The main argument

We begin with the idea that social practices are not just an aspect of the social world but its most basic constituent. Our central thesis is that practices are institutions defined as rule-based domains of interaction produced by intelligent individuals. This is a departure from the orthodox view of practices as actions, especially habitual or routinised actions.

In our vocabulary ‘intelligence’ is a term of art which designates the capacity to respond to a situation in a way that is not pre-programmed and that may involve mistakes. The notion of making a mistake gives us a definition of a rule in the normative sense of the term. As Winch points out, ‘The notion of following a rule is logically inseparable from the notion of making a mistake’ (1958: 32; PTIR: 40 (n 24); 98 (n 4)).Winch shows that it is warranted to speak of a rule (in the sense in which a rule guides action) only if it is possible to distinguish correct performances from incorrect ones. In this sense, all rules are norms (standards of appropriate conduct). Also, rules are meaningful propositions which an intelligent agent can understand or misunderstand.

We argue that the proper procedure for understanding practices that are not yet understood is the internal point of view (internalism). This is the point of view of competent practice participants. It involves both the epistemological claim that each participant knows that he or she is following the rules of a common practice together with others, and the normative claim that such rules represent standards of appropriate conduct. The first claim suggests that the participants must be following the same rules, and must know that they are doing so. According to the second claim, familiar from the writings of Hart (Reference Hart1961), acting against the rules is not just inconvenient or inefficient; it is considered wrong or inappropriate by the offending agent. The internal point of view has two modalities: it may refer to an agent perspective that a novice must adopt in learning how to participate in a practice ab initio, or to an observational perspective by a third party observer (social scientist, analyst, philosopher) who studies what the participants are doing.

What are the conditions for a competent action within a practice, or correlatively, for the competent observation of a practice? We hold that these conditions are captured by the already mentioned internalism as well as by descriptivism. In order to observe an action that takes place within the framework of a social practice (a ‘practice-dependent’ action in our terms), it is necessary to have a prior description of the basic rules of that practice. Such rules – ‘constitutive rules’ (Searle Reference Searle1969) – identify the actors and specify their roles and rights inside a sphere of rule-governed activity. The constitutive rules of chess provide a good example: they define the pieces, the permissible moves, and the playing field (the chess board). Although practices resemble games in their dependence on constitutive rules, one basic distinction between them is that unlike a game, a practice has constitutive rules that are amenable to contestation and change (in IR see Wiener Reference Wiener2008).

Our notion of practices as institutions reproduced by conscious activity derives from the hermeneutic tradition and is indebted to Hegel’s conception of ‘ethical life’ (1821). Institutions in this sense do not represent devices for preference aggregation or formal decision-making procedures, as in rational choice methodology. They are ever evolving human artefacts that demand reflective consciousness, language, and an appreciation of the values embedded in them. Similar arguments for practices have been advanced by Oakeshott (Reference Oakeshott1975) and Wittgenstein (Reference Wittgenstein and Anscombe1968). Central to our account is Oakeshott’s and Wittgenstein’s concept of rule-following. In contrast to a formal statement of a rule, rule-following is an activity predicated on meaning in use. The meaning of a rule can only be learned in the course of its use, or less technically, in a context. When a surgeon says ‘Pass me the knife’ in the context of an eye operation it is understood that ‘knife’ means a scalpel and not a bread knife, even though this is not explicitly stated.

In Part I of the book we identify the core elements of a practice as an analytical object and show how they cohere. These elements include rule-following, internalism, constitutive rules, norms, and meaning. On this basis we proceed to criticise the recent ‘practice turn’ in the discipline of International Relations (PTIR: Chapter 2) arguing that it exhibits the defects of externalism, a standpoint that gravitates towards invariance, reduction and a search for foundations. While an interpretation of a social practice from the internal point of view employs the categories that the practice participants themselves use in making sense of their practice, the external point of view imports a different language of observation from the outside. The major drawback of this externalist understanding is its explanatory arbitrariness and its lack of intelligibility to the participants under observation. In Part II, we identify two macro practices in the contemporary international realm – the practice of sovereign states and the practice of global rights – and trace their dialectical relationship to one another. We argue that such a macro-level analysis of practices that takes seriously the problem of internally driven, dialectical change can advance our understanding of the international sphere better than the existing set of theories in the field of International Relations (IR).

With these preliminary comments in mind, we turn to the details of the points raised by the individual contributors to this volume.

Wittgenstein (Gunther Hellmann)

Gunther Hellmann welcomes our commitment to descriptivism and internalism, and our Wittgensteinian leanings. But he finds our interpretation of practices as institutions insufficiently Wittgensteinian. It is pertinent to indicate that we do not see ourselves as expositors of Wittgenstein or any other philosopher whose ideas have been an inspiration such as Georg WF Hegel (1821), Michael Oakeshott (Reference Oakeshott1975), HLA Hart (Reference Hart1961) and John Rawls (Reference Rawls1955, Reference Rawls1996, Reference Rawls1999). Wittgenstein’s insights concerning descriptivism, rule-following and language-games are important only insofar as they enable us to develop a coherent account of practices.

According to Hellmann, Wittgenstein associated practices with the concepts of praxis or practice, not with that of institutions. This contention is hard to defend considering that extensive passages in the Philosophical Investigations (1968 [1953], abbreviated as ‘PI’) where Wittgenstein claims that utterances and actions acquire meaning within ‘language games’, a term he treats as interchangeable with ‘customs’, and ‘institutions’ (PI, section 198–99, see also sections 337, 380, 540). That language games are sustained by the intersubjective agreement among individuals (PI, sections 198, 202, 208), and that each language game has rules of its own (PI, sections 66–7) represents a line of reasoning consistent with our take on practices as human institutions that require public criteria of interpretation and judgement.

Hellmann reads Wittgenstein’s understanding of practices as being bound by the normal usages of both ‘praxis’ and ‘practice’. Our response to this is that the term ‘practice’ has a diverse set of usages. Here are a few: ‘Fighter pilots need to practise their flying skills.’ ‘The diplomat’s practice was to read the on-line news reports first thing every morning.’ ‘In campaigning, politicians often promise a lot, but in practice they seldom fulfil their promises.’ ‘The doctors set up their medical practice in a squatter settlement.’ While in the book we are tackling the problem of meaning (meaning in use), our principal task is not to offer a linguistic analysis of the term practice (or praxis). That is, we are not attempting to ‘switch to a new vocabulary’ or to establish a new way of speaking about practices, as Hellmann suggests. Our objective is to identify the conditions under which a practice can be said to exist, including the conditions that enable observers to make sense of a puzzling or otherwise imperfectly understood practice. Hellmann is right to point out that Wittgenstein was interested in linguistic analysis. Our task however is social analysis. We are not just importing Wittgenstein’s statements as they stand in the Investigations but transforming them into social philosophy – a philosophy of practices – which he himself did not provide but only gestured towards (see Onuf Reference Onuf1989: 44, 49 in PTIR 9 (n 26)).

Another complaint is that by emphasising the concept of institutions (practices) we have eliminated the concept of action, responsibility and agency. It is, indeed, possible to make an argument of this sort. Michel Foucault famously presents modern individuals as passive subjects who are produced by impersonal, institutionalised processes of ‘biopolitics’ (Foucault Reference Foucault, Macey, Bertani and Fontana2003 [1997]) and disciplinary strategies (Foucault Reference Foucault and Sheridan1995 [1975]). Unlike Foucault, we theorise individuals in active terms, as agents, and institutions as devices that enable crucial forms of agency. The institution of citizenship for instance equips individuals with rights to political participation which they would otherwise lack. In short, individuals act within practices, and explaining their actions demands knowledge of these practices.

Knowing what a given practice is requires proper descriptions of its constitutive rules. The rationale here is that a particular set of rules enables us to distinguish one practice from cognate practices; for example, it enables us to contrast the practice of diplomatic mediation with that of legal adjudication. Hellmann questions our search for ‘coherent and accurate descriptions’. It entails, he suggests, a correspondence theory of truth that contravenes the spirit of Wittgenstein’s later philosophy. Here Hellmann seems to suggest that we seek to describe a mind independent reality. Such a postulate is foreign to our Hegelian world-view. For us, as for Hegel, reality is at bottom a social reality whose building blocks are not mind-independent objects or states of affairs, but meanings, interpretations, and rules – a position that projects a coherence theory of truth. The descriptions we have in mind are descriptions of rules. Because rules are meaningful propositions, they are not accessible via unmediated observation in the way natural kinds (cats, dogs) or states of affairs (being cloudy) are. We have no other cognitive access to a rule other than via its description.

In defending descriptivism, we do not put forward a universal, invariant procedure of framing ‘coherent and accurate descriptions’ of some mind-independent states of affairs. The procedure for observing socially constituted actions in our account involves descriptions that can only be internal to the social practices within which these actions take place. Like Wittgenstein in his later writings, we consider ourselves to be providing ‘a description of descriptions’. Another way of expressing this is to say that we are engaged in a double hermeneutic.

Further, Hellmann believes that the search for ‘appropriate descriptions’ implies strict criteria of appropriateness that vitiate the ethos of the Investigations. To us it seems clear that Wittgenstein’s account of rule following is based upon the notion of criteria in the already discussed sense of a distinction between mistaken and appropriate performance. More specifically, providing a proper description of some action within a practice involves, first, an internalist description of that action in that particular practice, and, second, an account of rule-following where this presupposes knowing the relevant criteria of appropriateness. For example, in order to understand what we, the authors, and our critics are doing now – having a scholarly exchange – the reader must understand in some broad sense the practice of academic life. The criteria of appropriateness (the norms) that structure this practice must be relevant to it in the sense that they are distinct from those relevant to alternative practices such as having a contest, a political rally, or a wake.

Another criticism is directed at our claim that the proper task of philosophy is the ‘logical clarification of thoughts’ (PTIR: 4). This is a line from Wittgenstein’s early Tractatus Logico-Philosophicus (1961 [1921]). Hellmann finds our reference to this early work problematic since in the Investigations Wittgenstein recants his former Tractarian views. Our riposte is that in the Tractatus the propositions of language are said to mirror the logical structure of the world (comprising combinations of simple objects which Wittgenstein calls states of affairs). In the Investigations language is no longer grounded externally, in states of affairs, but internally, via the concept of language games. Wittgenstein has changed his view on the nature of language. It is less clear whether he has changed his view of philosophy, a point to which we shall return in due course. However that may be, the idea of logical clarification of thoughts is not necessarily tied to any specific doctrine of metaphysics or semantics and as such does not entail the alleged contradiction.

A final objection raised by Hellmann is to what he calls our ‘pneumatic way of thinking’. By this phrase, which appears in paragraph 109 of the German text of the Investigations, Wittgenstein means some kind of foundationalism or scientism based on a representational view of language. As an alternative, Hellmann recommends understanding language as praxis – as a way of coping with the world. Coping here is an idiom that aligns Wittgenstein with the school of pragmatism. Since we are not foundationalist we do not insist that there is a single correct reading of Wittgenstein. Nonetheless, there is one interpretation of ‘coping with the world’ we find suspect. This would be one which sees social practices in purely instrumental terms, as tools that can be brought to bear on a world that is objectively given in a way that can be ascertained by simple observation. On our premises the world ‘to be coped with’ is one which is itself socially constituted by a wide variety of practices. Crucial in this respect are the class of non-instrumental practices (also known as constitutive or authoritative practices) that we explore in Chapter 1 of PTIR. In these, the status of the participants and the values they pursue are constituted within the relevant practice. That is, they are not goals which pre-exist the practice and for the pursuit of which the practice was generated as an instrument.

Wittgenstein, Gunnell and the internalist interpretation of practices (Maren Hofius)

In a carefully developed analysis, Maren Hofius challenges us to clarify two basic problems. The first is the relation between theory and practice. The second is the relation between an observer and the social practice under observation. Hofius suggests that our kind of practice theory leaves the first question unanswered, and the second, inadequately answered. With respect to the first, our response is brief. We do not operate with the orthodox distinction between theory and practice where ‘practice’ means ‘action’. For us, a practice is a social institution constituted by rules and usages that accompany the rules. Because such rules and usages have meaning, and because the rules operate as norms (guides to conduct), they cannot be construed as brute facts. Instead they are available only under a description, and in this sense are ‘theoretical’.

With respect to the second problem, our critic asks whether the language which the proverbial theorist cum internalist uses in describing a practice is the same, similar, distinct or categorically different from the language of the practice participants under observation. In essence, this is an invitation to explain the internal point of view and the observational procedure it involves. Since Hofius bases her critique on a paper by the Wittgensteinian scholar John Gunnell (2013; see also Gunnell Reference Gunnell1998, Reference Gunnell2004, Reference Gunnell2009) whose work we acknowledged in passing in PTIR (42 (n 28); 47 (n 37)) we wish to respond to this source of argument more fully.

It is notable that there are three points of convergence between Gunnell’s and our reading of Wittgenstein. To begin with, we are in agreement that in the Investigations Wittgenstein views philosophy as an enquiry whose point is the clarification of ‘the use of our words’ (PI, section 122; see also PI, sections 125, 133; Gunnell Reference Gunnell2013: 87, 90, 93, 97). (The complication here is that there is still a distinction by implication, between clarification of thoughts, Wittgenstein’s motto in the Tractatus, and clarification of meaning, namely, meaning as use). Some think, not entirely without justification that in the Investigations Wittgenstein adopts a view of philosophy as a therapeutic enquiry (that dissolves philosophical problems). But seeking clarity and being therapeutic are not mutually exclusive ideas, and Wittgenstein discusses both in the same paragraph (PI, section 133). In any event, we wish to underline Wittgenstein’s and Oakeshott’s conception of philosophy as a pursuit of conceptual clarity and thus its descriptive character.

This brings us to the problem of the ‘internal point of view’ as a perspective which a third-party observer – a theorist or a philosopher – adopts in attempting to understand a not-yet understood practice or an action within a practice. The perspective is internal provided the observer uses the terms which the participants use in making sense of their own practice. This does not mean that the theorist must become an actor in order to perform a proper internalist observation, as Hofius speculates. Spectators at a Wimbledon women’s tennis final understand the game although they are not themselves players in the match. The theorist counts as a participant in an epistemic sense whereby making sense of the practice depends on using interpretive materials taken from the practice itself, and not from a transcendental standpoint outside the practice.

A second point of convergence between us and Gunnell is that we, too, employ the idea of a double hermetic to show how the internalist procedure of understanding is supposed to work. To get at this idea we must differentiate the act of understanding into two logical components, one pertaining to its object (what is understood) and one pertaining to its subject (the one who understands). In natural science, the subject constructs the object by constructing a language of observational terms. As physicists, for instance, we observe what we have defined as ‘atoms’. We are free to change our vocabulary and by so doing change the thing observed. In social science and philosophy, we enjoy no such freedom in constructing the object. The object, a social practice or a Wittgensteinian ‘form of life’, pre-exists the observer’s attempt to understand it. Unlike natural scientists, social scientists and philosophers are trying to understand something that is already understood by its own participants. As Gunnell writes:

In the book, we captured this insight by citing Winch, ‘Our practice is prior to any interpretation’ (Winch Reference Winch and Winch1987: 57; PTIR: 47 (n 37)). Practices then involve a double hermeneutic: the understanding of understandings. Because such understandings belong to the actors observed, they offer a line of resistance against the theorist’s attempt to explain them away: this is the gist of internalism.

At this juncture, we draw different conclusions about the activity of philosophising than those reached by Gunnell. As radical internalists, we hold that the philosopher is supposed to explicate and clarify the self-understandings of the agents by using the terms that the agents under observation are themselves using. In contrast, Gunnell is inclined to think that philosophy contains an additional, interpretive layer that positions it at one remove from the understandings of the practice participants. Thus, in contrast to first-order social practices such as religion or politics, the practice of philosophy (or the practice of social science) constitutes a ‘metapractice’. The analogy here is between language and metalanguage. For example, consider ‘“Der Schnee ist weiß” is a sentence in German.’ Here the expression in double quotation marks is a first-order language and that in single quotation marks is a second-order or metalanguage. On this view metapractices cannot be purely descriptive, they must be explanatory in some sense. This argument seems to be at variance with Wittgenstein’s remark that philosophy describes and thus, in a quietist manner, leaves everything as it is (PI, section 124). The more interesting question is, why does Gunnell make this move? He distinguishes understanding that takes place within a practice from interpretation as a ‘rendition of that understanding’ (Gunnell Reference Gunnell2013: 88; see also 2004: 94). This corresponds, he claims, to Wittgenstein’s distinction between Verstehen and Deutung or between ‘seeing’ and ‘seeing as’ (PI, sections 121–222). So construed, philosophy does not aim to describe something but to reconstruct or reconceptualise it. In Gunnell’s words, ‘The descriptions presented in second-order analysis are thus reconstructions or interpretations rendered in a theoretically informed language of their own’ (2004: 88, original emphasis). Only in this reconstructive or representational sense can philosophy have a differentia of its own – by giving us a synoptic view of knowledge. ‘But behind such [philosophical] representations is, necessarily, a general theory of what kind of thing is represented’ (Gunnell Reference Gunnell2004: 88, original emphasis; see also 2004: 90). If we press this conception of philosophy further, it seems to make concessions to the external point of view (it does not collapse into a full-blown externalism since philosophy must remain in ‘dialogue with’ its subject matter, Gunnell Reference Gunnell2004, 93).^{Footnote 1} What Gunnell implies is that philosophy must transcend what it finds – namely, the self-understandings of those who participate in a social practice – in order to justify its existence as a distinct form of activity that counts as doing philosophy and not, say, as jogging, counting, and so on. In contrast, we regard philosophy as a parasitic activity which is limited to clarifying and explicating, on a more general, conceptual level, what it finds but which need not use what it finds in order to construct its own, as it were, philosophical method. For us it is important to recognise Wittgenstein’s admonition that ‘Philosophy may in no way interfere with the actual use of language; it can in the end only describe it. For it cannot give it any foundation either. It leaves everything as it is’ (PI, section 124; see also section 126).

Let us finally consider the distinction between understanding and interpretation that Gunnell appeals to in order to elucidate Wittgenstein’s argument for rule-following (Gunnell Reference Gunnell2004: 93–4; 2013: 89). At 201 of the Investigations we encounter the remark that an action can always be made out to accord with a rule. Saul Kripke attributed to Wittgenstein a sceptical paradox according to which, given the possibility of multiple interpretations of any rule-action nexus, we do not know (hence the scepticism) whether a particular action counts as an instance of rule-following or not (Kripke Reference Kripke1982). Kripke’s solution was to point to community agreement that stabilises this nexus, a reading we endorse. Gunnell doubts that community agreement matters per se (2013: 89). Instead he takes Wittgenstein to be saying that, although the possible interpretations of a rule are unlimited, at the end ‘there is a way of grasping a rule which is not an interpretation’ (PI, section 201, original emphasis; Gunnell Reference Gunnell2004: 94). The contrast here is between doing something and explaining what has been done. Grasping (‘understanding’ in Gunnell’s terms) refers to learning how to perform an action, and interpretation, to an account of action (Gunnell Reference Gunnell2013, 88), which, for all we know, may be arbitrary. This way of presenting the puzzle compares doings by the agent with explanations/descriptions by the theorist without assuming that the doings in question are rule-governed – the element of rule-following drops out. Our account fleshes out the rule-following aspect of Wittgenstein’s position in two ways. The first is that the regress of potentially unlimited interpretations of an action that accord with a rule can be arrested because eventually the agent gets (grasps, understands) what the rule demands. The emphasis here is on the fact of regress and not simply on a contrast between understanding and interpretation: we treat these two terms as interchangeable to mean ‘making sense’. Notice that what is grasped is a rule of action, and not just a meaningful doing. The further consideration is that rule-following presupposes a criterion of appropriateness. For, as Winch said, it makes sense to speak of following a rule only if it would be possible to distinguish appropriate from mistaken performances (1958: 32). This criterion depends on community agreement. Thus, we conclude with Wittgenstein that ‘To obey a rule, to make a report, to give an order, to play a game of chess, are customs (uses, institutions)’ (PI, section 199). In effect, we read Wittgenstein as a conventionalist, a third point of convergence with Gunnell (2004: 88–9; 2013: 88, 92) though we arrive at it via a different route.

The methodology of social science (Nora Stappert)

We turn now to Nora Stappert’s critical note on our book. She begins with an observation about the recent practice turn in the discipline of IR, represented by the volume International Practices (2011a) edited by Emanuel Adler and Vincent Pouliot. Most of its contributors embrace the editors’ proposal for bridge building which requires finding a commonality among the main theories of international relations: realism, liberalism, English School theory, constructivism and postmodernism (Adler and Pouliot Reference Adler, Pouliot, Adler and Pouliot2011b). This common denominator is the category of practices, a term used as a synonym for action. The proposal is to study what various actors (diplomats, states, international organisations, NGOs) do in the international domain, and then to generalise the findings. In focusing on these ‘doings’ – or ‘competent doings’ – it is possible, so it is claimed, to combine insights about the material and ideational world, insights about facts and values, and insights about freedoms and constraints. Instead of pursuing this inclusive approach via the unifying category of action, Stappert suggests that our practice theory is ‘narrow’ and ‘conservative’. In our view there are reasons for the existing discontinuities between different discourses. The strategy of reducing their internal complexity to a common problematic simply leads to oversimplification and trivial findings such as the claim that certain people (diplomats) do certain things (acts of diplomacy). But while we reject the methodology of bridge building, we think that communication across paradigm lines is possible and theoretically productive. The distinctiveness of discourses does not entail complete incommensurability.

The charge that our approach is narrow seems to conflate the narrow with the basic. Our analysis of practices is basic in the philosophical sense of the term: we seek to identify the core properties of a practice as an analytical object (see above). It is not part of our task to catalogue every possible meaning of the term practice, every possible discourse in which this term (or terms) may circulate, or every possible form of practices: doing this would only result in confusion.

The attribution of conservatism presumably means that our brand of practice theory cannot theorise change. As the readers can see, the entire second part of the book is devoted to the problem of change, construed as ‘dialectics’ in the Hegelian sense. By this term Hegel referred to an internally driven process of contradiction, be it logical or phenomenological, that transforms an identity. In our case we consider the transformations within and between the two global macro practices – the practice of sovereign states and the practice of global rights – that determine the institutional structure of the international realm today. Changes of this sort may be positive such as the proliferation of a discourse of rights expressed in the doctrine of humanitarian intervention, but also negative ones such as ruthless economic competition among global actors, including global private security firms.

Another criticism of our account is its alleged reductionism, a term that Stappert leaves undefined. In the philosophy of social science reduction stands for the procedure of expressing the propositions of one discourse in the terms of another, more fundamental one (with the help of connecting principles): chemistry thus can be reduced to physics (Nagel Reference Nagel1961). It is easy to see that our approach is antireductionist. In PTIR we defend a thesis of practice differentiation: like Wittgenstein and Oakeshott we reject the idea that practices have a common bedrock and emphasise their irreducible distinctiveness. For instance, following the music of Mozart requires a kind of understanding that is completely different to following a cricket match.

One of the points raised by Stappert is that we have not considered the full range of theorists that have contributed to the ‘practice turn’ in IR. She points to two empirical papers on the Cold War (Schindler and Wille Reference Schindler and Wille2015) and maritime security (Bueger Reference Bueger2013), and a third one (Hopf Reference Hopf2018) that was published after our book. In the material we have read so far, there is no evidence of a compelling, theoretical account which rebuts our central points in the book about the need for understanding (international) practices from the internal point of view, about the merits of descriptivism, about the centrality of constitutive rules and rule-following. Similarly, we find Stappert’s charge that ‘the discussion of the practice turn is so sweeping that it is puzzling to read’ unsupported. Compare Adler and Pouliot’s (Reference Adler and Pouliot2011c) single page outline of the components of practices with our analysis that spans three chapters (PTIR: Part I). Pouliot’s discussion of Bourdieu’s theory of practice is five pages long (Pouliot Reference Pouliot2010: 31–6), ours is 16 pages (PTIR: 63–79). But the length of a discussion is not per se decisive. A central concern for us is that the readers should understand our philosophical argument. An argument is a line of reasoning that connects premises to conclusions, by using consistent terminology. As such it is not a list of concerns: it is a form of analysis disciplined by criteria of consistency and relevance. Our book is not a survey of the literature on practices. We discuss a few of the major theorists in order to make our case for a specific, internalist understanding of social practices.

Relatedly, it may be helpful to clarify our assessment of scientific realism. Scientific realism is a metatheoretical platform in IR theory whose premises represent a form of scientific structuralism. Practice theory embodies normative structuralism. So contrary to Stappert’s reading, in Chapter 6 of PTIR we were not engaged in a ‘response’, and certainly not in a hostile response, to our colleagues from the scientific realist camp. In spelling out the distinctions between these two major metatheoretical alternatives our task was to differentiate theorising grounded in science from one grounded in normative analysis.

As Stappert rightly points out, our analysis brackets international history. This is not because we think that historical explorations are not worthwhile. Quite the contrary, fully to understand either global practice (the practice of states and/or the global rights practice) would require a detailed exploration of its historical evolution. Nonetheless, historical enquiry and philosophy represent distinct realms of discourse. Our philosophical aim in the book was to describe the general architecture of the contemporary macro practices in which we all find ourselves participating, directly, as actors or indirectly, as observers. Stappert suggests that our case would have benefited from a more fine-grained analysis instead of limiting ourselves to their bare bones. True, but such detailed accounts require, as a starting point, a thoroughgoing acquaintance with the basic constitutive rules of these practices. Any attempt at detail would fail absent some knowledge of the general architecture.

Finally, the question is raised, what, if anything, our practice theory has to say about the role of international law in world affairs. This is an invitation to articulate a ‘practical’ conception of law and of international law. On this conception law represents a set of rules that have to be, in some measure, understood and accepted by a social group. This is in sharp contrast to the Austinian conception of law as a command of a sovereign backed up by sanctions. But law must have its own differentia – according to some authors it consists of norms that enable the authoritative resolution of uncertainty in moral systems (Hart Reference Hart1961). Others see law as comprising norms of justice which therefore it would be morally right to enforce (Kant Reference Kant and Gregor1996). On our premises, international law is a constitutive component of the two major global macro practices. With respect to the practice of sovereign states it involves juridical rules governing the conduct of diplomacy and of warfare. It also specifies, inter alia, what counts as a sovereign state, a border, or a properly constituted government.

Anthropological universals (Jorg Kustermans)

As we have seen, it is important to distinguish between different meanings of the term practice. In his commentary Jorg Kustermans highlights the difference between ‘practice’ defined as ‘all of our doings’, ‘practical knowledge’ defined as ‘our skill at doing things’, and ‘practices’ defined as ‘what we do’. These three definitions correspond, we take it, to the idea of praxis (as opposed to theoria); the idea of meanings that must be learned and used in the process of performing an activity; and the idea of actions (meaningful doings). These are useful distinctions. What we hasten to add is that there is yet another, distinct idiom of a practice (‘practices’ in the plural) meaning social institution akin to a Hegelian custom (Sitte) or a Wittgensteinian language-game. Thus, in order to understand a large class of actions that have a special salience in our common social life, it is necessary to have antecedent knowledge of the pertinent practice (social institution or language game) and thus to know – or to have a relevant description of – the concrete practice defining rules.

Kustermans endorses our descriptivism but challenges us for not adopting the methodology of ‘thick description’. This methodology for understanding artefacts, including practices, was introduced by the anthropologist Clifford Geertz (Reference Geertz1973). It is an internalist, but unstructured (by rules, for example) procedure of interpretation that posits a multiplicity of interpretive layers revealing ever finer details of the object under investigation. Once again, we would underline that moving towards thick description is only possible once one has mastered the bare bones of the practice under analysis. Analogously, the finer points of chess cannot be grasped unless one has already learned the basic rules of the game.

Kustermans expresses a worry about our language of spatial metaphors and recommends that it be replaced with a functionalist language. In describing practices as ‘realms’ or ‘domains’ we do not imply that they are territories that exist side-by-side but spheres of human activity with distinct logics. A central task in the book was to outline the architecture of the social constitution of our freedom as individuals. Modifying Hegel’s account in the Philosophy of Right (1967 [1821]) we distinguished between five kinds of constitutive practices – the practice of family life, the practice of civil society, state practice (expounded by Hegel), the interstate practice, and the global rights practice. These practices constitute the identity of individuals as family members, civil society participants, citizens, sovereign states, or rights holders in a global realm. As identity-conferring devices constitutive practices defy any functionalist language. The reason why a function cannot uniquely determine an identity is that actors with different types of identity can perform an identical function – to wit, a mathematical computation can be carried out by a person but also by a computer.

We are opposed to foundationalism and to foundationalist ethical theories. Kustermans contends that in rejecting the case for a world state in favour of an order among free and equal states we must have smuggled in a foundational ethical principle. Notice that our claim is a conditional one – it states that if freedom is a value, then it can be actualised within a pluralist, ‘anarchical’ international order (society of states), not an hierarchical one (empire).

A further puzzle is what enables us to compare different societies and social practices? For it seems that the differences among practices presuppose some shared features. Kustermans points out that writers like Nicholas Onuf, Reinhart Koselleck and Peter Winch have invoked ‘anthropological universals’ including notions of birth, sexual relations, and death that are found in any human society. Here we would stress that the meaning of these universals varies in line with the internal point of view of the particular society being investigated. To talk of such universals in abstracto, as if they were removed from a particular social context, is about as useless as saying that all games have rules. This tells us nothing about the meaning of chess, rugby, boxing or bridge.

Kustermans holds that anthropological universals can be identified empirically. However, the ‘anthropological universals’ identified above are not empirical findings; they are features of our practice as English speakers. When a competent English speaker employs the term ‘society’ it normally refers to certain rules, including those regulating birth, sexual relations, and death. Somebody who denies this is simply not understanding English. But learning how to use the English language properly does not tell us anything about how specific societies have constituted themselves over time. To gain an understanding of such societies it would be necessary to adopt the internal point of view and come to learn the relevant descriptions of their core practices. This is true of our present global social arrangements and all the other meso and micro ones.

In a final point, Kustermans insists that we ought to ground our understanding of practices in extra linguistic reality. But what is this? We only have access to social reality through the descriptions embedded in practices: if our account bears scrutiny, there is no deep, metaphysical reality that overarches practices.

Practice theory and the study of IR (Amy Skonieczny)

In her commentary Amy Skonieczny asks about how our interpretation of the practice turn differs from those of others. To reiterate, the distinguishing tenets of our practice theory include holism, internalism, descriptivism and a focus on rule following. In our account practices are not just doings but normative domains structured by common rules and propped up by shared sensibilities. Hence, to understand socially determined actions, we must have antecedent knowledge of the pertinent social practice/social institution/language game. That is, it is requisite to know – or to have a relevant description of – the practice defining rules. We speak of relevant descriptions because different practices are determined by different complexes of rules.

A second question is how our theory advances empirical research in international relations. Because we are committed to radical interpretivism, we depart from the positivist view of empirical enquiry as consisting of brute facts that can be detected by direct ‘empirical observation’. Only a very limited class of actions such as throwing a bomb or dispatching a drone – we term them ‘simple actions’ – are amenable to direct observation. The crucial class of practice-dependent actions such as declaring war, policing a border, signing a ceasefire cannot be understood by direct observation: they are intelligible under a description. The description of an action, say ‘Russia violated the Intermediate-Range Nuclear Forces Treaty’, presupposes knowledge of the background practice, international law in this case, and its practice-defining rules and in particular the rule ‘treaties must be kept’.

In the book we often compare practices to games like chess, bridge, football, and so on. This analogy is useful because it allows us to spell out the concept of constitutive rules – rules whose point is to define the players and stipulate their roles and rights. Observing a putative action, a hand moving a piece of black wood over a chequered board, does not count as a move in a game of chess unless we already understand the rules of chess. Observing people in uniforms moving on a road in a non-violent manner would not register as troops crossing an international border without an understanding of the rules of international law (which define ‘international border’) and the rules of war and diplomacy (which specify how armies may be deployed in peace time across borders).

Of primary importance, though, is not the similarity but the difference between participating in a practice and playing a game. In chess, for example, there is a finite end to the game in sight. But practices are ongoing social arrangements: they evolve thereby raising new problems that require new solutions. Social arrangements of this kind do not ‘go out of business’ figuratively speaking. Because conflict and contestation take place inside them, they are subject to internal transformation or dialectical change. One such change in the current practice of the society of states, which originally granted rights to states only, is the incorporation of human rights. This modification qualifies the sovereignty right of states who are now required to respect a minimum set of basic human rights (a right to life and bodily integrity).

All contributors to this forum ask what our practice theory implies for the world of international relations and for the study of IR. The answer is that we are sketching the core institutional contexts within which certain ‘going-ons’ (to use Oakeshott’s term for action) in the international sphere such as Brexit, Trumpism, Russia’s contemporary foreign policy decisions are taking place. In particular, we draw attention to two macro practices – the practice of sovereign states and the practice of global rights – that represent such contexts. The claim is that in seeking adequate interpretations of actions, passages of action, or the implementation of policy decisions, one has to start with a basic comprehension of these two global practices. This represents a holist methodology that locates the ‘parts’ inside a broader ‘whole’. There is no way in which a part, some single action, can properly be understood in isolation of the whole in which it is constituted as meaningful. It is not a matter of studying an action and then placing it in some wider context. Rather understanding of the action is only possible with an understanding of the practice in which it is constituted as an action of that kind.

Last but not least, with respect to the landscape of IR theorising, our practice theory bears a resemblance to both normative theory and to constructivism, and yet is not reducible to either. Like constructivists, we assume that identities and interests are socially constituted. Thus, states do not have predefined interests such as the pursuit of power or wealth, as is assumed by classical IR theories in the realist and liberal traditions. The relations between states members of the EU are vastly different to the relations between Taiwan and China, say. But while constructivism remains a label for a diverse set of theoretical positions, some of which make concessions to positivism (Katzenstein Reference Katzenstein1996) or rational choice and game theory (Wendt Reference Wendt1999), from our theoretical standpoint concessions of this sort are precluded. The theory of practices outlined in PTIR makes identities (including state identities) dependent on the concrete normative domains within which the actors are constituted as participants. What we propose to study, then, are not just isolated norms but entire normative systems.

Concluding remarks

In conclusion, we wish to highlight three tenets of our practice theory. The most perspicuous one is that the global practices we have discussed are common or shared practices. We are all involved in these practices whether we wish it or not (this comprises their coercive, non-voluntary character, see PTIR: 191, 193–5). There is no ‘outside’ to the global world constituted by these two macro practices. On this view, liberal democracies, rogue states and the so-called failed states all share a common normative space not because they occupy analogous positions in the game (they do not), but because their disagreements can be understood by reference to a common body of settled norms.

Second, what this shows is that our commitment to the internal point of view does not entail relativism or disconnect between practices. Relativism is the idea that different societies espouse different social conventions and rules and therefore it is impossible to pass value judgements across societies. It is important to register that when we speak of disagreements between actors in world affairs, we are not taking the relativist position ‘One man’s terrorist, another man’s freedom fighter’. Indeed, we insist that it is possible to draw judgements about right and wrong in the moral sense because, in virtue of having learned to make sense of moral utterances, we inhabit a common moral world. This capacity for judgement is not negated by the fact that we simultaneously participate in a plethora of distinct social practices. Thus we can discern the moral wrongness of torture, mass exodus or the killing of innocent civilians in war, and the onus of proof falls on those who wish to contest these settled moral principles.

As theorists who study practices as normative systems, we object to the positivist fact-value distinction. Practices cannot be understood on the model of observation employed in the physical sciences. We are equally sceptical of the premise of rational choice methodology where the actors are assumed to have preferences over a predefined menu of choices whose ordering is value neutral. Our position is that choices are value laden and that even purely artificial individuals, states, can be understood as participants in practices animated by values. One central value that permeates both the practice of sovereign states and that of global rights is freedom. Any argument that declares sovereignty to be obsolete (Buchanan Reference Buchanan2000) therefore has to consider the implication of a fundamental value shift from freedom to alternative values such as managerial control or risk reduction.

The general thrust of our analysis in the book has been that a proper understanding of the actions of international actors requires attention to the larger social systems within which these actions are nested. This is not just a change in the level of analysis from micro to macro. In more basic terms, we set out to show, against interest-based theories of conduct, that practices determine the identity of actors and not merely their well-being. Analogously, against consequence-based theories we argued that agents do not modify their conduct in light of anticipated adverse consequences or positive rewards. A state whose sovereignty is being questioned (think of Brezhnev’s doctrine of limited sovereignty with respect to the Eastern European states during the Cold War) is not merely facing the prospect of suffering economic hardship or threats to its national security but an identity crisis. The participants in our global practices are constituted through commitment to norms and statuses sustained by elaborate forms of mutual recognition. In effect, we offered an Hegelian critique of power politics in the name of identity politics.