The road (not) taken? How the indexicality of practice could make or break the ‘New Constructivism’

Introduction

Over the past few years, what began as a narrow, if ambitious, methodological call to ‘return practice to the linguistic turn’ (Neumann, 2002) has developed into a far-reaching project for a ‘New Constructivism’ (McCourt, 2016). This ‘New’ edition would restore empirical inquiry to the heart of the Constructivist project within International Relations (IR) by firmly anchoring the analysis of the ‘International’ and its changing nature as a social order in a close examination of the practical activities by which it is enacted and reproduced. In providing a methodology for producing ‘experience-near’ (Jackson and Nexon, 2013) knowledge, it would correct ‘scholastic’ (Pouliot, 2007: 365) biases towards meta-theoretical debates and ‘armchair theorizing’ (Neumann, 2002), and provide a platform for anchoring theorizations of the International in its constitutive everyday contexts.

Since its first articulations, the practice turn and calls for a New Constructivism have retreated from the somewhat strained claims to dissolve the discipline’s meta-theoretical and philosophical ‘wagers’ (Jackson and Nexon, 2013) at the figurative stroke of a pen, which had attracted some of the most emphatic criticisms (e.g. Ringmar, 2014). What has received less critical attention is the methodological strategy that underpins this ‘practice turn’. By this, I do not mean particular methodological choices or specific procedures adopted by particular practice analysts. Rather, I am interested in the reasons for and the consequences of the pivotal claim that practice and the ‘implicit understandings’ or logic underlying and ordering it can be observed directly, preferably by some form of immersion in or proximity to its ‘natural’ contexts.

It seems that the time is ripe for asking some hard questions about this methodological strategy. From initially rather eclectic forays into different varieties of practice theory, subsequent debates have produced a discernible core understanding of what the IR practice turn is all about — but that has not yet hardened into any definite form of normal science. This seems an apposite moment to ask whether the strategy that is becoming apparent can deliver on the promises and objectives that it formulates for itself.

The starting point and motivation for such a ‘New Constructivism’ arguably stems from the difficulties that IR has experienced in coming to terms with the changing nature of international order, and the accompanying theoretical and meta-theoretical throes that this has provoked within the discipline. The changing empirical landscape of the International has been mirrored in a confusing variety of deeply entrenched camps and ‘-isms’ (Lake, 2011), which have put increasing strains on meta-theoretical and methodological but also substantive empirical and theoretical dialogue (Adler and Pouliot, 2011; see also Sil and Katzenstein, 2010). The evolving and increasingly heterogeneous forms of orders that now make up the ‘International’ (Neumann and Sending, 2007) continue to undermine attempts to formulate a consensual model of inquiry that could reconcile scholarly methodological and theoretical concerns with connectivity to practitioners’ perspectives (e.g. Friedrichs and Kratochwil, 2009). The proliferation of novel actors and forms of authority, and the emergence of a patchwork of partial but interconnected orders, have translated into strong (meta-)theoretical pressures to ‘foreground ontology’ (Jackson, 2008) in order to identify appropriate building blocks for a theorization of such multiple formats of International order. Empirically, it has entailed pressures to examine more carefully the practical forms of intersubjective ‘world-making’ (Onuf, 1989) by which this multitude becomes transformed into a broader form of social order, whose exact contours more decontextualized, static and general theoretical frameworks have found difficult to capture.

The project of a New Constructivism and of bringing practice back in promises to develop a new (this time, for real) ‘middle ground’ (Adler, 1997) on which these problems could become, if not resolved, then at least the object of a more productive dialogue. By focusing attention on the empirically tractable level of social practice, it would provide a platform from which processes of world-making, and their ordering effects, would become observable. This move would enable us to avoid the Scylla of paradigmatic imposition of ‘deep’ analytical premises preordaining the nature of the order to be discovered, while also steering clear of the postmodern Charybdis of endless inter-textual deconstructions and reinterpretations.

Such a New Constructivism thus (seemingly) offers a methodological strategy that, rather than postulating any a priori privileged epistemological or ontological standpoint (Jackson, 2011), would allow us to research such multiple spaces of social order on their own terms, but also to ask how they are nested in one another. It would also allow us to build an ontology of the International ‘from the bottom up’, reconciling the need for some degree of generality and connectivity with practitioners’ concerns (Friedrichs and Kratochwil, 2009; Lake, 2011) with a sensitivity to the multiplicity of knowledges and voices that nowadays constitute the ‘International’ (Aradau and Huysmans, 2014; Leander, 2016).

However, what is striking about the methodological strategy that this project has articulated is how conspicuously it remains tied to the meta-theoretical debates and divisions that it seeks to transcend. In particular, it promises to make the indexicality of practice — the fact that practice and its orderliness are, for both participants and observers, bound to specific context(s) and comprehensible only in and through context — the methodological basis for an empirically grounded New Constructivism. While this clearly resonates with an IR audience, the emphasis on immersion as the means to make practice logics observable is striking in its one-sidedness when compared to methodological debates in other disciplines, which conceive of immersion in natural context as one (crucial but not sufficient) methodological tool among others.

In the first section of the article, I show how this black-boxing of crucial methodological requirements for making the logic of practice explicit results from the attempt to plausibilize practice theory in terms of the very same meta-theoretical stand-offs that it was meant to transcend. In the second section, I show that the notion of indexicality (or context-boundedness) of practice on closer inspection entails methodological implications radically incompatible with its assimilation to IR discourse. Finally, I problematize the stark methodological separation between the first-order analysis of ‘situated’ practice and its second-order contextualization within the ‘broader sociocultural and political-institutional frameworks within which they operate’ (Sending et al., 2011: 532). Through a close analysis of recent, applied practice-theoretical work, I show how this separation undermines our ability to examine the lateral relations between practices ‘nested’ (Neumann, 2012a: 60) in one another — and thus our understanding of the social dynamics that result from such interrelations. The works discussed — Neumann’s, as well as Adler-Nissen and Pouliot’s work on diplomacy as an everyday site at which the International is enacted — are selected for the clarity with which they illustrate my main point. The specific methodological problems and omissions that I discuss are the result of the more general methodological strategy rather than simply a consequence of the specific technical differences between the methods adopted by any individual scholar.

My goal is thus to demonstrate that the ‘New Constructivism’ has been so concerned with shoring up the storyline that interpretation is possible without ‘forcing meaning upon’ its objects that it has ended up black-boxing the indispensable role of reflexivity (or, more methodologically speaking, achieving contrast on the practice to be analysed) for the process of interpretation. Building on this, I then also clarify the methodological preconditions for a concrete, experience-near analysis of practice — on the premise that ‘nothing is more necessary to comprehending what interpretation is, and the degree to which it is interpretation, than an exact understanding of what it means — and what it does not mean — to say that our formulations of other peoples’ symbol systems must be actor-oriented’ (Geertz, 1973: 14, emphasis added).

Praxeology’s programme: How ‘foregrounding ontology’ led to ‘experience-near’ theorizing

What is it that the practice turn does for or, rather, within IR? Such a question is best answered genealogically — by situating the practice turn in the disciplinary context from which it emerged. As is well known, since about the mid-1980s, the Constructivist critique had denounced the unacknowledged ‘world-making’ (Onuf, 1989) of positivist research. Against the search for observed regularities and constant conjunctions by imposing theoretical priors (in the form of concepts and hypotheses to be tested) on decontextualized data, Constructivists advocated constitutive analysis of the ‘social construction of reality’ (Berger and Luckmann, 1971). This implied making processes of world-making amenable to empirical analysis by reconstructing the (unobservable) meaning-structures instructing it, rather than prescribing the structures of the world to be discovered through our categories of analysis (Wendt, 1999: 47–91).

By focusing on understanding constitutively (rather than just explaining causally) (Hollis and Smith, 1991) how actors make worlds, and analysing the structures of the worlds that they make and remake, we would thus be able to:

re-appropriate [the concept of the International], to interrogate and problematize it, in order to (possibly) re-orient and reconfigure the practices that have constituted the ‘international’ as an abstract space distinct from the national space and supposedly governed by its own rules of functioning. (Bonditti, 2017: 6)

The Constructivist focus on constitutive analysis thus promised a mode of inquiry more flexible and open conceptually than the positivist testing of static and prefabricated categories of analysis on decontextualized data. By foregrounding the question of the ontological structure or texture of the International, it promised not only to keep up with ongoing changes in these structures, but also to re-anchor theoretical reflexion in practitioners’ categories of practice (see Kratochwil and Ruggie, 1986).

However, constitutive analysis of the ‘real’ worlds made by practical actors has proven elusive. The key difficulties encountered were that inferring or reconstructing general ‘unobservable generative structures’ (Wendt, 1987: 350) from specific, observable actions raised issues of indeterminacy, and establishing such a link thus required considerable amounts of interpretation (cf. Doty, 1997; Wendt, 1987). As their lack of reflexivity had been one of the key Constructivist charges against positivist research (see Neufeld, 1993), having to supplant contextual and theoretical fillers to infer constitutive links between structure and agency triggered considerable soul-searching. Constructivist scholarship thus developed a prolonged and extensive preoccupation with clarifying the epistemological and ontological (meta-theoretical) stakes and preconditions of empirical and theoretical research into the social construction of international social order(s) (Hamilton, 2017; Reus-Smit, 2013: 595).

In order to limit the vagaries of interpretive inquiry, it was concluded that theoretical frames or lenses were required to conceptually stabilize the ‘building blocks’ (Guzzini, 2013) of socially constructed reality for analysis. As a result, ‘foregrounding ontology’ (Jackson, 2008) and ‘ontological theorizing’ (Guzzini, 2013) were put forward as strategies for putting the genie of reflexivity back in the bottle. Even if identifying the ‘building blocks’ of the social prior to substantive inquiry could not wholly neutralize the indeterminacy of interpretation, it could provide a scaffolding for reducing possible sources of distortion, both by securing a joint intersubjective grid for analysis (Reus-Smit, 2013: 590) and by ensuring the basic commensurability of this grid with the implicit structure of social order(s).

However, attempting to stabilize fuzzy constitutive meaning-structures by providing ontological form did not quite prevent the imposition of outside meanings on analysis as hoped. By emphasizing the role of decontextualized, general analytical form over situated content, it encourages rather than constrains abstraction from meanings in natural context(s). It also encourages a turn to ‘meanings encoded in language’ (McCourt, 2016: 480) as a proxy for decoding processes of world-making. Foregrounding ontology thus facilitates a style of reasoning that fuels rather than minimizes (as intended) the ‘well-known perils of treating the social like a text’ (Ringmar, 2014: 10f.).

It is at this point that the so-called practice turn took off, rebelling against the penchant for ‘armchair analysis’ (Neumann, 2002: 628) and ‘scholasticism’ (Pouliot, 2007: 365). It identified (quite correctly) the failure to ground analysis of social ‘world-making’ in the practical activities through which it is carried out as a key obstacle to the Constructivist project. It also concluded (I think misleadingly) that this problem derived from the linguist turn’s penchant for analysing ‘meanings encoded in language’, rather than from the meta-theoretical preoccupations that the discipline had developed.

The practice turn promises to get to the root of the problems faced by IR Constructivism: the unobservability of the generative cultural or meaning-structures that inform agents’ world-making. It identifies the decontextualization of the process of interpretation as the key reason for this problem as it allows for the intrusion of various forms of ersatz-contextual fillers (whether from observers’ background assumptions or inter-textually), giving rise to the problems discussed earlier. The putative solution, then, is to observe world-making and meaning as it occurs indexically, that is, within (or, at the least, very proximate to) the ‘socially organized occasions of [its] use’ (Garfinkel, 1967: 3). By focusing analysis on such ‘natural’ context(s), one could observe the generative principles of world-making concretely at work, without having to adduce distorting contextual fillers and without compromising the structure or texture of meaning:

The focus on practices … solved, or at least transformed, the problem of how to link culture and action. If culture is only practices, the problematic relationship of culture to action disappears … practices are action, action organized according to some more or less visible logic, which the analyst need only describe. (Swidler, 2001: 85, emphasis added)

While the idea of social practice as a visible manifestation of culture is not specific to IR practice theory, it is crucial to note how thoroughly and seamlessly it has been mapped onto IR’s particular meta-theoretical topology. Practice analysis provides a solution to the meta-theoretical conundrums faced by IR if, and only if, it is assumed that making explicit the logic of practice is equivalent to solving the problem of indeterminacy in the relation between (observable) agency and (unobservable) generative structures.

Such an analogy is far from as straightforward as the IR practice turn has made it appear. It is therefore particularly important to be very clear about the precise nature of the tight assimilation of practice theory to IR’s meta-theoretical concerns. One of the key factors, I believe, is the family resemblance between practice-theoretical arguments and analyticist conceptions of interpretive research that have emerged in recent years within IR — with probably the most effective and influential formulation being Patrick Jackson’s (2011). According to him, the fact so troubling to IR scholars, namely, that ‘the very experience of the world is inescapably mediated by the conceptual and linguistic apparatus that we bring to bear when producing knowledge of the world’ (Jackson, 2008: 130), is not a problem, but part of the solution. Interpretive inquiry into social world-making, he argues, is possible because our knowledge practices are fundamentally ‘continuous with the world that they are investigating’ (Jackson, 2011: 114). Despite this, there remains an element of looseness of interpretation because any specific conceptual apparatus or ‘ideal type’ necessarily ‘accentuates’ particular facets of reality (Jackson, 2008: 146f.).

It is precisely this gap that the practice turn seeks to fill. While asserting (like Jackson’s analyticism) an interpretive continuity or commensurability between observation and its object, it remains wary of the ‘scholastic’ implications of requiring substantive theoretical concepts (‘ideal types’) for establishing it. Instead, it turns to natural, indexical meaning: by partaking in the situated context(s) in which actors instantiate meaning-structures, these constitutive meanings (though unobservable directly) become tractable for us, ‘100% empirically observable — publicly available and concrete’ (Andersen and Neumann, 2012: 472) — just like for actors themselves. Partaking in natural context thus allows for an ‘immediate, practical grasping’ of constitutive meaning, avoiding the distortions arising from ‘detached reflections on objects or activities’ (Andersen and Neumann, 2012: 472).

Andersen and Neumann have elaborated this practice-theoretical version of mind–world continuity by pointing out that we do not create theoretical representations of practices, but ‘practices should be the representation’ (Andersen and Neumann, 2012: 466), so that we ‘study practices directly, as a phenomenon rather than as a model of the phenomenon’ (Andersen and Neumann, 2012: 458):

As what practices are is specified in the model, and not dependent on the subjectivities of the objects of research, it is possible to study something as practice, despite the fact that the participants in it do not see it explicitly as one. Therefore, in using models, it would typically make sense to talk about practices as an ontological category. (Andersen and Neumann, 2012: 468)

Practice analysis thus interprets, but does so without ‘forcing meaning upon the practice it studies’ (Bueger, 2014: 389). All it does is to ‘make explicit’ the common presuppositions behind actors’ doings, the background knowledge that forms the scaffolding for coordinating actions into a joint practice.

The knowledge that we produce of the logic(s) of practice is thus representational, but ‘facts produced through the [ontological] model’ of practice remained aligned with actors’ implicit understandings or presuppositions by virtue of the continuity in terms of the natural, indexical context that makes them ‘visible’ (cf. Andersen and Neumann, 2012: 464). Pouliot (2007) has translated this meta-theoretical argument into methodological terms when he formulated a two-stage procedure for a ‘sobjectivist’ analysis of practice. According to him, we first recover situated meaning from its indexically bounded, natural context, which, in a second and separate step, we can input into formulating (theoretical) accounts of wider social or historical contexts. In this way, our theoretical accounts of the ‘International’ become anchored in the constitutive meaning-structures as we have recovered them through practice analysis rather than ‘imposing meaning upon them’.

As I shall endeavour to show, these attempts to assimilate practice analysis by analogy with IR’s meta-theoretical problems regarding the unobservability of ‘world-making’ entails other difficulties. These difficulties are not so much meta-theoretical, but concern the silences and omissions that this analogy imposes on IR practice theory and methodology with regard to what it means, concretely, to interpret practice. Specifically, while immersion in (or proximity to) practical activities is crucial for producing experience-near scientific knowledge, there is a second and equally important methodological condition for making explicit the logic(s) of practice, namely, achieving contrast to break through the normality of the ‘practical arrangements of [actors’] experiences’ (Andersen and Neumann, 2012: 458). As I will show in the next section, trying to cast practice in the role of the visible reflection of cultural structures confounds the problem of achieving contrast with the ancient villain of theory-laden observation, thus discrediting it as a proper topic for methodological reflexion. However, this methodological issue is particularly relevant to a field such as IR, where the social ‘domains’ whose structure(s) we seek to clarify are not naturally bounded (such as, for instance, organizations and bureaucracies), but ‘overflow’ the situated contexts provided by natural ‘sites’ (Burawoy, 2000; Candea, 2007; Marcus, 1995).

Indexicality and the multiplicity of practice: What can and does ‘experience-near’ really mean?

What, then, is wrong with a methodological strategy of immersion in the natural, ‘socially organized occasions’ (Garfinkel, 1967: 3) of practice and world-making? On the one hand, nothing: immersion or proximity is a crucial methodological component for their analysis. However, it is not the only crucial component.

At first sight, the logic underlying immersion in practice as analytical strategy seems rather compelling. Assuming that practical order is ‘organised around [actors’] implicit understandings’ (Neumann, 2002: 629), making the ‘experience of the participants, including [their] concepts’ (Andersen and Neumann, 2012: 470), the basis for experience-near theorizing of world-making processes seems only natural. It helps avoid the problems of theory-laden observation since ‘theorizing is a form of interpretation, and it destroys meaning. As soon as we begin to impose categories on evidence, the evidence stops meaning what it meant in its earlier context’ (Hopf, 2002: 25, cited in Pouliot, 2007: 368). Actors’ shared understandings thus form a sort of joint ‘coordinate system’ (Davidson, 1973), which provides the basis and precondition for their ability to assure the mutual adequacy of their doings in terms of this horizon, but also for observers’ identification of such a ‘logic’ of practice (Turner, 1994: 73f.; 2014: 30ff.). By reconstructing this coordinate system directly from its natural context, we preserve the texture of the ‘practical arrangements of [actors’] experiences’ (Andersen and Neumann, 2012: 458), in contrast to theorizations on the basis of decontextualized ‘data’ about it. Pouliot (2007) refers to this as a process of ‘objectification’ of the directly apprehended meaning-structure, whereas Andersen and Neumann (2012) speak of its ‘direct representation’. In a second step, we can then contextualize these representations to see what they can tell us about the wider institutional orders and social structures in which situated practices are embedded.

This assumes, however, that actors’ implicit understandings form ‘a particular set of meanings that are then taken as self-evident and prior to the practices themselves’ (Duvall and Chowdhury, 2011: 344). Practice analysis borrows the analyticist idea of observation and interpretation as direct observation of constitutive relations (Jackson, 2008: 146f.), so that our representations are of practice and its logic directly, and not interpretations of decontextualized data. In this way, we achieve ‘continuity’ (Jackson, 2011: 114) not between our own (background and theoretical) ‘conceptual and linguistic apparatus’ (Jackson, 2008: 130) and the world, but between our representations and the actual texture of practice as it becomes visible in natural context.

However, this assumption of a determinate, unitary (constitutive) relation between a generative structure (actors’ implicit understandings and the joint coordinate system that they form) and the ‘exhibited’ logic of practice is deeply problematic (Garfinkel, 1967: 99). As some contributions to the IR practice turn have also argued (see, e.g., Barnett and Duvall, 2005; Ringmar, 2014):

One can in principle always identify various performances as instances of the same practice in multiple ways, with no grounds to identify the relevant ‘practice’ (or its presuppositions) with any one of them. These alternative conceptions of the underlying regularity would … provide differing verdicts as to whether subsequent performances were in accord with prior practice…. Appeals to exhibited rules thus cannot resolve the difficulties. (Rouse, 2007a: 669, emphasis added)

A great source of confusion in this regard is that both observers of and participants in a practice must constantly presuppose such a constitutive relation in order to observe (and to know how to go on in) a practice. Garfinkel has usefully termed this operation ad hocing:

Ad hocing occurs (without, I believe, any possibility of remedy), whenever the coder assumes the position of a socially competent member of the arrangement that he seeks to assemble an account of and, when from this ‘position’, he treats actual … contents as standing in a relationship of trusted signification to the ‘system’ … he can ‘see the system’ in the actual content. (Garfinkel, 1967: 22)

However, the ‘system’ one sees in a practice as a means for decoding it is not ‘self-evident and prior to the practices themselves’ (Duvall and Chowdhury, 2011: 344; see also Garfinkel, 1967: 20f.). While ad hocing is a necessary operation for finding one’s way through a practice (and the world), it is not evidence for a unitary and determinate, constitutive (meaning-)structure ‘behind’ or ‘prior to’ practice. Ad hocing creates:

a persuasive version of the socially organized character of the … operations, regardless of what the actual order is, perhaps independently of what the actual order is, and even without the investigator [participant or observer] having detected the actual order. (Garfinkel, 1967: 23, emphasis added)

Practices turn out (just like rules in the linguistic turn) to be fundamentally indeterminate (Turner, 1994: 73). Treating them as self-contained units that can be understood on their own terms and that are relatively stable and durable (e.g. Adler and Pouliot, 2011: 6; Neumann, 2002: 633; Ringmar, 2014: 14) thus, indeed, means courting the ‘well-known perils of reading the social like a text’ (Ringmar, 2014: 10f.). As Garfinkel emphasizes, the accounts developed and given by both participants and observers of a (logic of) practice actually do not ‘represent’ such logics in a constitutive sense. What they do is to allow actors (and observers) to hold their performances ‘accountable’ by making identifiable and describable some doings as ‘deviant’ from a practice or as resisting its (ad hoc) postulated logic (Garfinkel, 2002: 35). However, they do not do so by constitutively extending or applying an actual, ‘prior, pre-existing’ logic; rather, they incrementally negotiate, define and specify a logic of practice by constraining it as it unfolds (Garfinkel, 2002: 35). The relation between a logic ‘of’ and ‘its’ practice is therefore negotiated and changing, and therefore normative, not constitutive (Rouse, 2007b) — there is no semantically substantive ‘shared understandings’ behind a practice, but only the ‘brute fact of agreement’ (Kripke, 2000):

‘Shared agreement’ refers to various social methods for accomplishing the member’s recognition that something was said-according-to-a-rule and not the demonstrable matching of substantive matters. The appropriate image of a common understanding is therefore an operation rather than a common intersection of overlapping sets [of presupposition or background knowledge]. (Garfinkel, 1967: 28f, emphasis added)

These insights pose significant problems for research strategies that one-sidedly emphasize immersion. Bourdieu (who is commonly appropriated as intellectual support for the IR practice turn) repeatedly insisted that a strategy built around immersion alone is unable to provide the analytical leverage necessary for decoding the logic of practice (Bourdieu, 2008: 34). He was expressly concerned with how it would favour what he called the ‘standpoint of the listening subject’ (Bourdieu, 2010: 1; 24), by which he meant a tendency to assume that practical performances correspond to and can be decoded by a unitary and determinate code. Where speech sequentially makes and remakes, and in the process incrementally (per)forms, meaning, the listening (or observing) subject treats the meaning conveyed in this way as if it formed a natural whole a priori. By treating performances as repeatable instances generated from a pre-existing code, the listening subject is ill-equipped for attending to the ambiguities, errors and deviations in which such a normally implicit code is revealed.

Rather than deriving constitutively from a shared, underlying code, the order of practice is an ‘ongoing accomplishment’ by actors constantly monitoring the joint definition of the situation, updating it to accommodate resistances and inscribe them into its logic as constraints. Garfinkel entertainingly and brilliantly illustrates the difference in his ethnomethodological account of trying to assemble a chair (speech/process) on the basis of the instruction manual provided (logic). As he shows, ‘what a particular instruction means can only be seen at a certain point in the assembly process’ (Garfinkel, 2002: 41). Rather than pre-existing performances, the logic of practice remains indeterminate until it meets a resistance that (in retrospect) would have mandated a different choice:

when and if I encounter an equivocality … I can wait for something later to happen with which I will possibly have seen what it was … that was relevantly there to be seen all along and was seeably there in the first place. (Garfinkel, 2002: 202)

What Bourdieu, Garfinkel and many others thus point to is a methodological problem that I would like to refer to (in order to convey both its processual and contrastive dimensions) as achieving contrast. In ethnographic, anthropological and sociological debates, it is often referred to as ethnographic distance, or ‘being a stranger’ to the context investigated (Alan Fine and Hallett, 2014; Neumann, 2012a: 186; Schütz, 1944), in order to be able to break through the ‘natural attitude’ and the implicitness of social practice (Garfinkel, 1967: 37). The IR practice turn, in contrast, has consistently sought to downplay the significance of achieving contrast for deciphering practice. Unfortunately, it is indispensable, even (and especially) in natural settings:

Establishing the reality of practices [requires] pointing to unanticipated orderliness [and] is itself dependent on the prior expectations of the reader, which are violated, and is relative to these expectations … the inquirer can ‘construct’ the tacit rules of another community only by a process of discovering that something done by the members of this other community is done differently or in an unexpected way.… The practice may then be ‘explained’ by showing it to be a variant of our own practices. (Turner, 1994: 36; see also Katz, 2001, 2002)

From the perspective of IR meta-theory, this must look like an invitation to theory-laden observation — precisely what we are trying to avoid! However, postulating an (ad hoc) ‘as if’ conceptual or symbolic structure for practice is, indeed, the precondition for decoding it (cf. Geertz, 1974: 30): such an analytical explication allows for the identification of errors and resistances and, on that basis, a refined and better representation of the logic of practice (Seifert and Hutchins, 1992). The fiction of ‘direct’ observation and representation is thus not only ontologically and epistemologically problematic, but methodologically counterproductive. The listening subject assumes such accounts to ‘directly represent’ the logic of practice, thus fundamentally misapprehending their purely methodological and processual function. It fails to take into account that the fact that such accounts have a role for decoding practice ‘does not mean that such descriptions are themselves … part of the reality they are ostensibly describing; they are anthropological — that is, part of a developing system of scientific analysis’ (Geertz, 1973: 15). Practices, then, are ‘not objects, but rather explanatory constructions that solve specific problems of comparison and unmet expectations.… But breaches in expectations occur adventitiously, and reveal no more than specific, limited differences. We cannot identify practice as such’ (Turner, 2014: 123).

This methodological principle has been famously and succinctly articulated by Alfred Schütz (1932: ch. 1) in his critique of Max Weber’s attempts to argue (transposed into a contemporary terminology) that there could be a form of direct or ‘objective’ interpretation of action (or practice) because of its ‘public’ and intersubjective nature. Schütz qualified Weber’s claim that practices or doings could be observed due to their invariant recognizability or ‘objective’ sense separable from specific ends and intentions (for a short exposition of this view, see Ringmar, 2014). He emphasized that ‘objectifying’ analogies to familiar situations provides a starting point for the analysis of situation-specific meaning; however, he also points out how further, subsequent doings not only have the potential to (quite radically) alter the (observed) meaning of practice, but make it decodable in the first place, allowing us to further revise, re-specify and refine our accounts of it.

Such accounts provide ‘vocabularies of motives’ that ‘stand for anticipated situational consequences of questioned conduct’ (Mills, 1940: 904), and that allow for formulating expectations as to the further unfolding of practice, whose confrontation with its actual unfolding thus provides instructions for modifying the initial account. Such vocabularies and situation accounts thus explicate the meaning of doings against a (postulated) joint context, and express actors’ strategies or intentions in terms of those contexts. Such an analytics of intentions thus in no way suggests that we need to get ‘inside actors’ heads’: ‘Intention or purpose … is awareness of anticipated consequence; motives are names for consequential situations, and surrogates for actions leading to them’ (Mills, 1940: 904; see also Ringmar, 2014: 12f.).

However, by abandoning the fiction of naturally bounded practical situations that could provide a determinate link between the ‘exhibited rules’ observed and some underlying logic of shared ‘implicit understandings’, these insights directly lead to a second methodological issue, namely, the problem of casing, the question of ‘what kind of sociological case one has, or establishing the sociological properties and boundaries of the situation at hand’ (Tavory and Timmermans, 2009: 244; see also Latour, 2005: 121ff.). This problem is widely discussed across the interpretive social sciences (Burawoy, 2000; Candea, 2007; Luhmann, 1994; Ragin and Becker, 1992; Tavory and Timmermans, 2009), but quite literally ignored by the IR practice turn. As we saw, our ability to make explicit the logic of practice depends not so much on immersion per se, but on following the actors through resistances they encounter and how they deal with them (Latour, 1988) — and on making the joint situation describable as it emerges through those contrasts. This means, however, that our problem is not just, or even predominantly, following the actors, but where to follow them — since different resistances offer different insights into the logic of practice!

Context therefore is not ontologically prior, but an analytical construct created by our methodological choices regarding which resistances and contrasts to create and trace — allowing us to see a, but not the, logic of practice. Whether contrast is achieved by techniques of securing ethnographic distance, by simply being a stranger (Alan Fine and Hallett, 2014) or by focusing on disruptions or ‘breaches’ in normality (Garfinkel, 2002; Latour, 1988) matters less than to realize that it is needed for detecting which ‘violations’ to the ascribed logic of practice can serve to further specify representations of this logic (Garfinkel, 1967: 20ff.; Turner, 1994, 19ff.). Whether ‘achieving contrast’ becomes an explicit methodological problem thus depends on the relation between the questions asked, the site(s) chosen and the specific background of the (participant) observer.

Iver Neumann’s excellent study of the bureaucratic modes of being a diplomat in the Norwegian Foreign Ministry (Neumann, 2012a) is a case in point. In this analysis, his previous professional background as someone concerned with conceptual argumentation rather than operative issues (as one of his interlocutors put it) provides a contrast position that enables him to dissect quite astutely and penetratingly the distinct professional relevancies of ‘being a diplomat’. It allows him to discern, in the diplomatic and bureaucratic practices within the ministry, a logic of consultation and speaking ‘in one voice’ rather than of conceptual innovation. Bureaucratic practices in the ministry are thus constrained by the (implicit) collective intentionality of producing speeches that ‘the entire ministry may stand for’ (Neumann, 2007). This situated logic, as Neumann shows, is bound up with a macrostructural logic of the International as a system of sovereign states in that it seeks to construct and represent the ‘national interest’ of Norway — a systemic logic that is increasingly under pressure as new actors, forms of authority and geographical fora start eroding the macro-order that it instantiates.

However, the assumption of there being constitutive logics or meanings-in-practice that can be determinately recovered by strategies of immersion (alone) can quickly become a liability — even for the very author of this excellent ethnographic study. In an article addressing the question of whether contemporary diplomatic practice is or remains Euro-centric (Neumann, 2012b), Neumann begins by decoding how contemporary diplomatic practice-logics originate from historical-religious cultural scripts hailing from a European ‘context’ or discursive domain. This procedure makes sense against the analyticist assumption that constitutive meanings exist and can best be decoded in their natural, original context(s); however, in the study at hand, this assumption quickly becomes a profound methodological liability. Neumann’s assertion that the diplomatic practices in question follow a ‘Euro-centric’ logic (a point he ends up casting considerable doubt on as analysis proceeds — without, however, being willing to completely jettison it) is substantiated (solely) by reference to and on grounds of the socio-historical context in which they originated, which is treated as holding the keys to a constitutive analysis of their inherent ‘logic’. Thus, while the question that Neumann asks is posed within the contemporary macrostructural context within which the practices are ‘nested’ (Neumann, 2012a: 60) laterally, the analysis is done against a different, ‘original’ historical context.

This mismatch between the context set up by the research question and the context from which the logic of practice is to be reconstructed leads to a number of follow-up problems. For one thing, as Neumann points out himself, very similar — even analogous — practices and logics of diplomacy have developed in other socio-historical contexts, suggesting that the ‘formative’ context inscribed in the logic of the practice analysed is not as clearly delineated as he presupposes. What is more, through careful historical analysis, Neumann reveals that the logic of contemporary practices of delegation is in important ways a hybrid between Western and non-Western cultural scripts and practices. While Neumann does show that the logic or constitutive ‘myths’ he uncovered ‘shine through in present-day diplomacy’ (Neumann, 2012b: 300), this is hardly surprising given that he set out to identify this present logic in past contexts in the first place. The real question of whether this logic could also be the result of the lateral embedding of practices (their being constitutively entangled with other diplomatic practices, forming a wider structurally constrained field of practices) rather than their diachronic origins falls by the wayside. These methodological tensions consistently resurface — yet, they cannot be properly addressed without exploding from within the meta-theoretical justifications and methodological strategy tailored to it that the IR practice turn has developed.

This mismatch between analytical contexts is somewhat accidental in the case of Neumann’s study. However, breaking apart the process of understanding of practices ‘as if’ they were fully contained within bounded contexts, and then contextualizing them in relation to the wider institutional orders and social structures in which they are embedded, has indeed been formulated as an explicit methodological injunction within IR practice theory. Therefore, we will need to dissect it in more detail in the next section.

The paradoxes of ‘situatedness’: How practice overflows its context(s)

As the previous section shows, the IR practice turn and New Constructivism have prioritized banishing the spectre of theory-laden observation and reflexivity over a more profound methodological engagement with the problems of interpreting social practice. Their express goal has become to ‘make the concept of practice an analytical one, rather than a hermeneutic one, and do away with the need to postulate unobservable “motivations” when doing empirical research’ (Andersen and Neumann, 2012: 458). Crucially, this has also meant trying to insulate practice analysis meta-theoretically from what its proponents perceive to be the greatest weakness of ‘hermeneutic’ interpretation: the (perceived) need and (actual) procedures of clarifying meaning by either adducing empirical background knowledge about the wider social embedding of practice, or by exploring such backgrounds inter-textually. Ironically, in defence against these perceived sources of contamination, the IR practice turn has ended up articulating a reified conception of practice contexts as naturally bounded, self-contained and interpretable on their own terms — that is, it has come to (implicitly) define practice in analogy to text (for a clarification of textual versus inter-textual interpretation, see Lotman, 1990: 11ff.).

The fiction of ‘naturally bounded’ contexts as a self-sufficient site for studying practice or action is, in fact, a theme of numerous micro–macro debates in the social sciences. This very distinction rests on the notion that ‘experience-near’ knowledge production means examining social (micro-)process, ‘conceived as a self-contained enterprise that is locally productive’ (Cicourel, 1981: 52). However, ‘neither micro- nor macro-structures are self-contained levels of analysis, they interact with each other at all times despite the convenience and sometimes the dubious luxury of only examining one or the other level of analysis’ (Cicourel, 1981: 54).

As Bourdieu, Latour and numerous others have shown, our ability to produce substantive empirical descriptions of the logic of practice requires us to ‘follow’ not so much the actors per se, but the relations in which they are embedded beyond the boundaries of seemingly natural contexts in order to be able to specify their internal logics. Trying to (fully) endogenize (Adler-Nissen and Pouliot, 2014) the analysis of practice to such contexts thus reduces our ability to make our accounts more ‘semantically contentful’ by describing the logic of practice in terms of its relational constraints. As a result, we may indeed produce less theory-laden accounts, but instead end up with purely formal descriptive categories that, lacking substantive semantic specification in terms of the constraints constitutive of a practice, remain fundamentally unobservable in empirical terms (cf. Turner, 2014: 3). As the example of Neumann’s work on diplomatic practice shows, this has profound consequences for our ability to empirically examine practices’ situatedness in the ‘broader sociocultural and political-institutional frameworks within which they operate’ (Sending et al., 2011: 533, emphasis added) — and thus the possibility of using the analysis of practice as a prism for understanding the changing nature of the International.

If practice were fully contained within situated contexts (or had invariant form across contexts), it could be exhaustibly accounted for as a form of rule-following or as commensurate with shared understandings (see earlier). By corollary, all the dynamics unfolding within a practice could indeed be explained as more or less (in)competent performances (see Duvall and Chowdhury, 2011). If, however, as we have seen earlier, the order of practice is the contingent outcome of how participants hold each other ‘accountable’ (thus specifying the logic of practice as they move along), (in)competence or (in)adequacy can neither be measured against an a priori, nor explain the power dynamics involved (except in purely formal terms). Instead, it becomes an explanandum, which has to be accounted for in terms of the constraints and modifications imposed on actors’ strategies by their encounters with unfolding, ad hoc context(s). Garfinkel has teased out this point with regard to how broader institutional orders are implicated and manifest themselves in practice contexts:

Institutional orders do not work through rules and rule-following, but rather by imposing accountability constraints on action. Instructed action is accountable action. But instructed action is also carefully oriented over its course. It is reflexive action. It is action prospectively constructed to meet constraints on praxeological validity. (Garfinkel, 2002: 41)

Institutional orders are thus interlaced with practice contexts through the (anticipated, encountered) constraints that actors impose on their own doings. As we have seen, we can make these constraints describable (only) by expressing them in terms of actors’ (assumed) intentions and strategies in terms of the ‘implicit understanding’ of the situation that actors place on their doings and that thus allow for expressing them. This, however, necessarily entails a conception of power and institutional order that not only ontologically, but also in methodological terms, overflows a vision of practice as ‘self-contained enterprise that is locally productive’ (Cicourel, 1981: 52).

To see why, let us turn to Callon and Latour’s (1981) explication of how actors ‘macrostructure’ social worlds and how this process bestows power on some, turning them into ‘macro-actors’, although they are ontologically no different from ‘micro-actors’. Macro-actors have power to ‘conduct the conduct’ of others not because of superior ability or competence, but because other actors (and their practices) are aligned (thus also shaping a joint context for each other) in a way that provides a context or system that is conducive to realizing macro-actors’ intentions through those of others. Their power thus derives from the structuredness of such contexts; they are empowered because others’ practices and actions are constrained — symbolically, materially and interactively — by being ‘nested’ in one another. This means, however, that even if power (very broadly conceived) is understood as allowing ‘actors to compel others to change’ (themselves and their practice) (Barnett and Duvall, 2005: 3), then ‘power-in-practice’ cannot be analysed as purely endogenous to situated intersubjective contexts (as attempted by Adler-Nissen and Pouliot, 2014).

Analysing power-in-practice (and thus the implication of wider institutional orders and social structures in practice contexts) means explicating in substantive empirical terms the ‘collective definition of the situation’ and the constraints inscribed in it — how and by whom they were manufactured, and which and whose practical strategies they benefit (Bourdieu, 1980: 94ff.). An empirically discriminating analytics of power must examine the origins of some actors’ ability to ‘manipulate the collective definition of the situation’ (Bourdieu, 2010: 40) by imposing constraints that force others to align their strategies with joint practice so that one can simply ‘go on’ oneself (Bourdieu, 2009: 43–65). Garfinkel’s (1967: 35–75) (in)famous breaching experiments, in which some participants in a practice were instructed to behave eccentrically or ‘incompetently’ so as to provide a disruption to what others treated (up to that point) as an unproblematic context and thus logic of practice, illustrate the implications of this point. Rather than simply disciplining the dissenters and restoring the order of practice, as the notion of practice as having a set logic more or less ‘competently’ executed by actors would imply, disturbed participants would seek to alter their definition of their joint situation in ways to accommodate deviant behaviour and adjust their own behaviour in accordance with what they hypothesized now to be a more correct situation definition. Power, in Bourdieu’s words, is thus ‘not a simple technical ability [or “practical mastery”], but a statutory ability’ (Bourdieu and Wacquant, 1994: 146) — it depends very much on the position that one occupies within a wider institutional and structural context.

Without such explicit situational analytics, power as an analytical category and, with it, our descriptions of how ‘broader social structures … enable and constrain practices’ (Pouliot and Cornut, 2015: 302) effectively collapse into other analytical categories such as competence, adequacy or appropriateness to a logic, instead of being explicated in empirically substantiated and clearly differentiable terms (see earlier). Instead, we are forced into an abstract meta-commentary of practice (and ‘power-in-practice’) in terms of analytical unobservables, that is, purely formal analytical categories that cannot properly be given substantive empirical meaning in terms of the evolving situations faced by participants in a practice. As we have seen in the case of Neumann’s study of Eurocentrism in diplomatic practice, the fixation on an alleged constitutive relation between practices and their ‘logics’ frequently works to black-box methodologically how their lateral embedding in broader sociocultural and political-institutional frameworks shapes and formats practices, and is crucial for understanding their role in wider structures of power and domination.

Not all analyses of practice will suffer from these problems to the same extent since the visibility of practices’ roles in wider systems of domination depends on the alignment of the researchers’ background knowledge with the sites and contexts chosen for inquiry and the questions asked. This means, however, that the best way to explicate the limitations for empirical inquiry entailed by the methodological strategy that I critique in this article is to focus on works in which these problems are clearly linked to this strategy itself, rather than to the more technical problems of methods.

Adler-Nissen and Pouliot’s (2014) study of the 2011 United Nations Security Council (UNSC) negotiations about international intervention in Libya provides particularly much analytical leverage in this regard. They deliberately develop an account of power as wholly ‘endogenous resource — social skills or competences — generated within practices’ (Adler-Nissen and Pouliot, 2014: 891). At the same time they explicitly contend that their analysis will not ‘obscure the broader context in which practices occur’ (Adler-Nissen and Pouliot, 2014: 890), but provide a better understanding of the ‘concrete workings of power in international politics’ (Adler-Nissen and Pouliot, 2014: 890). Through their focus on endogenous ‘power-in-practice’, they claim to provide ‘a deeper understanding of the everyday struggles through which particular skills become recognized as power resources and affect world politics’ (Adler-Nissen and Pouliot, 2014: 910).

Their methodological strategy for achieving this follows our reconstruction of the IR practice turn’s rationality almost to the letter:

In tune with practice theory, we refrain from using motives, which are often empirically intractable, as explanatory variables. Instead, we zoom in on the actual moves performed by national diplomats … in order to reconstruct the dynamics of influence that gave the intervention its peculiar shape and pace. (Adler-Nissen and Pouliot, 2014: 891)

As Pouliot (2016: 58) formulates it in his more recent book on International Pecking Orders, ‘in order to make sense of the struggle for practical mastery … one must get deeply immersed in the intersubjective context or situation at hand’, explicitly warning against the idea that ‘competence … can be assessed from an external point of view’.

One problem with this strategy becomes manifest already at the very surface. As we have seen, trying to explain the effects of power in terms of (relative) competence without carefully laying out situational accounts is difficult. Without explicating the specific configurations empowering certain actors over others, we are forced into forms of meta-commentary that invoke theoretical referents but cannot specify them in empirically substantive terms. Such meta-commentaries, trying to frame entire episodes as exhibits of more general theoretical claims without, however, anchoring these categories in the empirical texture of the episodes themselves, appear throughout their exposition. For instance, we are told that ‘once British leadership was recognized by others, this reputation for competence was easily cashed out in concrete influence’ (Adler-Nissen and Pouliot, 2014: 898), and that ‘recognized mastery of procedures turned into critical influence’ or ‘the skillful handling of this diplomatic opening paved the way to exerting power in practice’ (Adler-Nissen and Pouliot, 2014: 899). None of these meta-commentaries (in line with their avowed goal of avoiding motives) specifies accounts of situations and vocabularies of motives that would allow us to differentiate actors’ strategies in substantive terms. They thus do not achieve the contrast necessary for empirically differentiating the effects and sources of concepts such as power, competence or mastery in the episodes they exhibit. Instead, they ‘are treated as explainers, but are descriptions that need to be given some sort of additional explanatory background or force to actually explain’ (Turner, 2014: 3).

More structurally, the same problems appear within the analytical narratives meant to convey the practical ‘logic’ of the German and South African paths to (eventually) supporting the resolution championed by the UK and France (Adler-Nissen and Pouliot, 2014: 902ff.). As narrated in their article, it is exceedingly difficult to determine just what the correlates of key theoretical terms such as ‘competence’, ‘authority’ or ‘standing’ would refer to in terms of actors’ (implicit) strategic understandings of the parameters of the situation(s) in which they interacted. We learn that the abstention by German representatives in a first round of consultations (having been instructed by their government to do so) ‘seriously undermined Germany’s diplomatic standing as a competent player, in the eyes of its partners, and it consequently dented the country’s ability to wield influence in and through practice’ (Adler-Nissen and Pouliot, 2014: 903). While we are told that (as a result of this ‘exhibited incompetence’) Germany was subsequently compelled to engage practically in a way its partners expected of it, having to ‘“overcompensate” for its abstention, in an attempt to regain its lost authority and, potentially, its influence’ (Adler-Nissen and Pouliot, 2014: 903), we do not learn anything about how concrete subsequent German moves were actually constrained and compelled by situational logic. Likewise, we are not given any empirical criteria for what it meant for Germany to have ‘show[n] a delay in assuming responsibility for a country that has claimed greater responsibility in the maintenance of international peace and security’ (Adler-Nissen and Pouliot, 2014: 903).

This episode has all the potential for a fascinating analysis of how endogenous dynamics of power in practice are linked to broader dynamics of, and ongoing changes within, the International. There are repeated (if passing) references to ‘macro-’dynamics that could provide a background to inform concrete analysis of practice. The authors note in the abstract how ‘superior legal knowledge’ has become a source of authority constraining practice in a way incompatible with a logic premised exclusively on the representation of sovereign interests (Adler-Nissen and Pouliot, 2014: 896). They also note how maintaining ambiguity can become a source of power, or how language may serve as a form of ‘symbolic power’ (Adler-Nissen and Pouliot, 2014: 896). However, eager to avoid contaminating their analysis of contextually specific power in terms of the purely indexical, endogenous play of competence, Adler-Nissen and Pouliot choose not to engage such lateral background conditions that could help substantiate the practice logics that they are interested in.

By way of suggestion, the episode they recount could have provided intriguing insights into the role of ‘statutory power’ (see earlier), and the inscription of competing macrostructural logics into situational logics. As they do point out, German abstention was decided in Berlin, when the possibility of the US ‘going it alone’ was not yet on the table. Once this happened, Germany’s representatives in the UNSC became isolated as the representatives of other Western states moved to avoid the public appearance of a rift. A plausible interpretation could suggest that the US used its statutory power to performatively alter the collective situation in a way that favoured its own intended outcomes — by threatening a move that could have publicly weakened the ties of solidarity of the Western alliance, a price to high for the other players in the game to risk paying. Such a narrative — which is entirely compatible with the story told by Adler-Nissen and Pouliot — could also speak to a conflict, in practice, between different logics deriving from lateral context(s). The German delegation’s logic of purely ‘representing’ their government’s position, as in a classical conception of diplomacy between sovereign states, was thus invalidated by the US. It became unsustainable once the latter used its statutory power to unfold a course of events in which the logic underlying the German move had to give way to a rationality in terms of a shared ‘governmentality’, which can shift acceptable grounds of justification to more substantive registers and knowledges (see Sending and Neumann, 2006), and thus compelled a change in Germany’s strategy. Likewise, one could also interrogate the episode as to whether the specific logics of multilateral diplomacy (where the norm of argumentative rationality often trumps pure representational logics), or even simply a (perceived?) need to move along to preserve the outward legitimacy of the UNSC as a collaborative forum in which pure interest representation is not acceptable (a norm of rhetorical entrapment that seems to have been crucial to getting a number of other members of the UNSC to move along, but which receives hardly any analytical attention from Adler-Nissen and Pouliot), might have constrained German moves situationally.

Similar lacunae appear in the second episode analysed. In a complete reversal of their argumentation in the German case, the authors now suggest that South Africa’s support for the resolution made it appear profoundly incompetent to its African partners (Adler-Nissen and Pouliot, 2014: 904). Without an explication of how South Africa’s position was constrained by the linkages between its situated performance in the UNSC and wider institutional contexts in which it operates on other occasions, competence must be relied upon to explain diametrically opposed empirical effects compared to the first episode. While we are told that the South African attempt to frame non-intervention as the more humanitarian option ‘was skillfully framed as an unwillingness to defend human lives’ (Adler-Nissen and Pouliot, 2014: 908), we are given no criteria for deciding why one attempt to ‘manipulate the collective definition of the situation’ (Bourdieu, 2010: 40) trumped the other. Finally, while we learn that the Russian attempt to plug into a more traditional macro-logic of sovereign statehood to delegitimize intervention as interference in the internal affairs of a sovereign state failed (although it did not seem a lost cause from the beginning) (Adler-Nissen and Pouliot, 2014: 908), we never learn what substantively made the Russian attempt fail, nor why they would try such an obviously ‘incompetent’ move. ¹

In attempting to provide a purely ‘endogenous’ practice analysis of their case, Adler-Nissen and Pouliot have deprived themselves of crucial methodological procedures. Without the contrast(s) provided by a more explicit situational analysis, it becomes impossible to substantiate why the Western permanent members of the UNSC retain and execute their dominance — and thus how wider, macrostructural logics of the International are instantiated within this situated context. The case they examine contains multiple toeholds for achieving contrast in order to make the texture of practice describable and specifiable in this regard. However, the methodological strategy that they adopt means that they are invariably bypassed in the article — although the authors seem very well acquainted with relevant ‘background’ conditions.

The net result of this one-sided emphasis on immersion at the cost of a lateral analytic ‘embedding’ of practices in wider macrostructural contexts that could help make their logic empirically tractable is that it becomes impossible to analyse why practice unfolds in a certain way. This cautiousness about ‘imposing meaning’ on the analysis of practice is certainly understandable against the backdrop of IR’s intense meta-theoretical concern with the reflexivity of social-scientific inquiry. However, the shift to the opposite extreme of attempting to erect a methodological strategy on the fiction of ‘direct observability’ (due to the existence of an inherent ‘constitutive’ logic to practices) is no less problematic. The net result is that theoretical terms such as ‘competence’ remain abstract and decontextualized formal descriptors, decoupled from the substantive texture of the practices that they are meant to make empirically describable.

Conclusion

In this article, I have shown how the IR practice turn and New Constructivism’s attempt to banish the spectres of theory-laden observation have led to a one-sided methodological strategy premised on the fiction of a natural commensurability between practices and the structure of the contexts from which they ‘originate’. This emphasis is quite understandable as a response to the (partly perceived, partly real) ‘excesses’ (McCourt, 2016: 480) of the linguistic turn, and the sprawling meta-theoretical debates that have accompanied Constructivism’s march through the institutions. At the same time, it creates new obstacles with regard to the avowed objective of this project: to make the analysis of practice the methodological platform for a more empirically grounded, ‘experience-near’ theorizing of the changing logics, structures and world-making that make up the ‘International’ today.

In this article, I have above all attempted to give a precise diagnosis of what these obstacles are, and how they result from a widely shared (if partly implicit) methodological strategy rather than being merely accidental or technical. Situating this methodological strategy in the meta-theoretical debates that have motivated its formulation has allowed me to problematize some of its key omissions and silences. Rather than an end in itself, situating this problem genealogically and systematically thus helps (or so I hope) to break through the (seeming) normality of this strategy and recognize more clearly its severe limitations.

This has allowed me to unpack, in a second step, how the assimilation of practice analysis to IR’s distinctive debates has worked to reify practices and obscured that they are not ‘objects, but explanatory constructions’ (Turner, 2014: 123). As I show, this (deliberate) misunderstanding of the nature of practice has also entailed a black-boxing of the methodological preconditions for a purely ‘analytical’ use of practices as models or lenses, and thus undermines the possibilities of a lateral analysis of the ‘broader sociocultural and political-institutional frameworks within which they operate’ (Sending et al., 2011, 532).

This analysis has shown that the project of making the analysis of practice a prism through which we achieve better, experience-near theorizations of how heterogeneous processes of world-making — the forms of authority that they are based on and multiple forms of order that they instantiate — constitute the changing dynamics of the ‘International’ must make its peace with the crucial and methodologically indispensable role of achieving contrast, and thus reflexivity. The desire to foreground practice as an empiricist-analyticist ersatz-ontology in this regard seems symptomatic of a broader desire to secure, against all odds, a form of ‘epistemic sovereignty’ (Rouse, 2007a) for Constructivist inquiry that can banish once and for all the spectres of reflexivity and the endless meta-theoretical debates surrounding it. Despite its plausible ring to IR ears, this desire may end up doing more damage than good to the project of a New Constructivism (cf. Kessler, 2012): it may, paradoxically, reduce its ability to produce empirically tractable, experience-near theorizations of the International (to which it is, falsely, seen as a threat rather than a key ingredient).

Despite the (predominantly) bad name that it has acquired for wide circles of IR scholars, reflexivity is a fundamental methodological ingredient for the ability to distance oneself from the taken-for-grantedness of social worlds. As we have seen, this makes it a precondition for making them explicit: ‘The project of taking society as an object and describing the components of social life … [requires] positioning oneself outside this framework [of society] … in fact, a framework cannot be grasped from within’ (Boltanski, 2011: 7). In trying to formulate a doctrine of direct observability of practice that distances itself — not always overtly, but certainly in substance — from the reflexivity inherent in any such observation, the IR practice turn is in danger of cutting off the methodological legs on which Constructivist inquiry stands.

We need to leave this collective trauma behind, and, with it, the meta-theoretical and methodological fictions that we have produced to palliate it. Any account of practice and the making of social worlds that it entails is inevitably reflexive, and thus dependent on a position of externality (Reed, 2011: 22ff.). Any way of making a practice logic exhibitable is always ‘interested’ in (Bourdieu, 2010: pp. 2) and thus ‘interwoven’ (Reed, 2011: 16) with the making of social order. The explanatory constructions of practice that we use to interpret its logic, relying on them for reconstructing the implicit understandings and symbol systems under which we can explain actors’ practical activities, make some version of practice and social worlds more, and others less, ‘real’ (Law and Urry, 2004). They thus inevitably remain partial and, to some extent, help perform reality in ways we cannot fully control and cannot fully neutralize. Rather than try and deny this, the New Constructivism must face the fact that it is necessarily involved in ‘writing culture’ (Clifford and Marcus, 2008) — and start thinking of this as the methodological resource that it is, rather than a threat.