Towards a ‘theory of the gap’: Addressing the relationship between practice and theory

Abstract: In their ground-breaking philosophical investigation of the ‘practice turn’ Lechner and Frost prompt a standpoint debate in international relations theory, which touches upon the relationship between practice theory and its subject matter. Lechner and Frost decidedly opt for an internal standpoint, which promises to understand a social practice in terms of the meaning-in-use of its participants. This article argues that the internalist promise will ultimately remain unfulfilled, however, for the aim of collapsing the distinction between the ‘language of action’ and ‘language of observation’ is epistemologically impossible. Taking such an ‘internal’ perspective not only underestimates the problem of the double hermeneutic. It also disregards the gap between theory and practice. Any social enquiry that fails to acknowledge this gap inevitably becomes externalist, for it misses to reflect on its own normative presuppositions. The way ahead is to address this gap reflexively by way of a triple hermeneutics that is bolstered by abductive reasoning.


I. Introduction
Maybe the 'practice turn' in the International Relations (IR) discipline needed a philosophical investigation such as Silviya Lechner and Mervyn Frost's (2018) Practice Theory and International Relations to rectify the ontological bias prevailing in IR's contemporary practice theory. Maybe it needed a philosophical perspective on practices to clarify that a turn to practices can never solely constitute an ontological project, but is equally an epistemological one that seeks to foreground that background knowledge, however tacit or inarticulate it may be, normatively guides our everyday practices and is produced and meaningfully acted out in communities of practice. To be sure, IR's 'practice turn' does not need another theory of practice that loses itself in theoretical abstractions. Yet, this monumental work theorises both the ontological and epistemological building blocks of a practice in such thoroughgoing and illuminating ways that it challenges predominant assumptions in IR's 'practice turn' through a re-turn to some of its key inspirational sources in twentieth-century philosophy. With regard to ontology, the authors propose to particularly focus on constitutive rules and rule-following in order to understand how a practice becomes meaningful in the first place. Without a recourse to constitutive rules, the two argue, any action is bound to be meaningless, for it misses the social or normative standard that a community of practice sets for a practice to be appropriately, competently or correctly performed. Addressing the normativity of practices through a return to rules qua the ordinary language philosopher Wittgenstein is thus a timely and necessary advancement.
The book comes at a time during which IR practice theorists painfully realise that the question of normativity and critique is their Achilles heel (see self-critically Bueger and Gadinger 2018: 110ff.;Cornut 2017: 18;Gadinger 2016;Schindler and Wille 2019). To be able to address normativity it requires a conception of rules, which IR practice turners all too eagerly dismissed despite the prominence that rules took in early constructivist writings. 2 Nicholas Onuf's World of Our Making, for instance, already relied on Wittgenstein in 1989 to stress the central role rules play in the constitution of society and for constructivism as a metatheory. And it was Onuf who pointed to the nexus of rules and practices, highlighting that 'all rules in a socially constructed reality are related to practice ' (1989: 52) in that rules 'tell us how to carry on ' (1989: 51), while practices 'are the content of carrying on ' (1989: 52). Accordingly, we can already read in his early constructivist writings that practices are performed normativity qua rules.
The possibly political decision to erase rules and norms from the practicetheoretical vocabulary in order to push for another IR 'turn' and delineate the practice-based from norm-oriented constructivism now proves premature. As Lechner and Frost convincingly show, rules and norms are an intrinsic part of practice-dependent action, for they are the shared and normative standard of conduct that tells participants how to 'go on'. Thus, if we wish to understand not only that practices are the building blocks of the social, but also why agents consider a practice worth pursuing or where the shared sense of knowing what is appropriate or not comes from, we cannot get around a conception of rules. Lechner and Frost offer a holist conception of practices that meaningfully links these to constitutive rules that are here understood as norms or 'standards of conduct' (Lechner and Frost 2018: 115, 127). This conception enables them to be attentive to questions of community, and thus the identity and ethical status of practice participants, a dimension of practice which practice theorists tend to neglect (for an exception see Adler 2005;2019).
Raising the community question in times of late modern societies lacking the general consensus, shared beliefs or cultural membership of early and modern societies is, however, crucial (Rawls 2011: 399). So investigating how normative standards of conduct emerge and change in and through practice as well as how they endow participants with an identity that is 'realised as a concrete ethical status within a practice' (Lechner and Frost 2018: 131) promises to provide insights into how cooperation and community is globally possible. At the most basic level, then, Lechner and Frost seek to offer an interpretive methodology for understanding the meaningful standards or rules of engagement of a not-yet-understood practice. In concreto, they seek to investigate what the two macro-practices of the society of states and global rights mean for questions of global cooperation or conflict, identity and political community in the global realm.
At a more profound level, however, Lechner and Frost also promise ethical guidance for a social science that is itself undergoing a profound legitimation crisis, a crisis that is compounded by a surging populism that challenges established forms of expertise. This they do by prompting a standpoint debate in IR, which touches upon the relationship between philosophy and its subject matter, or between practice theory and the practice of its subjects. What standpoint shall the practice theorist adopt in light of the 'new obscurity' of global practices? One that is 'internal' to the 'language of action' of the historical practitioner or one that is 'external' to it and remains in the 'language of observation' of the theorist? Given the delicate links between practice and theory in practice theory, this is a decisive, yet thus far largely neglected question in IR. Lechner and Frost's intervention is thus timely and urgently needed. Grounding their standpoint in hermeneutic philosophy, and following in the footsteps of Wittgenstein and Oakeshott, Lechner and Frost decidedly opt for an internal standpoint. As they explicate, an internalist standpoint 'involves an attempt by a thirdparty observer to understand a not-yet-understood social practice in the terms its own competent participants themselves understand it' (Lechner and Frost 2018: 12). Otherwise, the theorist might risk applying or even imposing her own precepts or standards of evaluation, thereby evoking a meaning that has no relevance for the participants or, worse, distorts their intersubjectively validated meaning. 3

II: The internalist promise unfulfilled
This epistemological turn towards understanding a social practice from 'within' has been a key objective for IR practice theorists. After all, the 'practice turn' implied a decisive turn against the 'bias toward representational knowledge' (Pouliot 2008: 258) from which the social sciences, including IR, predominantly suffered. This change in perspective now allowed scholars to return to the long-standing insight that our understanding is not only embodied and practical, but that much of our meaningful action and norms guiding our 'sense of rightness' 'may be quite unformulated, or formulated only in fragmentary fashion' (Taylor 1993: 51). Accordingly, the 'practice turn' was about understanding world politics in its various multiplicities, about studying how seemingly coherent and abstracted wholes play out in concrete and immediate action contexts and take on different meanings depending on the situation. Practice theorists, hence, sought strategies that would be able to 'look down' and 'apprehend the local and the non-coherent' (Bueger 2014: 389) through a turn to people's everyday practices.
From the start, practice turners were confronted with the challenge of uncovering non-representational forms of knowledge that are 'not immediately accessible' (Bueger 2014: 388) despite their public nature. While bodily objects and artefacts could be directly observed, the implicit meaning required interpretation (ibid). How, then, were they to make sense of practices whose meaning the practitioners would sometimes not even be able to describe themselves because they would be at a loss for words (cf. Taylor 1993: 50-1)? Ostensibly aware of their position as theorists whose 'theorization destroys meanings as they exist for social agents' (Pouliot 2007: 364), practice theorists have sought to evade the danger of the theorist imposing her own scientific categories when interpreting the meaning-in-use by their research subjects. With Pouliot leading the way, they have therefore opted for an inductive logic of enquiry, which demands the researcher to immerse herself in the lifeworld of her research subjects. Contra deduction that would a priori fix the categories of meaning to make the empirics fit theoretical hypotheses, induction promises to 'recover the meanings of the world as it exists for the actual agents of international politics' (Pouliot 2007: 364). This inductive inquiry is precisely what Lechner and Frost advocate when they urge practice theorists to occupy a position inside the practice under investigation to understand what practice members actually 'do'. As they make plain, 'the language of observation must match the language of action'; '[…] the observer, in interpreting a practice, […] [must] use the categories which its own participants use in making sense of their practice' (Lechner and Frost 2018: 34). Lechner and Frost's critique levelled against Vincent Pouliot in Chapter 2 is therefore surprising, if not difficult to follow. When they particularly criticise IR Bourdieusian practice theorists for missing constitutive rules as the defining feature that makes practices social in the first place, their critique is most convincing (see e.g. 87-9). Yet, when it comes to their epistemological and methodological critique, their criticism loses strength -not because they fail at explicating their hermeneutic philosophy. But, ironically, because they fall into the same 'externalist' trap of which they accuse Pouliot. As it turns out, the internalism they promote takes one hermeneutic 'loop' too few and it thereby inadvertently becomes externalist. It therefore amounts to an accidental externalism, which is grounded in the authors' failure to reflect on the theoretical background that guides and moulds the research puzzle concerning the dialectics of their two macro-practices. While the authors seek to foreground the question of standpoint, they precisely omit reflection on their own standpoint, that is, their identity as theorists.
Consequently, both Lechner and Frost and their sparring partner Pouliot (2007; cannot fulfil the 'internalist promise', simply because internalism, defined in the sense that it 'collapses the distinction between observer and participant, between the language of observation and the language of action' (Frost and Lechner 2016: 304) is gravely dangerous, if not epistemologically impossible. Taking such an 'internal' perspective not only underestimates the problem of the 'double hermeneutic' (Giddens 1984: 284), and thus the interpretive leap that is needed to close the inescapable gap between the meaningful action of participants in a practice and the sense-making by what Lechner and Frost (2018: 12) call a 'third-party observer'. 4 It also misconstrues the relationship between theory and practice.

III: Practice and theory: Two distinct language games
It is one thing to find that 'science' is 'an interpretive enterprise' like any other practice (Lechner and Frost 2018: 208). Given their Wittgensteinian leanings, Lechner and Frost's aim to rid science of its superior claim to knowledge, if not truth, seems hardly surprising. Grimmel and Hellmann (2019: 205) have recently highlighted that Wittgenstein refused to accord the same 'privileged status to "theory" in philosophy as in the natural sciences' precisely because philosophy, just like the social sciences, engages with an already interpreted world for whose understanding the scientist has no superior access point or method at hand than the 'lay' person. In an IR discipline that is still pervaded by an intellectualist science that strives for objectivity, this is a highly rewarding move. Yet, it is another thing to argue that these two practices can be collapsed into one and that 'we must go no further than describing the self-understandings of the practice participants observed' (Lechner and Frost 2018: 211, emphasis added).
Regardless of the fact that science is a practice, science remains a practice with its own rules, norms and specialised 'language of observation' that can only approximate, but not equal the 'language of action'. 5 Gunnell (2013: 88) has made plain that the language that is in use in social enquiry reflects different affordances, social standards and logics than the language, which is in use by its research subjects. It follows that the two languages are 'autonomous' or 'conceptually distinct'; they therefore constitute two different 'language games' (ibid). This is where Lechner and Frost probably take the internalist procedure of observation and Wittgenstein's aim to 'describe' too literally, and thereby fail to recognise that any inquiry by a theorist is necessarily an interpretive, not a descriptive endeavour. 6 While Wittgenstein may have used the word 'describe', he was well aware that philosophy's task was to interpret, re-present and re-construct (Gunnell 2013: 88-9). After all, he made a clear distinction between the understanding (Verstehen) and interpretation (Deutung) of a practice (ibid: 88). As Gunnell (ibid) specifies, '[u]nderstanding is what takes place within a practice, and interpretation is a rendition of that understanding'. This crucial distinction I miss in Lechner and Frost's approach, a consequence of which is that the qualitative leap required from a theorist to move from understanding to interpretation is not problematised.
Apart from representing two autonomous 'language games', the distinct practice domains of the research subject and theorist also operate on different analytical planes that confront us as practice theorists with the well-known problem of the double hermeneutic: while participants of the practice under study already engage in a meaningful practice, practice scholars as theorists engage in second-order or third-order theorising. The result is not a 'description', as Lechner and Frost have it (e.g. Lechner and Frost 2018: 16-17). Much to the contrary, it always is a re-description or re-presentation (Grimmel and Hellmann 2019: 210;Gunnell 2013: 97). The theorist, hence, cannot 'just' describe; she always interprets against the background of her own taken-for-granted understandings and inevitably engages in normative judgement from 'outside' the 'language of action'. Consequently, the interpretation of the practice theorist will never be able to convey the same meaning as that which has been interpreted by the subject under study (Taylor 1971: 17).
Far from arguing that the 'language of observation' is neutral, I nonetheless argue that it is impossible to leave behind our 'language of observation' and replace it by the 'language of action'. This does not mean that we should not aspire to 'estrange' ourselves as far as possible to understand 'what the hell is going on'. But it does mean that our understanding and scholarly knowledge will always be partial because our theoretical and conceptual lens only allows us to see 'what we want to see' (Ravecca and Dauphinée 2018: 127). Believing we might 'become natives' (cf. Geertz 1973: 13 to the contrary) comes close to naïve empiricism and is only the reverse of scientific realism's claim that science is able to objectively reflect social reality. Unless the practice theorist decides not only to learn and understand, but also to become the practitioner of a given practice under study and remain within this domain of practice, she cannot close the gap between those two domains. She will be forever damned (or blessed!) to interpret the practice from the position of the theorist that operates within a specific theoretical framework, for she needs theory as a language to converse with her fellow scholarly community members. Even if she does not seek a 'correspondence theory of truth', as positivists would have it, the theorist is surely tasked to build a 'consensus theory of knowledge' among the members of a given academic community (Friedrichs and Kratochwil 2009: 705).
Without doubt, we must work with theory in the sense of 'sensitising concepts' or instruments (Blumer 1954: 7) to avoid 'superimpos[ing] an "explanation from the outside"' (Lechner and Frost 2018: 211). In the last instance, however, we cannot get around what Geertz (1973: 15) has poignantly called the 'imaginative act', indeed 'fiction', that is involved when theorists engage in second-or even third-order theorising. I add third-order theorising here because, irrespective of how close to 'the ground' we go in the microscopic study of intersubjective meaningmaking, we nonetheless aim for theory development that demands a third hermeneutic moment to translate or make one's interpretations of interpretations amenable to one's scientific community (Yanow 2009). Geertz has identified this tension 'between the need to grasp and the need to analyze' long ago, arguing that this tension is 'irremovable' (Geertz 1973: 24).

IV: The danger of the intellectualist bias
Despite Lechner and Frost's great effort towards recovering the internal point of view of practice members in international relations, I am doubtful whether, in practice, they can be truthful to both the subject matter of global practices and practice theory -or 'do justice' to the subject matter, as both Wittgenstein and Winch aspired to (Gunnell 2013: 84). Ultimately, adopting the internal viewpoint is honourable, for its wants to do justice to the internal standards a group of practitioners under study develop and have in use. Yet, it remains a utopian quest because it misses the theoretical baggage a researcher carries and ultimately disregards the epistemic constraints and customs of the scientific community, within which practice theorists are embedded. Refusing to acknowledge one's position in the field of production, one's identity as a theorist and its effect on the research subject, results in what Bourdieu once called the 'theoreticist or intellectualist bias' (Bourdieu and Wacquant 1992: 69, emphasis in original). As he details, it 'consists in forgetting to inscribe into the theory we build of the social world the fact that it is the product of a theoretical gaze, a "contemplative eye"' (ibid). This 'contemplative eye' is also at work in Lechner and Frost's work. Take the example of Pouliot's (2010) research on diplomacy in International Security in Practice, whose research strategy Lechner and Frost (2018: 80ff) chastise for its apparent externalism. Here, the authors' criticism is particularly directed against Pouliot's problem of 'missing identity criteria' that, according to Lechner and Frost (2018: 85), must be established before one looks at a concrete practice designated as 'diplomatic'. It follows that, as the necessary first step in one's analysis to determine who or what counts as a diplomat or diplomacy, they consider it of outmost importance to identify the constitutive or 'elementary rules which define a social practice' (ibid). For this identification, Lechner and Frost see neither the 'self-testimony of agents' nor the official status of a diplomat performing an action as sufficient proof of qualification (ibid). Instead, they propose that any conceivable action only counts as diplomacy, if 'it coheres with the canons of the practice of diplomacy' (ibid, emphasis added). The authors identify the 'canons' as Addressing the relationship between practice and theory 177 the constitutive rules of modern post-Renaissance diplomacy [which] include the system of embassy, the exemption of diplomats from local jurisdiction, the non-violability of the diplomat's person, and sovereign states as the principal (if not sole) agents represented by diplomats in international affairs (Lechner and Frost 2018: 86).
I have a hard time understanding these to be the constitutive rules of diplomacy other than to think that the above canons are an overly static and formal textbook re-presentation of diplomacy's legal framework. These are not the sole identity criteria that make diplomacy, simply because they are not the only rules that are of relevance to its participants and because they fail to capture the informal rules of the game that make it meaningful and worth pursuing (for an exceptional account to the contrary see e.g. Neumann 2005;2012;. Confining oneself to such textbook identity criteria that are abstracted from practice and abstain from a conversation with the research subjects cannot but remain externalist, for such an enquiry is guided by second-order conceptions that are primarily found in English School theorising (see paradigmatically Bull 2002Bull [1977).
I am not denying the overall usefulness of such legal accounts of diplomacy. Surely, the theorist should consult such representations in her study on diplomacy and use them as sensitising concepts that guide her provisional hunches in the investigation (see e.g. my proposal on an abductive research logic below). However, it is negligent to use these as the ground rule from which the study is subsequently developed and take this as the 'internally' operating standard. This reflects not only the 'inductivist fallacy' by which everything observed must be subsumed under an already 'known' whole and cannot generate any new knowledge (Reichertz 2014: 130). It also reflects Bourdieu's 'theoretical gaze' according to which the theorist or social scientist is blind to her own precepts that orient her. In the extreme case, such an approach risks reproducing the 'tendency [in rational choice theory] to allow the framework of analysis to function as if it were a presentation of reality' (Gunnell 2013: 91). The argument, then, to adopt an internal standpoint for reasons of 'getting the native's point of view' all too readily omits that we as scientists do not enter the theatre value-and judgement-free. The scientist is therefore tasked to once and for all bury the Grounded Theory Method 'myth' that scientists can enter the field tabula rasa as 'judgmental dopes' (see e.g. Kelle 2005: 45).

V: The gap between theory and practice -and how to cope with it
The only way out of this conundrum is to learn that even a practice theorist's account of social practices will never equal the practice at hand (cf. Bourdieu and Wacquant 1992: 70). There will always be a gap between theory and practice in the social sciences. Or rather: there will always be a gap between the practice of the practice theorist and the practice of her research subject, for the two practices are two distinct language games. To cope with rather than solve this problem, I suggest squarely addressing the gap that exists between the two practices. I take my cue from Bourdieu's conversation with Wacquant in which Wacquant summarises Bourdieu's plea for scholarly reflexivity by arguing that 'an adequate science of society must construct theories which contain within themselves a theory of the gap between theory and practice' (Bourdieu and Wacquant 1992: 70). I am not siding with Bourdieu's approach tout court, for I have my own reservations about his techniques of objectivation. Yet, I believe that such a theory is necessary -or better a 'thick methodology' (Pouliot 2016: 49). Such a methodology could reflexively address the gap by way of a 'triple hermeneutics' (see e.g. Friedrichs and Kratochwil 2009: 714) that goes beyond the mere understanding of a practice and moves towards rendering this understanding amenable to a specific research community, that is, practice theorists in our case. Rather than merely representing a 'metatheoretical "reflexive" stance' (Hamati-Ataya 2013: 681), this methodology provides practical guidelines for empirical enquiries by moving from double to triple hermeneutics and bolstering the enquiry by abductive reasoning. While triple hermeneutics adds a third, self-reflexive hermeneutic moment to the interpretive enterprise of the theorist, abduction celebrates rather than hides the interpretivist leap that it takes to reconstruct a rule in use (Reichertz 2014: 127).
Like double hermeneutics, triple hermeneutics problematises the relationship between the two contexts of action in which interpretation takes place. It problematises both the double interpretation that is involved when the theorist interprets the already meaningful world of the research subject and it takes into account that the theorist's 'analytical categories are intertwined with what Connolly calls the "terms of political discourse"' (Berenskoetter 2017: 156). In that sense, double and triple hermeneutics both advocate what Jackson calls 'mind-world monism' (2011: 35), a position that finds any rigid separation between the theorist and her research subjects nonsensical. Yet, and here lies the difference between double and triple hermeneutics, in the latter case, the theorist seeks to be reflexive about her own categories used and the theories from which she thinks and acts during both the fieldwork and desk study (Yanow 2009: 278-79). Here, the researcher neither dispenses with theory and fully immerses in the 'language of action' nor imposes her theoretical categories onto the subject from 'outside' of the context of action. Instead, she consciously works with theory as 'thinking tools' that help her construct the research puzzle as a 'problem', formulate a research question or approach her research subject in the field (Bourdieu and Wacquant 1992: 50; in anthropology see Hirschauer and Amann 1997: 37;in IR see Leander 2008;Wiener 2014, Chapter 5).
An abductive research strategy is able to reflect this reflexive use of theory and allows for the 'iterative-recursive play' (Schwartz-Shea and Yanow 2012: 28) between the empirical context of action and conceptprone context of the theorist, the latter of which is primarily one of justification. 7 Along these lines, theory is neither ignored nor readily applied to one's research subject. It rather functions as a useful heuristic or the above-mentioned 'thinking tools' that need to be reinvested to build the empirical. In a similar fashion, it was Herbert Blumer (1954: 7) who pointed to the significant role that concepts play throughout the research process: in his view, they function as 'sensitizing concepts' that remain vague and yet provide the frame within which research problems are generated (see also Kruse 2015: 109-10).
So in contrast to what the inductive logic implies, the abductively proceeding researcher does not enter the field value-or concept-free; rather, her research strategy is 'concept-driven' (see Friedrichs and Kratochwil 2009: 716-17). Yet, contrary to concepts being defined ex ante, they are specified and filled with 'content' during the course of the research process so that they align closely with the data generated in the field. Coined and advocated by the pragmatist Charles Sanders Peirce as an addition to deduction and induction, abductive reasoning is probably the most adequate logic to adopt for practice theorists since it constitutes a 'rule-governed way to new knowledge' (Reichertz 2010: 5). While it allows reconstructing the meaning-creating rules guiding a practice, it also incorporates the theorist's conceptual creativity in constructing a rule that is 'usable' (ibid: section 24) for the puzzle at hand, meaning that the identified rule fits the theorist's aim to understand and explain the identity of a not-yet-understood social practice. Accordingly, as Reichertz (ibid) notes, 'abductively discovered orders are neither (preferred) constructions nor (valid) reconstructions, but usable (re-) constructions'.
Here, the theorist neither pretends to 'objectively' uncover the meaningin-use of her research subjects nor a priori imposes her theoretical concepts. What she seeks to achieve instead is similar to what anthropologist Clifford Geertz (1973: 13, emphasis added) has proposed in The Interpretation of Culture: Finding our feet, an unnerving business which never more than distantly succeeds, is what ethnographic research consists of as a personal experience; trying to formulate the basis on which one imagines, always excessively, one has found them is what anthropological writing consists of as a scientific endeavor. We are not, or at least I am not, seeking either to become natives (a compromised word in any case) or to mimic them. Only romantics or spies would seem to find point in that. We are seeking […] to converse with them, a matter a great deal more difficult, and not only with strangers, than is commonly recognized.
From anthropology we thus learn that, at the end of the day, it is conversations with our research subjects that we should strive for in order to 'find our feet' -the latter being a phrase coined by Wittgenstein in Philosophical Investigations (Wittgenstein 1989(Wittgenstein [1953 II: 223). We should strive for conversations not only because it is in conversations that we speak to and encounter the research subject on an equal footing (as far as the power asymmetries allow for it). But also because it is an approach by which we as philosophers or IR scholars can most truthfully do 'justice' to both our research subjects and us as theorists (cf. Gunnell 2013: 99). It is in such conversations that we acknowledge that we cannot 'find our feet' unless we interpret and judge from an evaluative standpoint that is inherently linked to our own scientific practice. This standpoint is as much affected by our theoretical presuppositions as by our personal background or experiences as practitioners of a norm-and value-based research community. Greater engagement with their research subjects in the form of conversations would have helped Lechner and Frost clarify both the normativity of the practices of their research subjects and that of their own scientific practices.