contrast to realism’s politically-focussed designation, ‘the international of everything’ (Rosenberg and Tallis, 2022).

And finally, with differentiated societies necessarily entangled with each other, the fifth consequence of multiplicity follows too: their socio-historical development (and that of the human social world as a whole) cannot take a uni-linear form. Conjoining reproductive with interactive logics, it must have a dialectical character. Crooked lines, tangential offshoots and rupturing intersections must, and do, proliferate; and these, unpredictably but nonetheless intelligibly, overdetermine the course of events and the patterns of socio-historical change. Our definition of society as a geopolitical sub-division of a wider, interconnected human reality makes sense of this necessarily dialectical form.

Conclusion

Throughout history the human species has been divided among multiple populations - many ‘troops and companies’. A fuller account of why this multiplicity obtains would need to go beyond the nature of politics alone, and to include the spatio-temporal unevenness of socio-historical existence on a global scale - its uneven and combined development (Rosenberg, 2010). The simple fact that it obtains, however, entails the transhistorical existence of the international as a dimension of the human world, and one with considerable implications for how we conceptualise all other parts of that world too.

Why then has this dimension proven so difficult to pin down? The answer apparently is that it takes almost infinitely various forms, and hence has appeared to be undefinable in itself. Quantitative multiplicity is repeatedly crowded out by the effects of its own noisier sibling: qualitative multiplicity. But this is not a necessary outcome, for quantitative and qualitative multiplicity are not in fact the logical siblings they appear.

From the fact of quantitative multiplicity alone, we can already derive at least five consequences that hold throughout human history, whatever the concrete configuration of bands, tribes, states, empires and so on. Indeed, it is these consequences - co-existence, difference, interaction, combination and dialectical change - which generate the possibility and actuality of qualitative multiplicity itself. Specified non-anachronistically, they reveal, for any given setting, both the international dimension of social existence and its reach into other dimensions. Of course, this means that what `the international’ is must be identified and redefined afresh for every setting in which it is analysed. Thus, qualitative multiplicity has its day too. But this is a requirement that the international shares with all general abstractions - for example, labour, politics, culture, wealth and so on.

Nonetheless, in order to be applied, the five consequences do require a very particular definition of ‘society’: one which includes the distinctive element of geopolitical agency (meaning #3) and yet which is otherwise wide enough to include all other aspects of social existence without suppressing the varied historical forms these will take (without, i.e. violating meaning #2).

The definition that meets this requirement is `a social formation which is functioning simultaneously as a geopolitical sub-division of human existence’. Perhaps we could call it simply `a geo-social formation’, for the term itself matters less than the content we give it. Such formations will indeed take radically different shapes, and interrelate with others in radically different ways, in different cases. But `geo-social formation’ is the definition of society that matches in its generality the concept of Societal Multiplicity, while also preserving the latter’s claim to identify the determinacy of the international dimension. And it also - perhaps unexpectedly - allows for both the inter-societal separations and the trans-societal interconnections which have elsewhere been allocated to opposed (realist and anti-realist) visions of the international. Not only does it fill the definitional gap at the heart of the idea of Societal Multiplicity; it also further strengthens the latter’s claim to grasp the international as an object of social theory.

We return, then to the original purpose of the Multiplicity idea, which was precisely to find a way of formulating the international that would enable it to be reintegrated with the field of social theory in general (Rosenberg, 2016). That exercise had two steps: the first was to posit that society - the fundamental object of social theory - always co-exists in the plural; and the second was to reason out the consequences of this plurality for social reality at large. But Ole Wæver - along with all the other critics who challenged this idea - was right. A crucial third step was missing. Without an analytical definition of `societies’- in effect, a re-definition in the light of the international - the idea of Multiplicity was a precariously unfinished proposition. It was in constant danger of tipping over into one of the very fallacies it was attempting to overcome - either a proto-realist vision of bounded entities or, as some of its supporters indeed advocated, a pluralist claim about the multiplicity of almost anything.

Doubtless the definition presented here - necessarily multiple social formations functioning as geopolitical sub-divisions of the social world - will not end the debate. But it does at least show that the units of Societal Multiplicity can be defined. And if, in the course of that exercise, the realist and pluralist fallacies have been successfully avoided, this further supports the intuition that the solution to the challenge of the international does not lie in international theory alone. It lies rather in the latter’s reintegration into those more general theories of social reality from which it has been so long separated, and at such cost to both sides.